JP2012037425A - Method for inspecting polycrystal silicon wafer and device thereof - Google Patents

Abstract

An object of the present invention is to reduce the contrast of a polycrystalline silicon wafer and to inspect the polycrystalline silicon wafer surface properly without over-detecting foreign matter on the surface of the polycrystalline silicon wafer. Is to provide.
A method for inspecting a polycrystalline silicon wafer according to the present invention includes: an imaging unit provided on a cover with the illuminated wafer surface indirectly illuminating the wafer surface by irregularly reflecting light from a light source on the inner surface of the cover; In this case, an image is acquired by an imaging means provided outside the cover, and a defect on the wafer surface is detected from the image.
[Selection] Figure 3

Description

  The present invention relates to a method and apparatus for inspecting a polycrystalline silicon wafer, and more particularly to a method and apparatus for inspecting a surface defect of a polycrystalline silicon wafer manufactured for a solar cell panel.

  Polycrystalline silicon is used as a material for various electronic components. In particular, in recent years, in the production of solar cells, solar cells using a polycrystalline silicon wafer as a material are becoming mainstream from the viewpoint of a good balance between cost and performance.

The polycrystalline silicon wafer is formed by cutting a silicon ingot into a certain thickness with a wire saw or the like.
At that time, fine foreign substances (diameter of about 100 to 1000 μm in diameter) such as carbon and silicon nitride may be mixed in the silicon ingot, and these may be cut out together with the wafer.
Although the cut wafer is cleaned in a cleaning process, foreign matter may remain on the wafer even after cleaning, and the characteristics of a solar cell manufactured using the wafer with such foreign matter remaining deteriorate. There's a problem.

In general, therefore, defect inspection is performed on the cleaned polycrystalline silicon wafer to check for foreign matter and dirt.
As an inspection method, for example, there is a method of irradiating a polycrystalline silicon wafer with light, capturing an image of the wafer surface irradiated with light, and analyzing the captured image to determine the presence or absence of a defect.
Here, the polycrystalline silicon wafer has a different crystal orientation depending on the portion. Therefore, when the directional light is irradiated onto the polycrystalline silicon wafer, the main reflection direction of the light differs depending on the portion, so that the contrast of the mosaic crystal pattern appears strongly, and the defect such as foreign matter mixed in the wafer and the It becomes difficult to distinguish the crystal pattern from the crystal pattern, and the pattern may be erroneously detected as a defect.

  On the other hand, for example, Patent Document 1 utilizes the fact that the crystal pattern is substantially similar on both surfaces of the polycrystalline silicon wafer, compares the images of the opposing positions on both surfaces, and cancels the contrast of the crystal pattern, There are described a method for inspecting for defects on both sides of a wafer and a method for inspecting for defects by analyzing a plurality of images taken by sequentially irradiating a wafer with light from a plurality of angles.

JP 2007-67102 A

However, in the method of comparing the images at the opposite positions on both sides described in Patent Document 1, if the crystal grain boundaries do not exactly match in the front and back images, the boundary may be erroneously detected as a defect. There is. That is, there is a problem that a minute shift when matching the crystal patterns on the front and back surfaces of the wafer, a minute difference in imaging magnification when imaging the front and back surfaces, and the like become fatal.
Further, in the above-described technique, in the method of sequentially irradiating the wafer with light from a plurality of angles and detecting a region having a small change in brightness as a defect candidate, there is a crystal region having a small change in brightness with respect to a plurality of angles of illumination There is a possibility that the area is erroneously detected as having a defect.
Furthermore, in these techniques, in order to determine the presence / absence of a defect from a plurality of images, there is a problem that imaging is required a plurality of times and the processing speed of the defect inspection is slow.

  An object of the present invention is to solve the above-mentioned problems, and an object of the present invention is to enable inspection of the presence or absence of defects from a single image in which the contrast of the crystal pattern of the polycrystalline silicon wafer is sufficiently reduced. An object of the present invention is to provide an inspection method and apparatus for a crystalline silicon wafer.

