JP4586272B2 - Substrate inspection apparatus and substrate inspection method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a substrate inspection apparatus and a substrate inspection method suitable for use in inspection of a large substrate, and a method for manufacturing a liquid crystal display device.
[0002]
[Prior art]
In general, a visible light microscope, an electron microscope (SEM: scanning electron microscope, TEM: transmission electron microscope), or the like is used for inspection of a substrate manufactured in the manufacturing process of a semiconductor device or a liquid crystal display device. The visible light microscope can cope with inspection from a wide area to a minute area, and the electron microscope can cope with inspection of a minute area even more than the visible light microscope.
[0003]
[Problems to be solved by the invention]
However, the visible light microscope is effective for low magnification and wide-range observation, but has limitations in terms of magnification and resolution. Therefore, it is not possible to sufficiently cope with observation of a minute region on the substrate surface.
[0004]
In addition, since an electron microscope has a relatively small sample chamber due to the necessity of vacuuming or the like, a large substrate used for a liquid crystal display device or the like cannot be observed as it is. Therefore, in the inspection, it is necessary to destroy the substrate and prepare a small sample. In addition, depending on the way of observation, sample pretreatment may be required. For this reason, inspection takes time, and only a very small part of the substrate is subject to inspection, so the distribution of the entire substrate cannot be obtained. Furthermore, the substrate to be inspected cannot be commercialized due to the use of sample destruction or pretreatment. Therefore, it is not suitable for inspection during the manufacturing process.
[0005]
The present invention has been made to solve the above-described problems, and an object of the present invention is to provide an apparatus capable of easily observing a minute area with sufficient resolution without destroying a large substrate and in a short time. And providing a method.
[0006]
[Means for Solving the Problems]
A substrate inspection apparatus according to the present invention includes a sample table that supports a large substrate, and deep ultraviolet light that irradiates the surface of the large substrate supported by the sample table with deep ultraviolet light to obtain a reflected image from the large substrate. Image acquisition means.
[0007]
In this substrate inspection apparatus, a sample stage for supporting a large substrate is provided, and the surface of the large substrate supported by the sample table is irradiated with deep ultraviolet light, and a reflected image from the large substrate is obtained by a deep ultraviolet light image acquisition means. By acquiring, it is possible to obtain an observation image with high magnification and high resolution by nondestructive inspection.
[0008]
In the substrate inspection method according to the present invention, the surface of a large substrate is irradiated with deep ultraviolet light to obtain a reflection image from the large substrate, and the surface state of the large substrate is inspected using the reflection image.
[0009]
In this substrate inspection method, by irradiating the surface of a large substrate with deep ultraviolet light to obtain a reflected image from the large substrate, an observation image with high magnification and high resolution can be obtained by nondestructive inspection. As a result, the surface state of the large substrate can be easily inspected in a short time.
[0010]
In the method for manufacturing a liquid crystal display device according to the present invention, the surface of a large liquid crystal display device substrate is irradiated with deep ultraviolet light to obtain a reflected image from the liquid crystal display device substrate, and a liquid crystal is obtained using the reflected image. It has a substrate inspection process for inspecting the surface state of the display device substrate.
[0011]
In this method of manufacturing a liquid crystal display device, the surface of a large liquid crystal display device substrate is irradiated with deep ultraviolet light to obtain a reflected image from the liquid crystal display device substrate, thereby enabling high magnification in nondestructive inspection. A high-resolution observation image can be obtained. This makes it possible to easily inspect the surface state of the large liquid crystal display device substrate in a short time.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0013]
FIG. 1 is a perspective view showing an overall image of a substrate inspection apparatus according to an embodiment of the present invention. In FIG. 1, a substrate inspection apparatus 1 is provided with a substrate inlet / outlet 2 through which a substrate to be inspected is carried in and out. A large substrate to be inspected (hereinafter referred to as a substrate to be inspected) is carried in and out while being accommodated in the cassette 3. The cassette 3 is set in the vicinity of the substrate entrance 2. A single cassette 3 accommodates a predetermined number (for example, 10) of substrates to be inspected in the vertical direction.
