JP4906602B2 - Defect inspection apparatus and defect inspection method for polycrystalline silicon substrate - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a defect inspection device and defect inspection method capable of removing a false crack image by X-ray scattering on a substrate surface occurring when a sample such as a solar cell substrate having a surface texture structure constituted by a polycrystal silicon substrate is inspected by X-ray. <P>SOLUTION: Data related to, for example, spatial variation of the differential value of concentration between pixels and X-ray absorption rate of each part of the sample is calculated from X-ray transmission data taken while the sample is rotated, and a plane image is reconstituted and operated using the spatial variation data. Thus, the plane image where part having high spatial variation of the X-ray absorption rate is emphasized inside the sample is obtained, and a minute crack or the like on the substrate can be clearly imaged. <P>COPYRIGHT: (C)2009,JPO&amp;INPIT

Description

  The present invention relates to a defect inspection apparatus and a defect inspection method, and more particularly to a defect inspection apparatus and a defect inspection method for a polycrystalline silicon substrate using X-rays.

  In an industrial X-ray inspection apparatus, generally, a sample to be used for measurement is placed between an X-ray source and an X-ray detector that are arranged opposite to each other, and the sample is rotated or an X-ray source is provided. X-ray transmission data obtained by irradiating the sample with X-rays while the pair of X-ray detectors is rotated with respect to the sample, and a rotational axis is obtained by tomographic image reconstruction calculation using the X-ray transmission data Construct a tomographic image of the sample along a plane perpendicular to. By using such an X-ray apparatus, for example, the presence or absence of defects such as cracks existing in the article can be investigated non-destructively (see, for example, Patent Document 1).

JP-A-9-145541

  However, according to the conventional X-ray apparatus that reconstructs a tomographic image of a sample using the X-ray transmission data as described above, a tomographic image of a desired part may not be obtained accurately depending on the sample to be measured. is there. That is, for example, in the case of a polycrystalline silicon substrate such as a solar cell, X-ray scattering due to the texture structure formed on the surface of the substrate may appear as if there are minute cracks. There was a problem that it might be detected.

  In addition, in the case where the object to be measured is an article in which substances made of a plurality of materials such as a substance having a high X-ray absorption rate such as a metal and a substance having a low X-ray absorption rate such as a resin are mixed, In some cases, fine contrast such as minute cracks cannot be imaged well due to the influence of a substance having a high absorption rate. In such a case, there is a problem that detection cannot be performed with high accuracy.

  The present invention has been made to solve such a problem. For example, X-rays are scattered due to a texture structure formed on a substrate surface of a polycrystalline silicon substrate, particularly a solar cell substrate. An object of the present invention is to obtain a defect inspection apparatus and defect inspection method for a polycrystalline silicon substrate capable of appropriately inspecting a minute crack or the like in a sample that seems to have a crack.

  In the present invention, a polycrystalline silicon substrate is disposed between an X-ray source and an X-ray detector, and a relative rotation is applied to the X-ray source, the pair of X-ray detectors, and the polycrystalline silicon substrate. However, the polycrystalline silicon substrate is irradiated with X-rays, and the X-ray transmission data taken at every predetermined rotation angle is used to distinguish between the grain boundaries and cracks of the polycrystalline silicon substrate and inspect for defects. A defect inspection apparatus for a polycrystalline silicon substrate, wherein spatial variation data for calculating spatial variation data relating to spatial variation in X-ray absorption rate of each part of the polycrystalline silicon substrate is calculated based on the X-ray transmission data. Using the variation data calculation means and the spatial variation data, a planar image of the polycrystalline silicon substrate was constructed, and the portion where the spatial variation of the X-ray absorption rate was large inside the polycrystalline silicon substrate was emphasized. Emphasis to generate a planar image Reconstruction calculation means, threshold setting means for setting a threshold value for the crack based on the grain boundary size of the polycrystalline silicon substrate, and the region in the planar image generated by the enhanced image reconstruction calculation means. A defect inspection apparatus for a polycrystalline silicon substrate comprising defect detection means for comparing a size with the threshold value and detecting a portion larger than the threshold value as a defect.