The inventors have intensively studied to solve the above problems.
As a result, a part or all of the surface of the polycrystalline silicon wafer is covered with a cover, and the surface of the wafer is indirectly illuminated by irregular reflection on the inner surface of the cover without directly irradiating light from the light source on the surface of the wafer. As a result, the irregularly reflected light is further diffusely reflected on the surface of the polycrystalline silicon wafer, and multiple irregular reflections of the light are generated, so that the wafer surface can be irradiated in various directions and angles.
Thereby, the contrast based on the difference in the main reflection direction of the above-mentioned polycrystalline silicon wafer can be reduced, and new knowledge that defects on the wafer surface can be appropriately detected from a single image obtained by one imaging. Got.

This invention is based on said knowledge, The summary structure is as follows.
(1) A method for inspecting a polycrystalline silicon wafer,
Covering at least part of the surface of the polycrystalline silicon wafer with a cover;
Without directly irradiating light on the wafer surface, the surface of the wafer is indirectly reflected by irregular reflection on the inner surface of the cover,
The illuminated wafer surface is imaged through an imaging unit provided on the cover to obtain an image,
A method for inspecting a polycrystalline silicon wafer, wherein a defect on the wafer surface is detected from the acquired image.

(2) The acquired image is obtained by dividing the luminance value of the corresponding pixel of the acquired image by a predetermined illuminance pattern to correct illuminance unevenness,
The method for inspecting a polycrystalline silicon wafer according to (1), wherein a defect on the wafer surface is detected from the corrected image.

(3) Calculate the average brightness of the acquired image or the corrected image,
According to the average brightness, set at least one of a lower threshold and an upper threshold,
The polycrystalline silicon according to (1) or (2) above, wherein a pixel that is not more than the set lower threshold and upper threshold is detected as a defective portion in the acquired image or the corrected image. Wafer inspection method.

(4) A polycrystalline silicon wafer inspection device,
A cover having an imaging unit that covers at least part of the surface of the polycrystalline silicon wafer and diffusely reflects light from the light source on the inner surface;
Imaging means for imaging the wafer surface indirectly illuminated with the irregularly reflected light from the outside of the cover via an imaging unit provided on the cover;
And a defect detection unit for detecting defects on the wafer surface from an image acquired by the imaging unit.

  (5) The defect portion detecting means includes illuminance unevenness correcting means for correcting the illuminance unevenness by dividing the luminance value of the corresponding pixel of the acquired image by a predetermined illuminance pattern. ) Inspection apparatus for polycrystalline silicon wafers.

(6) The defect detection means, average brightness calculation means for calculating an average brightness of the acquired image or the image after correction,
Threshold setting means for setting at least one of a lower limit threshold and an upper limit threshold according to the average brightness calculated by the average brightness calculation means;
(4) or (5), characterized by comprising threshold processing means for detecting, as a defective portion, pixels that are equal to or lower than the set lower threshold and upper threshold in the acquired image or the corrected image. ) Inspection apparatus for polycrystalline silicon wafers.

  According to the present invention, the wafer surface is covered with a cover, light is diffusely reflected on the inner surface of the cover and the wafer surface, and the direction and angle of light incident on the wafer are made uniform, so that the wafer can be used under non-directional illumination conditions. The surface can be illuminated. Accordingly, the contrast of the crystal pattern of the wafer is reduced, and the S / N ratio with respect to the defect on the wafer surface portion is improved, so that the defect can be detected appropriately.

1 is a schematic perspective view of a polycrystalline silicon wafer inspection apparatus according to an embodiment of the present invention. It is sectional drawing of the inspection apparatus of the polycrystalline silicon wafer concerning one Embodiment of this invention. It is a flowchart of the inspection method of the polycrystalline silicon wafer concerning one Embodiment of this invention. It is the schematic perspective view and sectional drawing of the inspection apparatus of the polycrystalline silicon wafer concerning one Embodiment of this invention. (a) It is a figure which shows the image which image | photographed the wafer surface with the camera of the inspection apparatus of the polycrystalline silicon wafer of this invention. (b) It is a figure which shows the image which image | photographed the wafer surface using the highly directional line illumination. FIG. 6 is a diagram showing luminance on the line A-B in FIG. It is a figure for demonstrating illuminance unevenness. It is a figure which shows the illumination intensity pattern in the apparatus of this invention. It is a figure which shows the image and brightness | luminance which correct | amended illumination intensity irregularity. (a) It is the figure which expanded the part shown by E of FIG. (b) It is a figure which shows the brightness | luminance on the FG line of (a). It is a figure which shows the micro defect of the polycrystalline silicon wafer detected by the apparatus of this invention.