[0014]
Incidentally, in FIG. 1, two cassettes 3 are set side by side, but this is for accommodating uninspected substrates to be inspected and inspected substrates in separate cassettes 3. It is possible to cope with only one cassette 3.
[0015]
A substrate transfer robot 4 for loading and unloading the substrate to be inspected is disposed in the vicinity of the set position of the cassette 3. The substrate transfer robot 4 includes a robot hand 6 that supports the substrate 5 to be inspected, a robot arm 7 that transfers the substrate to be inspected supported by the robot hand 6, and a drive system that drives the robot arm 7. (Not shown).
[0016]
An inspection stage 8 is provided at the transfer destination of the substrate to be inspected by the substrate transport robot 4. The inspection stage 8 includes a table (sample stage) 9 that supports the substrate to be inspected in a mounting state, and a table driving unit 10 that supports the table 9 so as to be movable in three axial directions of X, Y, and Z. Yes. The table 9 supports the substrate to be inspected by a vacuum suction method.
[0017]
The inspection stage 8 is mounted on the upper surface portion of the surface plate 11. As a driving source of the table driving unit 10, an AC servo motor or a pulse motor that realizes highly accurate positioning can be used. The surface plate 11 is supported horizontally by four vibration isolation tables 12. Further, a portal column 13 is mounted on the surface plate 11 so as to straddle the inspection stage 8. An optical microscope unit 14 for substrate inspection is provided at the center of the column 13.
[0018]
The optical microscope unit 14 is for observing the surface state of the substrate to be inspected supported by the table 9 and has an observation function using deep ultraviolet light (DUV) and an observation function using visible light. FIG. 2 shows a specific configuration example of the optical microscope unit 14.
[0019]
In FIG. 2, the visible light observation unit 15 is provided with a visible light CCD camera 16, a red cut filter 17, a visible light illumination unit 18, and an autofocus unit (hereinafter referred to as AF unit) 19. A halogen lamp can be used as a visible light source. The visible light CCD camera 16 captures a visible light image (reflected image from the substrate to be inspected) obtained by light irradiation from the visible light illumination unit 18.
[0020]
A reflection mirror 20 and an objective lens 21 are arranged in the light emitting direction of the visible light illumination unit 18. The reflection mirror 20 has a characteristic of transmitting visible light and reflecting deep ultraviolet light, and is disposed close to the inspection target portion of the substrate 5 to be inspected. The objective lens 21 can be switched according to the type of irradiation light and the observation magnification.
Switching of the objective lens 21 is performed by an electric revolver.
[0021]
On the other hand, the deep ultraviolet light illumination system 22 is for irradiating deep UV light to the inspection object part. An optical fiber 23 for introducing light from a light source (not shown) is connected to the deep ultraviolet illumination system 22. As a light source of deep ultraviolet light, a laser light source, a gas sealing lamp, or the like can be used. A reflection mirror 24 is disposed in the light emission direction of the deep ultraviolet illumination system 22, and a polarization beam splitter 25 is disposed in the light reflection direction of the reflection mirror 24.
[0022]
A quarter-wave plate 26 is disposed between the polarization beam splitter 25 and the reflection mirror 20. The polarization beam splitter 25 reflects the reflected light from the reflection mirror 24 toward the reflection mirror 20 described above, and transmits the light from the reflection mirror 20 to the imaging lens unit 27. The imaging lens unit 27 includes two imaging lenses 27a and 27b having different magnifications, and means (not shown) for switching between these imaging lenses 27a and 27b. The imaging lenses 27a and 27b are switched electrically. The deep ultraviolet light CCD camera 28 captures a deep ultraviolet light image (reflected image from the substrate to be inspected) imaged by the imaging lens 27a or 27b.
[0023]
FIG. 3 is a functional block diagram of the substrate inspection apparatus according to the embodiment of the present invention. In FIG. 3, the visible light microscope 29, the deep ultraviolet light microscope 30, and the switching mechanism 31 thereof are configured by the optical microscope unit 14 described above. The visible light microscope 29 irradiates the surface of the substrate to be inspected, which is a sample, with visible light, thereby acquiring a reflected image (visible light image) reflected from the substrate to be inspected. The deep ultraviolet light microscope 30 irradiates the surface of the substrate to be inspected with deep ultraviolet light and thereby obtains a reflected image (deep ultraviolet light image) reflected from the substrate to be inspected.
