JPH0972722A - Substrate visual inspection device - Google Patents

Abstract

(57) [Abstract] [PROBLEMS] To reliably inspect defects on a substrate surface. A parallel light generation unit 8 generates parallel light based on light from an illumination unit, and the parallel light is vertically incident on a surface of a substrate via a half mirror 9. Then, the reflected light from the substrate 4 is imaged by the CCD camera 6, the number of defect detection pixels px in an abnormal portion of the imaged image is counted, and this is converted into a defect volume v to determine the size of the defect. . Also, while performing general diffuse illumination on the surface of the substrate 4,
The surface of the substrate 4 is imaged by the CCD camera 11, and the defect area A is obtained from the imaged image. Then, the defect height or the defect depth d = v / (k · A) is calculated, and the defect determination is continued.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a substrate appearance inspection capable of optically non-contact and high-speed measurement of three-dimensional shape such as fine defects and patterns on a substrate surface and positional information regardless of the substrate surface reflectance. Regarding the device.

[0002]

2. Description of the Related Art Conventionally, when performing a visual inspection of a board, CC
The surface of a substrate is photographed by a D camera, and the presence / absence and size of a defect are determined from the image information.
It is also practiced to irradiate the surface of a substrate with a laser beam, detect the defect by scattering, diffraction, and reflection, and measure the shape thereof.

Furthermore, detailed three-dimensional shape measurement of defects is also performed by STM (scanning tunneling microscope) and AFM (atomic force microscope).

[0004]

However, CCDs
In normal shooting with a camera, the area of the defect can be calculated from the image information, but the height or depth of the defect and the volume of the defect must be known. I couldn't. In addition, it is difficult to detect defects in a substrate having a low surface reflectance by the method using laser light irradiation.

Further, fine defects and patterns on the surface of the substrate can be measured by STM or AFM, but they are unsuitable when the inspection time is long and high-speed measurement is required. In view of such conventional problems, an object of the present invention is to provide a substrate appearance inspection device capable of inspecting minute defects on the substrate surface at high speed and accurately.

[0006]

Therefore, in the invention according to claim 1, as shown in FIG. 1 (A), parallel light incidence means for injecting parallel light in the vertical direction onto the surface of the substrate and the parallel light incidence means. Provided is a reflected light image pickup means for picking up an image of the reflected light of the light from the substrate, and a defect judgment means for judging the size of the defect based on the image of the abnormal portion picked up by the reflected light image pickup means.
A board appearance inspection device is configured.

Here, the abnormal portion is an uneven portion derived from a defect. The incident light is not specularly reflected but the reflected light is condensed and looks brighter than the normal portion, or the reflected light is irregularly reflected and normal. The part that appears darker (shadow) than the part. Further, the invention according to claim 2 is characterized in that the defect determining means determines the size of the defect by converting the image signal of the abnormal portion into a defect volume.

The inventors of the present invention have conducted experiments to generate parallel light, irradiate the surface of the substrate in a vertical direction, and analyze the reflected light, specifically, image it with a two-dimensional CCD camera, and capture it. The image thus obtained was binarized by an image processing device, and the number of pixels detected in the abnormal portion was counted. This is because it is inferred that if there is an uneven defect on the surface of the substrate, only that part is not specularly reflected, so that there is a certain relationship between the defect and the number of detected pixels in the abnormal part. According to this experiment, the number of defect detection pixels does not correlate with the actual defect area, and the actual defect volume (=
It was found that there is a strong correlation with [defect area] × [defect height or depth]).

Therefore, image information is obtained using parallel light,
By counting the number of defect detection pixels in the abnormal portion, the size of the defect is determined based on the number of defect detection pixels in the abnormal portion.Furthermore, the defect detection pixel number in the abnormal portion is converted into a defect volume. , Determine the size of the defect. In the invention according to claim 3, as shown in FIG. 1 (B), parallel light incidence means for injecting parallel light onto the surface of the substrate in the vertical direction, and reflected light for imaging the reflected light of the parallel light from the substrate. Image pickup means, defect volume measurement means for converting an image signal of an abnormal portion picked up by the reflected light image pickup means into a defect volume, and auxiliary image pickup means for picking up the surface of the substrate while performing general diffuse illumination on the surface of the substrate. A defect area measuring unit that captures an image of the abnormal portion imaged by the auxiliary imaging unit and measures the area thereof, and a defect that calculates a defect height or a defect depth from the defect volume and the defect area of the abnormal portion. A substrate appearance is provided by providing height / depth calculation means and defect determination means for determining the size of a defect based on at least one of the defect volume and defect area of the abnormal portion and the defect height or defect depth. Configure an inspection device.