Hereinafter, the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic perspective view showing a polycrystalline silicon wafer inspection apparatus according to an embodiment of the present invention.
As shown in FIG. 1, in order to detect defects on the surface of the polycrystalline silicon wafer 1, the inspection apparatus of the present embodiment includes a cover 2 that irregularly reflects light on the inner surface, and a line sensor camera 3 that images the wafer surface. In addition, a defect portion detection means (image processing computer) 4 for detecting defects on the wafer surface from the captured image is provided. In addition, the inspection apparatus according to the present embodiment further includes a transfer stage 5 for transferring a wafer in the illustrated example.
FIGS. 2 (a) and 2 (b) are cross-sectional views of the inspection apparatus of FIG. 1 as viewed from the directions of arrows (a) and (b), respectively.
As shown in FIGS. 1 and 2 (a) and 2 (b), the cover 2 of the inspection apparatus of the present invention has an imaging unit 21 for imaging the wafer surface. The so-called arcade shape covers part or all of the wafer surface (a part in the illustrated example).
Further, in the illustrated example, the cover 2 has a bottom portion 2a, and a large number of light sources 22 are arranged at equal intervals on the bottom portion 2a.
Note that “covering part or all of the wafer surface” means that the cover part is located above part or all of the wafer surface (on the wafer surface side in the normal direction of the wafer surface).

Hereinafter, a method for inspecting the surface defect of the polycrystalline silicon wafer by the above-described inspection apparatus will be described.
First, light is emitted from the light source 22. Here, in the example shown in FIGS. 2 (a) and 2 (b), a large number of light sources 22 are arranged at equal intervals on the bottom 2a extending from the opening end of the cover 2 inward in the opening width direction. Light (for example, light having directivity) emitted from these light sources 22 indirectly irradiates the surface of the polycrystalline silicon wafer 1. That is, since the light source 22 is inside the cover of the bottom 2a, the light emitted from the light source 22 does not directly irradiate the surface of the wafer 1, but irradiates the inner surface 2b of the cover 2. As described above, the inner surface 2b of the cover 2 of this apparatus diffuses light, and this irregularly reflected light is further diffusely reflected on the inner surface 2b of the cover 2 and the surface of the wafer, and this multiple diffusely reflected light is incident on the wafer surface. To do. Accordingly, it is possible to irradiate the wafer surface with light in various directions and angles.
For this reason, in a single image, the contrast based on the difference in the main reflection direction of the polycrystalline silicon wafer can be reduced, and the S / N ratio of the defect can be improved.
Here, in order to diffusely reflect the light on the inner surface of the cover 2, for example, a glossy paint can be applied to the inner surface of the cover, or the surface of the inner surface of the cover can be roughened. In consideration of the above, white is preferable. Further, by providing symmetry to the shape of the cover and the arrangement of the light sources, it is possible to further diversify the direction and angle of incidence on the wafer surface.For example, in this embodiment, as shown in FIG. The shape of the cover is symmetrical with the imaging unit 21 as the center, and the arrangement of the light sources is also symmetrical.

Next, a defect inspection method according to an embodiment of the present invention will be described in detail with reference to the flowchart of FIG.
When imaging the surface of the wafer irradiated with light in the various directions and angles described above using the line sensor camera 3, for example, as shown in FIG. 1, it is synchronized with the transfer of the wafer 1 by the transfer stage 5. Thus, an image can be acquired by imaging (step S1).
Here, imaging by the line sensor camera is performed through the imaging unit 21. The imaging unit 21 may be a hole in the cover 2 or a window with a glass or the like that transmits light.
Since the imaging unit 21 is a portion that does not perform irregular reflection of light, it is preferable that the imaging unit 21 has as small an area as possible while ensuring a size that allows imaging.
In acquiring an image, the pixel resolution of the line sensor camera 3 and its associated optical lens can be adjusted according to the size of the defect to be detected. For example, a defect with a diameter of 100 μm can be detected. Therefore, the pixel resolution can be set to 50 μm or less on the wafer surface per pixel, preferably 20 μm or less.
The obtained image can be processed as, for example, an image I (X, Y) in the defect detection means 4. Here, I (X, Y) represents the luminance at the pixel coordinates (X, Y), X is a coordinate in the one-dimensional visual field direction of the line sensor camera 3, and Y is a coordinate in the transport direction of the transport stage 5.