  In the present invention, a polycrystalline silicon substrate is disposed between an X-ray source and an X-ray detector, and a relative rotation is applied to the X-ray source, the pair of X-ray detectors, and the polycrystalline silicon substrate. However, the polycrystalline silicon substrate is irradiated with X-rays, and the X-ray transmission data taken at every predetermined rotation angle is used to distinguish between the grain boundaries and cracks of the polycrystalline silicon substrate and inspect for defects. A defect inspection apparatus for a polycrystalline silicon substrate, wherein spatial variation data for calculating spatial variation data relating to spatial variation in X-ray absorption rate of each part of the polycrystalline silicon substrate is calculated based on the X-ray transmission data. Using the variation data calculation means and the spatial variation data, a planar image of the polycrystalline silicon substrate was constructed, and the portion where the spatial variation of the X-ray absorption rate was large inside the polycrystalline silicon substrate was emphasized. Emphasis to generate a planar image Reconstruction calculation means, threshold setting means for setting a threshold value for the crack based on the grain boundary size of the polycrystalline silicon substrate, and the region in the planar image generated by the enhanced image reconstruction calculation means. The polycrystalline silicon substrate is a defect inspection device for a polycrystalline silicon substrate that includes a defect detection means that detects a portion having a size larger than the threshold value by comparing it with the threshold value. In particular, X-rays are scattered due to the texture structure formed on the substrate surface of the solar cell substrate, and minute cracks and the like on the substrate that appear as if there are minute cracks are appropriately inspected. It is possible.

Embodiment 1 FIG.
1 is a schematic configuration diagram showing the configuration of a defect inspection apparatus for a polycrystalline silicon substrate according to Embodiment 1 of the present invention. In FIG. 1, 1 is an X-ray source, 2 is an X-ray irradiated from the X-ray source 1, 3 is a subject to be examined (sample) to be subjected to measurement, and 4 is a sample on which the subject 3 is placed. It is a stage. Reference numeral 5 denotes an X-ray detector composed of an X-ray light receiving element for detecting the X-ray 2 transmitted through the test pair 3, and reference numeral 6 denotes an image capturing for capturing X-ray transmission data detected by the X-ray detector 5. An apparatus 7 is an arithmetic unit that analyzes the X-ray transmission data captured by the image capturing apparatus 6.

  The X-ray source 1 irradiates the subject 3 with X-rays 2 and transmits the subject 3. The subject 3 is a solar cell substrate made of, for example, a polycrystalline silicon substrate. The solar cell substrate is characterized by having a structure called a texture structure in which numerous irregularities are formed on the surface.

  When a normal semiconductor substrate is irradiated with X-rays 2, a transmission image corresponding to the X-ray dose transmitted through the substrate can be obtained. However, when there are innumerable irregularities called texture structures on the substrate surface, X-rays 2 are scattered at the texture portion, and the amount of transmission increases at the boundary between the collision portions of the scattered X-rays. In some cases, a linear image may be obtained. For this reason, it is difficult to distinguish this linear image from that due to defects.

  However, the linear image seen at the boundary part has a feature that the position detected on the transmission image of the subject 3 differs when the irradiation angle of the X-ray 2 with respect to the subject 3 is changed.

  When the object 3 is a substrate made of polycrystalline silicon, a grain boundary between crystals may be seen in an X-ray transmission image depending on an X-ray incident angle and observation conditions, and this grain boundary part is also distinguished from a defect. It is difficult.

  The subject 3 is placed on the sample stage 4. The sample stage 4 includes a rotation mechanism for rotating the subject 3 around a rotation axis along the z-axis direction (height direction) and three axes (x, y, z) orthogonal to each other including the z-axis. And a mechanism for rotating in the vertical direction of the z axis with respect to the center of the stage. At least one of these mechanisms is used to provide relative rotation to the pair of X-ray source 1 and X-ray detector 5 and the test pair 3.

  Therefore, the measurement is performed by detecting the X-ray 2 transmitted by the X-ray detector 5 at every predetermined rotation angle while rotating the subject 3 on the sample stage 4 in the vertical direction of the z-axis, for example. The obtained output of the X-ray detector 5 at every predetermined rotation angle, that is, X-ray transmission data, is sequentially taken into the arithmetic device 7 via the image taking device 6.