Further, in the invention according to claim 4, the defect height / depth calculating means calculates the defect height or the defect depth by dividing the defect volume by the defect area. And That is, if the surface of the substrate is imaged while performing general diffuse illumination on the surface of the substrate, the defect area can be measured from the imaged image. Here, the measurement magnification is selected according to the target defect size. At this time, by determining the defect detection pixel number and calculating the absolute coordinates of the defect at the same time, positioning at the defect location can be easily performed regardless of the magnification. Therefore, the defect detection pixel number is obtained from the image information of the abnormal portion obtained by using the parallel light, and this is converted into the defect volume and divided by the defect area to calculate the defect height or the defect depth equivalent value. be able to. Therefore, the size of the defect is determined from this defect height or defect depth, or in consideration of this.

In the invention according to claim 5, at least the surface of the substrate holding member for holding the substrate is made of a material having a low reflectance. That is, in order to prevent image noise due to diffused reflection from the surface of the substrate holding member during inspection, a member having a low reflectance is used for the substrate holding member or is covered with a low reflectance material.

In the invention according to claim 6, further, means for adjusting the distance between the reflected light image pickup means and the substrate surface, and the distance from the image pickup center to the abnormal portion are measured, and a weighting coefficient according to the distance is set. And means for correcting the size of the defect. That is, in the inspection apparatus, in order to keep the detection sensitivity of the abnormal portion constant regardless of the position in the measurement visual field, the known abnormal portion is brought to an arbitrary position in the measurement visual field, specifically, the hard disk. In the case of a doughnut-shaped substrate such as a substrate for measurement, the spindle that supports the inner diameter of the substrate is rotated with the optical system position (measurement field of view) fixed, and the detection position of the abnormal part from the center of the measurement field (optical axis). (Here, the absolute coordinate axes are set in advance so that the position of the defect can be measured.) The focus position of the inspection apparatus is adjusted so that the same number of defect detection pixels can be obtained regardless of the distance to the defect position. That is, the distance between the optical system and the surface to be measured is adjusted.

Further, when the adjustment of the focus position is not sufficient, the number of defect detection pixels in which a weighting factor corresponding to the attenuation amount of the detection sensitivity due to the distance from the center of the measurement visual field to the detection position of the abnormal portion is actually measured. Correct by multiplying by. In the invention according to claim 7, there is further provided means for measuring the dimension of the substrate or the substrate edge portion or detecting the abnormal portion of the substrate edge portion based on the edge image of the substrate imaged by the reflected light image capturing means.

That is, the edge image is binarized by an image processing device to create an edge-processed binarized image. Since absolute coordinates are set in the binarized image and the position of the edge line can be easily measured, the dimension of the substrate is measured simultaneously with the defect inspection, which is the main function of the apparatus. In addition, by performing image processing by comparing the standard image of the board edge portion with the actually captured image, the shape of the board edge portion, specifically, the chamfered shape or chamfered shape of the board edge portion, the board edge portion The abnormal portion, specifically, the edge portion of the substrate is detected, and the abnormal shape portion is detected simultaneously with the defect inspection, which is the main function of the apparatus.

In the invention according to claim 8, the substrate is a carbon substrate. Carbon substrate
Compared to aluminum and glass substrates, it is lighter, has superior shock resistance and heat resistance, and can be made thinner, and is extremely excellent as a substrate.However, because of its low reflectance, defects can be found by ordinary methods. It was difficult, but with this inspection device,
Defect inspection becomes easy.