Next, the illuminance unevenness can be corrected as necessary (step S2).
Hereinafter, correction of illuminance unevenness will be described.
Depending on the shape of the cover 2 or the like, there may be a difference in brightness (irradiance unevenness) of the irradiated light within the field of view of the line sensor camera, for example, slightly darker at both ends of the field of view than at the center of the field of view. In order to correct such illuminance unevenness, the corrected image I2 is obtained by dividing the luminance value I1 (X, Y) of the pixel at the corresponding coordinates of the image by a predetermined illuminance pattern L (X, Y) set in advance. (X, Y) can be obtained. That is, each coordinate (X, Y) can be calculated as in the following formula 1.

(Formula 1)
I2 (X, Y) = I1 (X, Y) / L (X, Y)

Illuminance pattern L (X, Y), for example, by placing a reference plate (such as a uniform white plate) in advance on the transfer stage 5 instead of the wafer 1, and line sensor camera 3 under the same illumination conditions as during defect inspection The image J (X, Y) of the reference plate can be obtained from the above and each coordinate can be obtained based on the following equation 2. However, Mean [J] is the average value of J (X, Y).
Here, the illuminance pattern L (X, Y), for example, acquires an image J (X) of a reference plate for one line, and a pattern only for the X coordinate, L (X) = J (X) / Mean [ J].

(Formula 2)
L (X, Y) = J (X, Y) / Mean [J]

  Thereby, the illuminance unevenness is corrected, and the contrast of the polycrystalline silicon wafer can be further reduced.

Next, in order to detect (determine) a defect from the obtained image, for example, first, obtain the average luminance Lm for the acquired image I1 (X, Y) or the corrected image I2 (X, Y) (step S3).
Hereinafter, an example of a micro defect detection method will be described with the corrected image I2 (X, Y) as an example.
The average brightness Lm of the corrected image may be obtained, for example, for all pixel coordinates (X, Y), or a representative line in the image (X = constant or Y = constant), An average may be obtained.
Next, at least one of a lower limit threshold and an upper limit threshold is set according to the average luminance Lm.
In the example shown in FIG. 3, only the lower limit threshold is set (step S4).
The lower threshold Lb can be set by the following formula 3, for example.
However, the following coefficient Kb is a real number satisfying 0 ≦ Kb <1.

(Formula 3)
Lb = Kb × Lm

Next, the corrected image I2 (X, Y) and the lower threshold Lb are compared, and the coordinates (Xn, Yn) (n = 1, 2,..., I2 (X, Y) <Lb are satisfied. N) is obtained (step S5). The pixel detected here becomes a defective part.
The value of the coefficient Kb is preferably in the range of 0.6 to 0.8 when detecting a black foreign object or dirt, and may be set according to other target defects to be detected.
Thus, by setting the threshold value according to the average luminance of the image, there is an effect that the defect detection result does not change depending on the overall brightness of the light irradiation.
Further, the upper limit threshold and the lower limit threshold can be made single by correcting the illuminance unevenness. In other words, in the case where illuminance unevenness is not corrected, depending on the degree of illuminance unevenness, it may be necessary to set a plurality of threshold values according to the pixel coordinates, but by correcting the illuminance unevenness, Can be processed, and simple defect detection processing becomes possible.

Subsequently, from the detected pixels (Xn, Yn) (n = 1, 2,..., N), an area where pixels are sequentially connected is found, and an identifier indicating that a defect has been detected in each of the found areas. It can be given (step S6).
In general, since one defect includes a plurality of pixels, pixels included in the same defect are connected by this step, and can be identified in units of defects.
For the distance between pixels, a distance function d (Xa, Ya, Xb, Yb) is defined in advance for two pixels (Xa, Ya), (Xb, Yb), and d (Xa , Ya, Xb, Yb) ≦ d0, the two pixels (Xa, Ya) and (Xb, Yb) can be considered to be included in the same region and can be connected and given the same identifier k. In addition, the definition of the distance function d may be arbitrary, for example, it can be set as the following formulas 4-6.