[0024]
In the visible light microscope 29, by switching the magnification of the objective lens from 2.5 times to 100 times, it becomes possible to observe an enlarged image having a visual field of 1 mm square to 50 μm square and a magnification on the CRT of 700 times to 7000 times. On the other hand, in the deep ultraviolet microscope 30, by switching the magnification of the objective lens to 100 times and the magnification of the imaging lens from 1 to 4 times, the field of view is 50 μm to 12 μm square, and the magnification on the CRT is 5000 to 20000. Double-magnified images can be observed. Deep ultraviolet light has a shorter wavelength than visible light, and its wavelength range is generally defined in the range of 200 nm to 300 nm.
[0025]
The image processing unit 32 performs recording (filing), display, and digitization of an observation image obtained by the optical microscope unit 14. Image recording is performed using a hard disk or the like, and image display is performed using a CRT (cathode ray tube). Further, the digitization of the image is performed by arithmetic processing by a CPU (Central Processing Unit).
[0026]
The main controller 33 controls the operation and processing of the entire substrate inspection apparatus. In the main controller 33, the use of the visible light microscope 29 and the deep ultraviolet light microscope 30 is switched by outputting a switching signal to the switching mechanism 31. The main control device 33 instructs the image processing unit 32 to execute various processes, and performs distribution management, statistical processing, and the like using the digitized data obtained from the image processing unit 32. Further, the main control device 33 evaluates the surface state of the substrate to be inspected based on the digitized data obtained by the image processing unit 32, and determines pass / fail. Further, the main control device 33 stores coordinate data of a teaching position referred to at the time of board inspection, and gives a command for moving the teaching position to the position control device 34 according to the coordinate data.
[0027]
The position control device 34 outputs control data to the XY axis drive control device 35 and the Z axis drive control device 36 based on the movement command given from the main control device 33. The XY axis drive control device 35 controls the drive of the XY axis drive unit 37 that moves the table 9 that supports the substrate to be inspected in the XY axis direction (horizontal orthogonal biaxial direction). The drive of the Z-axis drive unit 38 for moving the table 9 in the Z-axis direction (vertical direction) is controlled.
[0028]
Subsequently, a method for manufacturing a liquid crystal display device using the substrate inspection apparatus having the above configuration will be described by taking as an example the case where the substrate to be inspected is a TFT substrate of a liquid crystal display device (TFT is an abbreviation for a film transistor). This manufacturing method is a process for manufacturing a liquid crystal display device using a low-temperature polysilicon TFT. For example, amorphous silicon is laminated (film-attached) on a glass substrate serving as a base of a TFT substrate by a CVD (Chemical-Vapor-Deposition) method. After that, the surface of the TFT substrate is crystallized by irradiating the surface with excimer laser annealing (hereinafter abbreviated as ELA) with a linear laser beam, thereby transforming amorphous silicon into polysilicon (Poly-Si). Applies to inspection. The TFT substrate to be inspected is, for example, a large substrate having a length of 600 mm, a width of 720 mm, and a thickness of 0.7 mm.
[0029]
The TFT substrate after the ELA is accommodated in the cassette 3 and carried into the substrate inspection apparatus 1 through the substrate entrance 2. Then, when the operator activates the apparatus via the operation panel or the like, the substrate is inspected according to the following procedure.
[0030]
First, the uninspected TFT substrate accommodated in the cassette 3 is supported by the substrate transfer robot 4 and transferred to the inspection stage 8 and placed on the table 9 waiting there. Next, the XY-axis drive control device 35 and the Z-axis drive control unit 36 control the driving of the XY-axis drive unit 37 and the Z-axis drive unit 38 in accordance with the teaching position movement command from the main control device 33, respectively. After 9 moves to the lower end of the Z-axis, it moves in the XY-axis direction. As a result, the observation target portion of the TFT substrate is positioned directly below the objective lens 21 of the optical microscope unit 14.