  The arithmetic device 7 is, for example, a system that includes a computer, its peripheral devices, and an analysis program. The arithmetic unit 7 irradiates the X-ray 2 while giving a relative rotation to the pair of the X-ray source 1 and the X-ray detector 5 and the pair 3 to be examined, and transmits the X-ray transmitted at every predetermined rotation angle. Using the data, for example, the space of the X-ray absorption rate of each part of the subject 3 such as a difference value of X-ray intensity data (or pixel density data expressed in 256 gradations) between adjacent pixels. Variation data calculating means for calculating the spatial variation data related to the spatial variation, and a plane image of the subject 3 is constructed using the spatial variation data, and the space of the X-ray absorption rate inside the subject 3 Enhanced image reconstruction calculating means for generating a planar image in which a portion having a large variation is emphasized, threshold setting means for setting a threshold for a crack to be detected based on the grain boundary size of the polycrystalline silicon substrate, enhanced image reconstruction In the plane image generated by the configuration calculation means Compared to set the size of the site threshold, if there is greater site than the threshold value, and a defect detection means for detecting it as defective.

  Here, the incident angle of the X-ray 2 to the subject 3 is changed by rotating the sample stage 4, but of course the sample stage 4 is fixed and the X-ray source 1 and the X-ray detector 5 are fixed. The angle of incidence of X-rays on the subject 3 may be changed by rotating.

  Data of the X-ray detector 5 captured via the image capturing device 6 is sequentially analyzed by the arithmetic device 7, for example, a differential value arithmetic process is performed, and the X-ray absorption rate space in each part of the subject 3 is analyzed. Since the fluctuation data is obtained and the plane image of the subject 3 is constructed using the spatial fluctuation data, the linear image data and the crystal grain boundary image at the boundary portion are offset on the data, and the object is thus corrected. A planar image is obtained in which a portion where the spatial change in the X-ray absorption rate is large is emphasized inside the specimen 3, and only an image based on a defect image can be obtained. At this time, a threshold value is set based on the size of the grain boundary of the polycrystalline silicon substrate constituting the object 3, and a defect having a size larger than the threshold value is determined as a crack. Thereby, it is possible to distinguish between the grain boundary of the polycrystalline silicon substrate and the crack having the same order size as the grain boundary, and to detect only the true crack as a defect. As described above, in the present embodiment, the X-ray transmission data acquired at each predetermined rotation angle is used to synthesize the X-ray image of the subject 3 from the plurality of X-ray transmission data acquired at each rotation angle. In this case, data related to the spatial variation of the X-ray absorption rate of each part of the test pair 3 is calculated, and based on the output of the X-ray detector 5 using the spatial variation data of the X-ray absorption rate. Since the X-ray intensity distribution images for each rotation angle are integrated and the planar images are synthesized, cracks can be detected appropriately and accurately even on a substrate with X-ray scattering.

  In this embodiment, only one set of X-ray source 1 and X-ray detector 5 is used. However, two or more sets of X-ray source 1 and X-ray detector 5 are used. Image reconstruction that integrates the X-ray intensity distribution images when reconstructing an image along the plane of the central axis of rotation using the X-ray transmission data acquired from the X-ray detector 5 described above. By using the calculation means, the sample stage 4 and the X-ray source 1 may not be moved.

  As described above, the defect inspection apparatus according to the present embodiment arranges the subject 3 as the sample between the X-ray source 1 and the X-ray detector 5 and rotates the sample stage 4 to rotate the sample stage 4. The X-ray transmission data acquired at every predetermined rotation angle is used by irradiating the subject 3 with the X-ray 2 while giving a relative rotation to the pair of the radiation source 1 and the X-ray detector 5 and the subject 3. Thus, for example, spatial variation data of the X-ray absorption rate in each part of the subject 3 such as a difference value of X-ray intensity data (density data) between adjacent pixels is obtained, and the subject is used using the spatial variation data. 3 is constructed, a planar image is obtained in which a region where the spatial change in the X-ray absorption rate is large is emphasized inside the subject 3. Further, by using the difference value of the X-ray intensity data, it is possible to obtain a planar image excluding a specific scattered image that appears when the solar cell substrate is observed by X-ray. As described above, according to the present embodiment, X-rays are scattered due to the texture structure formed on the substrate surface of the polycrystalline silicon substrate, particularly the solar cell substrate, as if minute cracks exist. It is possible to properly inspect minute cracks and the like on the visible substrate.

Embodiment 2. FIG.
2 is a schematic configuration diagram showing the configuration of a defect inspection apparatus for a polycrystalline silicon substrate according to Embodiment 2 of the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the description thereof is omitted here. In FIG. 2, reference numeral 10 denotes a fixed rotating jig for fixing the X-ray source 1 and the X-ray detector 5 and rotating them simultaneously. Reference numeral 11 denotes a rotation drive motor for driving the fixed rotation jig 10 to rotate.