Even when the surface treatment / treatment is performed on the carbon substrate, specifically, texture / grinding / polishing
Even when thin film formation, etching, and pattern formation are performed, defect inspection is possible if the surface reflectance is several percent or more. By adding a means for measuring the reflectance of the surface to be measured and a means for controlling the amount of incident light according to the reflectance, and by the reflected light image pickup means, the amount of reflected light imaged can always be kept at an optimum value. This is because the same defect detection sensitivity can be obtained regardless of the surface reflectance (several percent or more) of the substrate.

Therefore, the substrate is not limited to a carbon substrate, but a magnetic disk substrate, a semiconductor wafer, a liquid crystal substrate,
For optical discs, stampers, magnetic tapes, floppy discs, metal substrates, ceramic substrates, plastic substrates, films, etc., the surface to be measured is flat and the surface reflectance is white light or single light or several% or more in a specific wavelength range. If

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below. Here, the hard disk substrate is the inspection target. FIG. 2 is a system diagram of a board appearance inspection apparatus showing an embodiment of the present invention. The table 1 is movable in the X direction in the figure, and a spindle 3 rotatable by a pulse motor 2 is provided on the table 1. The spindle 3 holds the hard disk substrate 4 to be inspected horizontally.

Here, a low-reflectance member is used for the spindle 3 which is a holding member for the substrate 4, or a low-reflectance material is coated (blackened) to prevent irregular reflection from the surface. It is preferable. Further, in this example, the inner diameter of the substrate 4 is held, but the outer diameter may be held. As for the side surface shape of the portion of the spindle 3 above the surface of the substrate 4, it is necessary to measure the edge of the inner diameter (or outer diameter) of the substrate 4 held thereby, so that the incident and reflected optical paths are Attach an appropriate inclination as shown in the figure to avoid interference.

A CCD camera 6 is fixed via a lens unit 5 above one side in the moving direction of the table 1 with the central portion (spindle 3) of the substrate 4 to be inspected in between. The lens unit 5 includes a parallel light generation unit 8 that generates parallel light based on light from the external illumination unit 7.
And a half mirror 9 that directs the parallel light from the parallel light generation unit 8 downward and makes it incident on the surface of the substrate 4 as incident parallel light. The reflected light of the incident parallel light from the substrate 4 passes through the half mirror 9 and enters the CCD camera 6. The illumination unit 7 can control the amount of light according to the surface reflectance of the substrate 4 to be inspected.

Therefore, the parallel light generating section 8 and the half mirror 9 constitute a parallel light incidence means for making parallel light incident on the surface of the substrate 4 in a vertical direction, and the CCD camera 6 reflects the parallel light reflected from the substrate 4. And a reflected light image pickup means for picking up an image.
A CCD camera 11 is fixed via a general-purpose high-magnification lens 10 above the other side in the moving direction of the table 1 with the central portion (spindle 3) of the substrate 4 to be inspected interposed therebetween.

On the side of the general-purpose high-magnification lens 10, a diffuse illuminator 13 for performing general diffuse illumination by light from an external illumination unit 12 is provided. Therefore, the CCD camera 11 and the diffuse illuminator 13 constitute an auxiliary image pickup means for taking an image of the surface of the substrate 4 while performing general diffuse illumination on the surface of the substrate 4. The image signals obtained by the two CCD cameras 6 and 11 are input to the image processing device 14 and the monitor 15.

In the image processing device 14, absolute coordinates are set on the surface of the substrate 4, and the abnormal portion obtained by the lens unit 5 (parallel light generating portion 8, half mirror 9), CCD camera 6, and illumination unit 7. The defect position is obtained from the image signal of. General-purpose high-power lens based on this defect position information
10, CCD camera 11, illumination unit 12, diffuse illuminator 13
When a defect is detected by (this system preferably has an automatic focusing function), high-speed positioning is possible by moving the table 1 in the X direction and rotating the pulse motor 2.

When inspecting the entire surface of the hard disk substrate 4, the spindle 3 is rotated by a predetermined angle by the pulse motor 2 and, for example, as shown in FIG.
In the case of a disc, it is divided into 12 images with a field-of-view size of 18.5 mm x 13.9 mm and captured. Further, when the diameter of the substrate 4 is large, the field of view size is changed or the table 1 is moved, and the spindle 3 is rotated by a predetermined angle by the pulse motor 2 for each moving position, as shown in FIG. 4, for example. As described above, in the case of a 2.5 "disc, the image is divided into 12 images on the inner circumference side and 15 images on the outer circumference side with the same visual field size.