(Formula 4)
d (Xa, Ya, Xb, Yb) = | Xa−Xb | + | Ya−Yb |

(Formula 5)
d (Xa, Ya, Xb, Yb) = min {| Xa−Xb |, | Ya−Yb |}

(Formula 6)
d (Xa, Ya, Xb, Yb) = [(Xa−Xb) ² + (Ya−Yb) ² ] ^1/2

Here, if the distance function of Equation 5 is adopted and d0 = 1, then 4 neighboring pixels are connected, or if Equation 5 is adopted and d0 = 1, 8 neighboring pixels are connected.
Finally, if a defect is found in the previous step on the monitor 41 (that is, if one or more identifiers are assigned), a defect determination (NG, etc.) is displayed. A determination of none (such as OK) is displayed (step S7).
Further, an image of the entire surface of the wafer may be displayed to indicate the position of the defect found on the image, and an enlarged image of each defective portion may be displayed separately from the image.

In step S4, only the lower limit threshold Lb is set. However, the upper limit threshold La is further set to La = Ka × Lm by the coefficient Ka (> 1), and the pixel region satisfies I2 (X, Y)> La. You may also find white or light colored foreign matter or dirt.
In this case, Ka is preferably in the range of 1.2 to 1.4.

In the above description, the combination of illumination by the line sensor camera and the arcade cover has been described. However, as shown in FIG. 4, another combination of illumination by the area camera and the dome cover is used. An effect can be obtained.
In FIG. 4, the line sensor camera 3 is replaced with an area camera 30 and the arcade-shaped cover 2 is replaced with a dome-shaped cover 20 with respect to the previous embodiment.
Just as in the previous embodiment, omnidirectional illumination by the dome-shaped cover is realized, and the crystal pattern contrast of the wafer 1 can be sufficiently reduced.
In the case of the present embodiment, when the illumination range of the dome-shaped cover 20 or the imageable range of the area camera 30 is narrow relative to the wafer 1, the transfer stage 5 is relative to the dome-shaped cover 20 and the area camera 30. The entire surface of the wafer 1 can be divided while moving the image to perform imaging multiple times.
The dome shape includes all forms that are convex in the center.

It shows that the contrast of the crystal pattern can be reduced by the illumination means of the present invention.
The two images shown in FIG. 5 are an image (a) obtained by photographing a polycrystalline silicon wafer according to the embodiment of the present invention, and an image (b) obtained by photographing the same wafer with a highly directional line illumination as a comparative example. is there.
In photographing the image (a), the illumination of the arcade cover and the line sensor camera shown in FIGS. 1, 2 (a) and 2 (b) were used, and the pixel resolution of the image was 13.5 μm / pixel.
The other image (b) is illuminated with line illumination from a direction inclined by 40 degrees with respect to the normal direction of the wafer surface, and is inclined with 60 degrees on the opposite side of the illumination with respect to the normal direction of the wafer surface. The image was taken with a line sensor camera, and the pixel resolution was 130 μm / pixel.
As shown in FIG. 5, the contrast of the crystal pattern is strong in the image (b) using line illumination, whereas the contrast of the crystal pattern is reduced in the image (a) photographed by the inspection apparatus of the present invention. You can see that

FIG. 6 shows the relative luminance when the average value of the luminance on the wafer surface is set to 1 for one line indicated by the symbols AB in the image shown in FIG.
As shown in FIG. 6, in the case of line illumination (b), the brightness of the wafer surface has a variation of about ± 30% of the average value, whereas in the illumination condition (a) of the present invention (wafer It can be seen that the variation can be reduced to about ± 10% (excluding the effect of the decrease in illuminance at both ends), about 1/3 of the variation.

Next, it will be shown that defects can be detected by the image processing of the present invention.
When the same wafer as in FIG. 5 was imaged by the line sensor under the same conditions as in the image of FIG. 5, an image as shown in FIG. 7 (a) was obtained. The graph of FIG. 7 (b) shows the luminance on the broken line indicated by the symbol CD in the upper image. It can be seen that the luminance at both ends is slightly reduced due to the uneven illuminance.