  The X-ray source 1 emits X-rays 2 and transmits the subject 3. The subject 3 is a solar cell substrate made of, for example, a polycrystalline silicon substrate. The surface of the solar cell substrate is characterized by having a structure called a texture structure in which numerous irregularities are formed on the surface. The X-ray source 1 and the X-ray detector 5 are installed along the vertical axis, and are rotated by the fixed rotating jig 10 while being positioned diagonally along the vertical axis. A visible light source and a visible light image detector (not shown) are arranged beside the X-ray source 1. The visible light image detector is composed of a CCD camera or the like. Since the visible light source and the visible light image detector are fixed to the fixed rotation jig 10, they rotate together with the X-ray source 1 and the X-ray detector 5. Note that the X-ray source 1, the visible light source, and the visible light image detector may be configured integrally.

  The image capturing device 6 according to the present embodiment detects image information due to scattering of X-rays on the substrate surface from a means for reading an image by visible light and an image by X-rays and an image reading means by X-rays. And a correcting means for correcting cracks on the substrate surface based on information on cracks based on image information obtained from the reading means for visible light and information on linear images based on X-rays scattered on textured portions of the substrate surface. is doing. The determination of the crack and the linear image by the image capturing device 6 is made by a threshold value, and the threshold value has means for changing according to the resolution of the reading means and the number of integrations. The threshold is composed of two thresholds a and b, and at least one of the thresholds changes discontinuously with respect to the reading resolution, or at least one of the thresholds is expressed by a function having the reading resolution as an argument. To do. Further, the other threshold value changes discontinuously with respect to the integration function regardless of the resolution, or is expressed by a function having the number of integrations as an argument.

  Next, a specific flow of detection will be described based on the flowcharts of FIGS. 2 and 3 to 5. First, as shown in FIG. 3, in the image capturing step 1 of FIG. 2A, a visible light image on the surface of the subject 3 is captured (step S1). The visible light source is set on the subject 3 and a visible light image reflected from the surface of the subject 3 is photographed by a CCD camera attached at substantially the same position as the visible light source. Further, the substrate position information, resolution, and magnification at the time of photographing are captured together with the visible light image (step S2). Next, the X-ray 2 is irradiated from the X-ray source 1, the X-ray 2 transmitted through the subject 3 is detected by the X-ray detector 5, and an X-ray transmission image is captured (step S3).

  By the way, a metal electrode wiring mainly composed of silver is formed on the surface of the solar battery cell of the subject 3 made of a polycrystalline silicon substrate as an example. In general, the metal electrode wiring does not transmit X-rays or transmits only a small amount compared to the polycrystalline silicon portion. For this reason, in the X-ray transmission image, the X-ray transmission amount of the polycrystalline silicon portion is large, and in the metal electrode wiring, the X-ray transmission amount is zero or very small. When expressing an X-ray transmission image with 256 gradations according to the amount of X-ray transmission, if a portion where 100% of X-rays are transmitted is 255 and expressed in white, 0%, that is, X The portion where the line has not been transmitted at all is set to 0 and expressed in black. Of course, the number of gradations may be set to 256 or more, or it may not be performed in black and white binary representation. Here, assuming that the amount of X-ray transmission of the polycrystalline silicon substrate is 80% and expressed in this way, the polycrystalline silicon portion is expressed in gray of about 200 gradations on the X-ray transmission image, The metal electrode wiring portion is expressed in black. On the other hand, if there are cracks on the surface and inside of the subject, X-rays are scattered or absorbed by the cracks, so that the amount of X-ray transmission increases or decreases. For this reason, a crack part will be expressed white or black. In addition, when there are innumerable irregularities called texture structures on the substrate surface, X-rays are scattered at the textured portion, so that the amount of transmission increases or decreases at the boundary between the scattered X-ray collisions. A linear white or black image (hereinafter referred to as a linear image) may be obtained at the boundary portion.