When the images are divided and imaged in this way, portions of the image are duplicated. However, it is preferable to mask the images as shown in FIG. 5 during image processing so that the images do not overlap as much as possible. Fig. 5 (A) shows the mask area of the inner peripheral portion of the 1.89 "disc or 2.5" disc, and Fig. 5 (B) shows 2.
The mask area on the outer circumference of the 5 "disk is shown. However, if the overlapping area is completely eliminated, an uninspected area may occur. Therefore, Fig. 6 shows an example of the 1.89" disk. As shown, a minimum overlap area is provided between adjacent measurement images. At this time, although there is a low probability, an abnormal portion may exist in the overlapping area. For example, the abnormal portion B in FIG. 6 is detected as an abnormal portion in both the image 1 and the image 2 adjacent to the abnormal portion B, resulting in duplicate detection. Therefore, if any abnormal part is detected redundantly at the same coordinates in the overlapping area, the defective data from the second time onward is deleted, or the average / maximum value of the duplicated data is deleted.
It is preferable to have a duplicate data editing function such as taking one of the minimum values.

Further, in the inspection apparatus, in order to make the detection sensitivity of the abnormal portion constant at any position in the measurement visual field, the known abnormal portion is moved to an arbitrary position in the measurement visual field while the optical system is moved. With the position (measurement field of view) fixed,
By rotating the spindle 3 that supports the inner diameter of the substrate 4, the focus is adjusted so that the same defect detection pixel number can be obtained without depending on the distance from the measurement visual field center (optical axis) to the known abnormal portion detection position. Focus position,
That is, the distance between the optical system and the surface to be measured is adjusted.

Further, when the adjustment of the focus position alone is insufficient, the number of defect detection pixels for which the weighting coefficient is actually measured according to the attenuation amount of the detection sensitivity depending on the distance from the center of the measurement visual field to the detection position of the abnormal portion. Should be corrected by multiplying by. FIG. 11 shows the variation in the number of defect detection pixels with respect to the distance from the center of the measurement field of view, as compared with the case of the conventional method (optical adjustment by monitor observation), and the focus position adjustment by image binarization observation (the present invention (A) method). And the effect of performing the correction by the weighting coefficient (the method of the present invention (B)) after the focus position adjustment.

In the image processing apparatus 14, not only the images from the CCD cameras 6 and 11 are subjected to image processing, but also defect determination is performed for each corresponding image according to the flowchart of FIG. This will be described according to the flowchart of FIG. This example is a one-sided inspection of a 1.89 "size hard disk carbon substrate.

In step 1 (denoted as S1 in the figure, the same applies hereinafter), each variable is initialized. That is, initialization of the measurement field of view n (n = 1), initialization of the sum total of the dirty error counts mpsum (mpsum = 0), and the number of defect detection pixels px which is a defect data array for temporarily storing the defect data group px , Defect detection radius r 1, defect detection angle th 1, defect volume v 2, defect area A
, The defect height or the defect depth d 1 and the 30% error count mp are secured.

In step 2, the defect detected based on the abnormal portion of the image of the measurement field of view n from the CCD camera 6 is labeled, and the number of defect detection pixels px i and the defect detection are performed for each defect i. The radius r i and the defect detection angle th i are obtained. At this time, the processing of each mask and the duplicated data editing function, and the correction by the weighting coefficient are also performed.

At step 3, it is selected whether or not the defect height or the defect depth is calculated. If this calculation is performed, the processing from step 4 to step 8 is performed.
In step 4, for each of the defects i detected in the measurement visual field n, based on the image from the CCD camera 11,
(Here, the table 1 and the pulse motor 2 which are high-speed positioning mechanisms are driven.) The defect area A i is estimated by image analysis. This part corresponds to a defect area measuring means.