Next, correction by an illuminance pattern was performed.
FIG. 8 shows an illuminance pattern in the apparatus of the present invention.
As described above, the illuminance pattern is placed on a uniform white plate conveyance stage, and an image of the reference plate for one line is acquired from the line sensor camera under the same illumination conditions as during defect inspection. , L (X) = J (X) / Mean [J].
When the image of FIG. 7 (a) was processed using this illuminance pattern, the image shown in FIG. 9 (a) was obtained.
The graph of FIG. 9 (b) shows the luminance on the broken line indicated by the symbol CD in the upper image.
In Fig. 9 after correction, the illuminance unevenness is improved compared to Fig. 7 before correction, and the brightness of the wafer surface is within ± 10% of the average brightness (100) including both ends of the wafer. I was able to fit it.
On the other hand, a minute black foreign substance as shown in the image of FIG. 10 (a) was present in the portion indicated by symbol E in FIG. The luminance on the broken line indicated by symbol FG of the image is shown in FIG. 10 (b).
As shown in FIGS. 10 (a) and 10 (b), the S / N of the defective portion can be sufficiently increased by sufficiently reducing the contrast of the crystal pattern.
Finally, when the lower limit threshold Lb = 70 (coefficient Kb = 0.7) and the upper limit threshold La = 130 (coefficient Ka = 1.3) are set for the image of FIG. 9, defect detection is performed as shown in FIG. Many small defects (foreign matter and dirt) were found in Fig. 11 (in FIG. 11, the detected pixels are emphasized and enlarged for easy understanding). It should be noted that the defect indicated by symbol E in the figure is also correctly detected.

1 Polycrystalline silicon wafer
2 Arcade cover
20 Dome-shaped cover
21, 210 Imaging unit
22 Light source
2a Bottom
2b, 20b Cover inner surface
3 Line sensor camera
30 area camera
4 Image processing computer
41 Monitor
5 Transfer stage

Claims (6)

A method for inspecting a polycrystalline silicon wafer,
Covering at least part of the surface of the polycrystalline silicon wafer with a cover;
Without directly irradiating light on the wafer surface, the surface of the wafer is indirectly reflected by irregular reflection on the inner surface of the cover,
The illuminated wafer surface is imaged through an imaging unit provided on the cover to obtain an image,
A method for inspecting a polycrystalline silicon wafer, wherein a defect on the wafer surface is detected from the acquired image. The acquired image divides the luminance value of the corresponding pixel of the acquired image by a predetermined illuminance pattern to correct illuminance unevenness,
2. The method for inspecting a polycrystalline silicon wafer according to claim 1, wherein a defect on the wafer surface is detected from the corrected image. Calculating the average brightness of the acquired image or the corrected image;
According to the average brightness, set at least one of a lower threshold and an upper threshold,
3. The inspection of a polycrystalline silicon wafer according to claim 1, wherein pixels that are not more than the set lower threshold and upper threshold are detected as defective portions in the acquired image or the corrected image. Method. An inspection apparatus for polycrystalline silicon wafers,
A cover having an imaging unit that covers at least part of the surface of the polycrystalline silicon wafer and diffusely reflects light from the light source on the inner surface;
Imaging means for imaging the wafer surface indirectly illuminated with the irregularly reflected light from the outside of the cover via an imaging unit provided on the cover;
And a defect detection unit for detecting defects on the wafer surface from an image acquired by the imaging unit.   5. The illuminance unevenness correcting unit according to claim 4, wherein the defect portion detecting unit includes illuminance unevenness correcting means for correcting illuminance unevenness by dividing a luminance value of a corresponding pixel of the acquired image by a predetermined illuminance pattern. Inspection equipment for polycrystalline silicon wafers. The defect detection means includes an average luminance calculation means for calculating an average luminance of the acquired image or the corrected image;
Threshold setting means for setting at least one of a lower limit threshold and an upper limit threshold according to the average brightness calculated by the average brightness calculation means;
6. The threshold processing means for detecting, as a defective portion, pixels that are equal to or lower than the set lower threshold and upper threshold in the acquired image or the corrected image, according to claim 4 or 5. Inspection equipment for polycrystalline silicon wafers.