[0031]
Subsequently, the AF unit 19 focuses the visible light lens. As a result, the surface of the TFT substrate is irradiated with visible light, and a reflected image (visible light image) from the TFT substrate is captured by the visible light CCD camera 16. The observation image acquired using the visible light is used for macro observation of the TFT substrate (visual confirmation using a CRT screen or the like) and identification of an inspection position (observation position).
[0032]
Subsequently, the lens used for observation is switched to deep ultraviolet light. The switching of the lens is performed by the switching mechanism 31 based on a switching signal from the main control device 33. Next, by driving the Z-axis drive unit 38, the table 9 is lowered by a preset offset amount.
[0033]
Next, a reflection image (deep ultraviolet light image) from the TFT substrate obtained by deep ultraviolet light irradiation by the optical microscope unit 14 is captured by the deep ultraviolet light CCD camera 28. Next, after the Z-axis drive unit 38 is driven to lower the table 9 by a predetermined pitch amount, the deep ultraviolet CCD camera 28 captures an image again. Thereafter, the lowering of the pitch amount of the table 9 and the image capture by the deep ultraviolet CCD camera 28 are repeated a predetermined number of times. Thereby, the observation image is captured a plurality of times (for example, four times). The reason is that processing variations occur in the thickness of the TFT substrate, so that an observation image of deep ultraviolet light that is a just focus position within the range of the variations is reliably obtained.
[0034]
The observation images repeatedly captured in this way are sequentially sent to the image processing unit 32. Then, the image processing unit 32 extracts an observation image corresponding to the just focus position of the deep ultraviolet light lens from the plurality of observation images, and records, displays, and digitizes the extracted observation image. When this TFT substrate was inspected, a good observation image was obtained when deep ultraviolet light having a wavelength of about 250 nm was used.
[0035]
Here, the crystalline state of polysilicon by ELA depends on the appropriateness of ELA energy. Further, in the observation image using deep ultraviolet light, if the ELA energy is appropriate, a periodic pattern along the laser beam line is obtained. However, when the ELA energy is low, a random pattern is formed. Conversely, when the ELA energy is high, the periodicity of the pattern is destroyed.
[0036]
Therefore, in the present embodiment, the crystal state is evaluated based on the periodicity of the pattern on the polysilicon surface that appears in the observation image by irradiation with deep ultraviolet light. Regarding the periodicity of the pattern, the image processing unit 32 performs image processing on the previously extracted observation image, for example, the degree of coincidence (similarity) of the pattern of a certain local portion with respect to the pattern of another local portion. (Auto-Correlation) can be expressed as a numerical value.
[0037]
The data digitized by the image processing unit 32 is sent to the main controller 33, where the crystal state of the polysilicon surface of the TFT substrate that is the substrate to be inspected is evaluated. FIG. 4 is a pattern drawn based on the data obtained by digitizing the observation image of the polysilicon surface by image processing. (A) is a pattern image having periodicity, and (B) is a pattern image having no periodicity. Represents. By scanning this pattern image in a direction perpendicular to the line direction of the laser beam in ELA, a numerical value representing the periodicity of the pattern is obtained by the autocorrelation method.
[0038]
If the numerical value obtained by the autocorrelation method is an AC value, this AC value is defined as “AC value = R1 / R0 (%)” as shown in FIG. As a result, the AC value (= R1 / R0) increases as shown in FIG. 6A in the pattern having periodicity, and the AC value decreases in the pattern having no periodicity as shown in FIG. 6B.
[0039]
In view of this, the main controller 33 quantitatively evaluates the crystalline state of the polysilicon surface by comparing the level of the numerical value (AC value) obtained as described above with, for example, a preset threshold level. Judgment is made. The TFT substrate that has been inspected is moved to the original position by the XY-axis drive unit 37 and the Z-axis drive unit 38, transferred to the substrate transfer robot 4, and then accommodated in the inspected cassette 3. Thereafter, the next TFT substrate is inspected by the same operation. When a predetermined number of substrates have been inspected, the inspected cassette 3 is carried out of the apparatus through the substrate entrance 2.