In step 5, the defect detection pixel number px i of each defect i is converted into a defect volume v i according to the formula (1) or the formula (1 ′). (Here, due to the characteristics of the device, there is a strong correlation between the defect detection pixel number and the defect volume as shown in FIG. 8.) This conversion is performed for all defects i. This part corresponds to the defect volume measuring means. v i = 10 ^{((px i +63.608) /44.917)} (1) or px i = 44.917 × log ₁₀ v i −63.608 (1 ') [(1') Relation number 0.999635] The defect volume v on the horizontal axis of FIG.
It is the AFM actual measurement value by Nanoscope III manufactured by the company.

In step 6, for each defect i, the defect volume v i is divided by the defect area A i according to equation (2) to calculate the defect height or defect depth d i.
At this time, the characteristic constant k is selected depending on how the defect shape model is imagined. For example, in the case of a quadrangular pyramid approximation model as shown in FIG. 9, k = 1/3. This portion corresponds to a defect height / depth calculating means.

D i = v i / (k · A i) ... (2) In step 7, the defect height or defect depth d i of each defect i is compared with a predetermined value 1, and d i ≦ If the predetermined value is 1, it is determined to be OK and the defective data is eliminated (step 8). This portion corresponds to a defect determination unit. In step 9, for each defect i judged to be NG because d i> predetermined value 1 in step 7, the defect detection pixel number px i is calculated according to the equation (3) to the search error count mp i.
Convert to (Here, when the device is applied to a hard disk substrate, due to its characteristics, there is a strong correlation between the number of detected pixels and the error count, as shown in FIG. 10.) mp = 0.013 × px i ² +0.0312 × px i (3) In step 10, the sum Σmp i of the search error counts mp i in the measurement visual field n is calculated, and mpsum is calculated.
Is added to.

At step 11, 1 is added to the value of the measurement visual field n. In step 12, all measurement fields (n = 1 to 12)
Is calculated or not. If all inspections have been completed, proceed to step 13. If not, the process returns to step 2 and the same inspection is performed. Step 13
Then, the total sum of the certifying error counts mpsum is compared with a predetermined value 2, and when mpsum ≦ predetermined value 2, OK
(Step 14), NG if mpsum> predetermined value 2
(Step 15) This portion also corresponds to a defect determination unit.

Further, when the one-sided inspection is completed, the other side may be inspected by the same inspection as described above.
Further, in the above-described inspection method, since the substrate defect inspection data can be converted into a 30-error count indicating the final quality of the hard disk medium, the substrate appearance inspection can be reliably performed. At the same time as the board appearance inspection, C
Based on the image (edge image of the substrate) from the CD 6, it is possible to measure the substrate dimensions and detect an abnormal portion of the substrate edge portion.

Table 1 shows the results of measuring defects of the Cr film-formed hard disk substrate using the inspection apparatus according to the present embodiment. Here, the results of measuring the same sample with a MAXIM NT roughness meter manufactured by ZYGO are also shown.
It was found that the measurement results of both agree very well. In addition, while MAXIM NT required several minutes to measure one abnormal portion, the inspection apparatus according to this example was able to perform measurement in 10 seconds per one side of the substrate.

[0038]

[Table 1]

In this embodiment, the substrate (inspection object) 4, the lens unit 5, the CCD cameras 6, 11, the illumination units 7, 12, the general-purpose high-magnification lens 10, the image processing device 14 are used.
The following was used as Substrate (Inspection target): 1.89 "carbon substrate for hard disk 6.0 mm inner diameter, 24.0 mm outer diameter, 25 mils thickness Reflectance approx. 20% (400-800 nm) Ra 1 nm or less, flatness several μm or less Lens unit (vertical epi-illumination device) ): Dynascope New Creation DS-18
5A Measurement field size 18.5mm x 13.9mm CCD camera: Sony XC-75CE (two-dimensional type) Illumination unit: Hoyashot HL100E (100W halogen lamp) General-purpose high-magnification lens: Microscope 50-400x Image processor: Takadake Manufactured by GV-3000-R

[0040]

As described above, according to the first or second aspect of the present invention, image information is obtained by using parallel light, and the image information of the abnormal portion is obtained, and further, the defect volume is obtained. The effect of being able to determine the size of the defect at high speed and accurately is obtained.

According to the invention of claim 3 or 4, it is possible to further calculate the defect height or the defect depth, and it is possible to obtain the effect that the defect can be determined more accurately from this value. According to the invention of claim 5, it is possible to obtain the effect that the image noise of the device can be reduced and the defect can be detected more accurately.