  Therefore, based on the position information of the metal electrode wiring part of the visible light image obtained in steps S1 and S2, from the X-ray transmission image obtained in step S3, image processing of only this metal electrode wiring part is performed, Wiring position information is removed from the X-ray transmission image (step S4). Note that, as the removal processing, here, the image of the metal electrode wiring portion is converted into gray having the same gradation 200 as that of the polycrystalline silicon portion. Further, when acquiring the X-ray transmission image information in step S3, the substrate position information, the resolution, and the magnification are captured in the same manner as when the visible light image is acquired. Naturally, when image processing is performed only on the metal electrode wiring portion, information relating to various image captures, such as substrate position information, image resolution, and magnification, is consistent between the visible light image and the X-ray transmission image. Or it needs to be adjusted. In addition, a metal electrode mainly composed of aluminum is formed on the back surface of the solar cell as an example, but unlike the front surface, it is uniformly formed on the entire back surface, so that the amount of X-ray transmission is uniformly reduced. Just do it.

  Moreover, in order to prevent the electrical short circuit of a photovoltaic cell, the part which does not form the metal electrode which has the said aluminum as a main component may exist in a back surface side edge part. Since the X-ray transmission state is different in the portion where the back electrode is not formed as described above, a line image similar to a crack is obtained at this boundary portion. Such a line image can be removed by applying steps S1 to S4 to the back side.

  Furthermore, when assembling a solar cell module using solar cells, an electrode pad may be formed on a part of the back surface of the solar cells in order to connect module electrodes for connecting the solar cells. As an example of this electrode pad, an electrode mainly composed of silver is formed. In this way, when an electrode made of a different material is formed as the back electrode, the X-ray transmission state is different, so that a line image similar to a crack is obtained at the boundary portion of the different material. This line image can also be removed by applying steps S1 to S4 to the back side.

  Next, whether or not there is a crack using information obtained by excluding information on the surface metal electrode wiring portion from the X-ray transmission image, and information such as resolution and magnification of the image when the X-ray transmission image is captured For the location assumed to be a crack portion, the coordinate position, area, length, etc. in the substrate surface are sequentially specified and written to the crack detection location specification table (1), and the crack detection location is determined. The specific table (1) is completed, and finally, the quantity thereof is calculated (step S5). This determination process is performed according to the flow procedure shown in FIG. This will be described later.

  As a result of the determination in step S5, if a crack exists, the process proceeds to step S7. On the other hand, if there is no crack, the process proceeds to step S13, the substrate position is moved or the substrate is replaced, and then the process returns to step S1.

  The case where it progresses to step S7 is demonstrated. First, as in the image capturing step 2 in FIG. 2B, the X-ray source 1 and the X-ray detector 5 are rotated by a predetermined rotation angle in the direction of arrow A in the figure, and steps S1 to S6 are performed again. Similar processing is performed (steps S7 to S11). Here, the processing is performed with the rotation angle set to about 5 degrees, the crack detection position specifying table (2) is completed, and finally the quantity thereof is calculated. The crack detection position specification table (1) and the crack detection position specification table (2) thus obtained are output (step S12).

  Subsequently, the process proceeds to the process of FIG. 5 to identify the crack and the linear image. First, the crack information written in the crack detection position specifying table (1) is read (step S41). Next, the crack information written in the crack detection position specifying table (2) is read (step S42). Next, the crack information read from the crack detection position specification table (1) and the crack detection position specification table (2) is compared (step S43).

  A linear image of scattered X-rays generated due to the texture structure on the substrate surface has a feature that the appearance position thereof changes when the X-ray incident angle is changed. This is a unique phenomenon that occurs only in a polycrystalline silicon substrate having a texture structure formed on the substrate surface. Therefore, in the crack detection position specification tables (1) and (2), if a crack of the same size is detected on the same coordinate, it can be determined as a true crack, and the crack detection position specification table If the position coordinates of the cracks written in (1) and (2) are different, it can be determined that it is a fake crack, that is, a linear image.

  Therefore, based on the comparison result in step S43, it is determined whether or not the position coordinates of the cracks written in the crack detection position specification tables (1) and (2) are the same (step S44). In this case, it is determined as a crack (step S45), and when they are not the same, it is determined as a linear image (step S48), and the process proceeds to step S49.

  If it is determined as a crack in step S45, then the position coordinates and size of the crack are specified (step S46), and the crack is written in the fixed crack detection position specification table (step S47), and the process proceeds to step S49.

  It is determined whether all cracks written in the crack detection position specifying table (1) have been read (step S49). If all have been completed, the process is terminated as it is (step S50), and if not completed, the process returns to step S41 to repeat the process for all the crack information (step S51).