[0040]
Thus, by inspecting the TFT substrate using deep ultraviolet light having a short wavelength, it is possible to obtain an observation image with a higher magnification and higher resolution than when visible light is used. Thereby, it is possible to quantitatively evaluate the crystal state (crystallinity) of the polysilicon surface, which was difficult to discriminate with a visible light microscope. In addition, since the crystal state can be appropriately evaluated without relying on visual judgment with ambiguous criteria, the accuracy and reliability of inspection can be improved.
[0041]
Furthermore, since the TFT substrate can be inspected in the atmosphere, when actually inspecting the substrate, the TFT substrate is supplied to the inspection stage 8 as it is, and an observation image of the polysilicon surface (deep ultraviolet light) is obtained by the optical microscope unit 14. Image). Therefore, it is not necessary to perform destruction or pretreatment for sampling. Thus, by realizing non-destructive inspection of the TFT substrate, the inspection time can be greatly shortened. Furthermore, a liquid crystal display device can be commercialized using the TFT substrate that has been inspected.
[0042]
Further, by obtaining the correlation between the crystalline state of the polysilicon surface and the ELA energy and fordifying the result into the ELA treatment, the ELA energy can be optimized. Furthermore, it is possible to improve the yield by determining pass / fail based on the crystal evaluation of polysilicon and supplying only the TFT substrate that has passed to the next process. In addition, by distribution management and statistical processing in the main controller 33, it is possible to compare the location dependence of the micro area observation results in the TFT substrate, or to compare the observation results at the same location of a plurality of TFT substrates. It becomes.
[0043]
In the above embodiment, an observation image of one minute region specified by the instruction position movement command from the main controller 33 is obtained. For example, as shown in FIG. Along with the image of the teaching position P1 specified by the XY coordinates from P0, the images of the peripheral portions P2 to P6 are acquired at a plurality of locations by constant pitch feeding in the XY axis direction, and these images are individually digitized to obtain the numerical values. A representative value (average value, variation, bias, etc.) of the digitized data may be obtained.
[0044]
Further, as shown in FIG. 7B, in addition to acquiring images of the teaching position P1 specified by the XY coordinates and the peripheral portions P2 to P6 from the substrate origin P0, the teaching position P1 is moved in the XY axis direction. Similarly, an offset may be provided, and images of a plurality of places P7 to P12 may be acquired in the same manner to obtain representative values.
[0045]
Thus, by observing the surface of the substrate at multiple points with the teaching position as the center (starting point), macroscopic evaluation can be performed even with a deep ultraviolet light microscope with a narrow field of view. As a result, for example, even when a unique point exists only in the minute region indicated by the teaching position, it is not evaluated only by the digitized data of the image, so that the surface state of the TFT substrate can be grasped more appropriately. It can be reflected in the crystallinity evaluation.
[0046]
In addition, the substrate inspection apparatus and the substrate inspection method according to the present invention include a defect inspection of a large substrate including a TFT substrate, a foreign matter inspection, a line width of a linear unevenness processed on a substrate by a semiconductor process, and a hole processing. It can be applied to the measurement of the diameter and area.
[0047]
【The invention's effect】
As described above, according to the substrate inspection apparatus of the present invention, the sample table is provided to support the large substrate, and the surface of the large substrate supported by the sample table is irradiated with deep ultraviolet light to reflect the large substrate. Is acquired by the deep ultraviolet light image acquisition means, an observation image with high magnification and high resolution can be obtained by nondestructive inspection. As a result, it is possible to easily perform a micro area inspection of a large substrate in a short time. In addition, it is possible to continuously process and process a large substrate that has already been inspected to produce a product.
[0048]
Further, according to the substrate inspection method of the present invention, a high-magnification and high-resolution observation image is obtained by nondestructive inspection by irradiating the surface of a large substrate with deep ultraviolet light to obtain a reflected image from the large substrate. Is obtained. Therefore, the surface state of the large substrate can be easily inspected in a short time. In addition, the inspected large substrate can be continuously processed and processed for commercialization.