According to the invention of claim 6, it is possible to detect a defect more accurately because it is possible to perform measurement with the same detection sensitivity regardless of the position of the abnormal portion in the measurement visual field of the apparatus. . According to the invention of claim 7, it is possible to obtain the effect of simultaneously performing the defect inspection, which is the main function of the apparatus, the substrate dimension measurement, the substrate edge portion dimension measurement, and the abnormal portion of the substrate edge portion.

According to the eighth aspect of the present invention, since an apparatus that is particularly excellent for visual inspection of a carbon substrate having a low surface reflectance is provided, it can be said that an environment in which the usefulness of the carbon substrate is exhibited more effectively has been prepared. .

[Brief description of drawings]

FIG. 1 is a functional block diagram showing the configuration of the present invention.

FIG. 2 is a system diagram of a board appearance inspection device showing an embodiment of the present invention.

FIG. 3 is a diagram showing an image division example 1.

FIG. 4 is a diagram showing an image division example 2;

FIG. 5 is a diagram showing an example of a mask.

FIG. 6 is a diagram showing an example of an overlapping area between adjacent measurement images.

FIG. 7 is a flowchart for defect determination.

FIG. 8 is a diagram showing the correlation between the number of detected pixels and the defect volume.

FIG. 9 is a diagram showing a defect shape model.

FIG. 10 is a diagram showing the correlation between the number of defect detection pixels and the 30 error count.

FIG. 11 is a diagram showing an effect of defect detection sensitivity correction.

[Explanation of symbols]

 1 table 2 pulse motor 3 spindle 4 substrate (inspection target) 5 lens unit 6 CCD camera 7 illumination unit 8 parallel light generation unit 9 half mirror 10 general-purpose high-magnification lens 11 CCD camera 12 illumination unit 13 diffusion illuminator 14 image processing device 15 monitor

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G11B 5/84 G06F 15/70 355

Claims (8)

[Claims] 1. A parallel light incidence means for injecting parallel light into a surface of a substrate in a vertical direction, a reflected light imaging means for imaging reflected light of the parallel light from the substrate, and an image taken by the reflected light imaging means. A substrate appearance inspection apparatus including: a defect determination unit that determines the size of a defect based on an image of an abnormal portion. 2. The substrate appearance inspection apparatus according to claim 1, wherein the defect determining means converts the image signal of the abnormal portion into a defect volume and determines the size of the defect. 3. A parallel light incident means for making parallel light incident on a surface of a substrate in a vertical direction, a reflected light image pickup means for picking up reflected light of the parallel light from the substrate, and an image picked up by the reflected light image pickup means. Defect volume measuring means for converting the image signal of the abnormal portion into a defect volume, auxiliary imaging means for imaging the surface of the substrate while performing general diffuse illumination on the surface of the substrate, and an image of the abnormal portion imaged by this auxiliary imaging means A defect area measuring means for taking in and measuring the area thereof; a defect height / depth calculating means for calculating a defect height or a defect depth from the defect volume and the defect area of the abnormal portion; A substrate visual inspection apparatus comprising: a defect determination unit that determines a defect size based on at least one of a defect volume, a defect area, a defect height, and a defect depth. 4. The substrate according to claim 3, wherein the defect height / depth calculation means calculates the defect height or the defect depth by dividing the defect volume by the defect area. Appearance inspection device. 5. The substrate visual inspection apparatus according to claim 1, wherein at least a surface of the substrate holding member holding the substrate is made of a material having a low reflectance. 6. A means for adjusting a distance between the reflected light image pickup means and a substrate surface, a distance from an image pickup center to an abnormal portion is measured, and a size of a defect is corrected by using a weighting coefficient according to the distance. The device for inspecting a substrate according to any one of claims 1 to 5, further comprising: 7. A means for measuring dimensions of a substrate or a substrate edge portion or detecting an abnormal portion of the substrate edge portion based on an edge image of the substrate imaged by said reflected light image pickup means. Claim 1-Claim 6 characterized
The board appearance inspection device according to any one of the above. 8. The substrate visual inspection apparatus according to claim 1, wherein the substrate is a carbon substrate.