[0049]
Further, according to the method for manufacturing a liquid crystal display device of the present invention, the surface of a large liquid crystal display device substrate is irradiated with deep ultraviolet light to obtain a reflected image from the liquid crystal display device substrate. A high-resolution and high-resolution observation image is obtained in the inspection. Therefore, the surface state of the large liquid crystal display device substrate can be easily inspected in a short time. In addition, the inspected liquid crystal display device substrate can be continuously processed and processed into a product.
[Brief description of the drawings]
FIG. 1 is a perspective view showing an overall image of a substrate inspection apparatus according to an embodiment of the present invention.
FIG. 2 is a schematic diagram illustrating a specific configuration example of an optical microscope unit.
FIG. 3 is a functional block diagram of a substrate inspection apparatus according to an embodiment of the present invention.
FIG. 4 is a diagram showing a pattern image of a substrate surface obtained by image processing.
FIG. 5 is a diagram (part 1) for explaining a polysilicon crystal evaluation algorithm;
FIG. 6 is a diagram (part 2) for explaining a polysilicon crystal evaluation algorithm;
FIG. 7 is a diagram illustrating an example of setting specific observation positions for multi-point observation.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Board | substrate inspection apparatus, 3 ... Cassette, 4 ... Board | substrate conveyance robot, 5 ... Board | substrate to be inspected (large board | substrate), 8 ... Inspection stage, 9 ... Table, 10 ... Table drive part, 14 ... Optical microscope unit, 29 ... Visible Light microscope, 30 ... deep ultraviolet light microscope, 21 ... switching mechanism, 32 ... image processing unit, 33 ... main controller, 34 ... position controller

Claims (5)

A sample stage for supporting a large substrate;
A deep ultraviolet light image obtaining means for obtaining a reflected image of the deep ultraviolet light to the supporting surface of the large substrate to specimen table from the large-type substrate by irradiating,
Image digitizing means for digitizing the image acquired by the deep ultraviolet light image acquiring means;
Evaluation means for evaluating the surface condition of a large substrate based on the digitized data obtained by the image digitizing means;
With
The deep ultraviolet light image acquisition means uses a polysilicon substrate formed by annealing the surface of the amorphous silicon substrate as a large substrate, and irradiates the surface with deep ultraviolet light to acquire a reflection image from the polysilicon substrate. And
The image digitizing means quantifies the periodicity of the pattern on the surface of the polysilicon substrate by expressing the degree of coincidence of the pattern of the other local part with the pattern of the local part of the reflected image by a numerical value called an autocorrelation function,
The evaluation means is a substrate inspection apparatus that evaluates the crystal state of the surface of the polysilicon substrate based on a numerical value representing the periodicity of the pattern on the surface of the polysilicon substrate . And the visible light image obtaining means for irradiating visible light to obtain a reflected image from the large-substrate to the support surface of the large substrate to specimen stage,
Substrate inspection apparatus according to 請 Motomeko 1 using you and a switching means for switching the deep ultraviolet light image obtaining means and availability Miko image acquiring means. Get the reflected image of the polysilicon substrate is irradiated with deep ultraviolet light to form polysilicon substrate surface by performing an annealing treatment on the surface of the amorphous silicon substrate, other with respect to the pattern of the local portion of the this reflected image A substrate inspection method for inspecting the crystal state of the surface of the polysilicon substrate by expressing the degree of coincidence of the pattern of the local portion of the substrate by a numerical value called an autocorrelation function and by digitizing the periodicity of the pattern on the surface of the polysilicon substrate. A request to inspect the crystal state of the surface of the polysilicon substrate by acquiring reflection images over a plurality of locations on the surface of the polysilicon substrate, individually digitizing based on each reflection image, and obtaining a representative value of each digitized data Item 4. The substrate inspection method according to Item 3. A reflection image from the polysilicon substrate is obtained by irradiating the surface of the polysilicon substrate for a liquid crystal display device formed by annealing the surface of the amorphous silicon substrate with deep ultraviolet light, and the reflection image is locally present. The degree of coincidence of the pattern of another local part with the pattern of the part is expressed by a numerical value called an autocorrelation function, and the crystal state of the surface of the polysilicon substrate is inspected by quantifying the periodicity of the pattern on the surface of the polysilicon substrate. method of manufacturing a liquid crystal display device that have a substrate inspection process.

Images