  In the above processing, in order to further improve the detection accuracy, it is only necessary to capture images when the X-ray incident angle is further changed, create more crack detection position specification tables, and use them for determination. It is also possible to improve the detection accuracy by comprehensively determining image information captured at a plurality of image resolutions.

  As described above, it is possible to specify a solar cell substrate having a crack determined to be a true crack.

  Here, the crack determination process shown in FIG. 4 (steps S5 and S9 in FIG. 3) will be described. FIG. 4A is the process of step S5 in FIG. 3, and FIG. 4B is the process of step S9 in FIG. First, FIG. 4A will be described. First, information such as the resolution and magnification of the image when the X-ray transmission image is taken in is read (step S21), and the information and the X-ray transmission image obtained by the processing of steps S1 to S4 in FIG. Based on the information on the metal electrode wiring portion and the image excluding the information on the influence portion that discharges the line image by the metal electrode on the back surface, the size information of the crack to be determined is read from the crack determination size threshold table (step S22). A detection determination size is determined (step S23). The threshold value information of the crack size to be determined is set so that the grain boundary of the polycrystalline silicon substrate and the crack having the same order size as the grain boundary can be distinguished. That is, the threshold value is set to the same value as the grain boundary size of the polycrystalline silicon substrate, and if there is a defect larger than the threshold value, it is determined as a crack. Further, the threshold information of the crack size is enhanced by feeding back information such as the size of the influence of the substrate crack or chipping in the manufacturing process after the inspection. Alternatively, a crack size that causes significant deterioration of the product due to aging can be calculated and fed back from a quality inspection process such as an accelerated test of the product. When the crack size of the substrate having cracks that are to be excluded as defective products from the quality inspection process of the solar battery cell is determined, the crack size to be judged is determined in consideration of the resolution, magnification, etc. of the image in the crack inspection process, and cracks are determined. It is previously written in the judgment size threshold table.

  Then, the crack judgment size threshold value information changes sharply than the reference gradation of the polycrystalline silicon portion, and any one of the length, width and area of the crack information is larger than the threshold information. (Step S24), if such a thing is detected (Step S25), the coordinate position, area, length, etc. in the substrate surface are specified (Step S26), and the crack detection position Write to the specific table (1) (step S27). Furthermore, crack determination is repeated, information such as the coordinate position in the substrate surface, area, and length is written to the detection position specifying table (1), and finally the quantity is calculated. In this way, crack determination is performed, and the detection position specifying table (1) is completed.

  Further, depending on the size and quantity of the determined crack, in the case where the solar cell itself is excluded as a defective product, the writing to the detection position specifying table (1) is completed in the middle without completing the measurement time. You may make it shorten.

  In FIG. 4B, the same process as in FIG. 4A is performed to complete the crack detection position specifying table (2). Since the process of FIG. 4B is basically the same as that of FIG. 4A, detailed description thereof is omitted here.

  Since other operations are the same as those in the first embodiment, description thereof is omitted here.

  As described above, in the defect inspection apparatus according to the present embodiment, the subject 3 as the sample is disposed between the X-ray source 1 and the X-ray detector 5, and the X-ray source is provided by the fixed rotation jig 10. 1 and the X-ray detector 5 are rotated to irradiate the subject 3 with the X-ray 2 while giving a relative rotation to the subject 3 and the pair of the X-ray source 1 and the X-ray detector 5. Using the X-ray transmission data captured at every predetermined rotation angle, for example, the X-ray absorption rate of each part of the subject 3 such as the difference value of the X-ray intensity data (density data) between adjacent pixels. Since spatial variation data is obtained and a planar image of the subject 3 is constructed using the spatial variation data, a planar image in which a region where the spatial change of the X-ray absorption rate is large is emphasized inside the subject 3 is obtained. Further, by using the difference value of the X-ray intensity data, it is possible to obtain a planar image excluding a specific scattered image that appears when the solar cell substrate is observed by X-ray. As described above, according to the present embodiment, X-rays are scattered due to the texture structure formed on the substrate surface of the polycrystalline silicon substrate, particularly the solar cell substrate, as if minute cracks exist. It is possible to properly inspect minute cracks and the like on the visible substrate.

Embodiment 3 FIG.
FIG. 6 is a schematic configuration diagram showing a case where two sets of X-ray sources and X-ray detectors are used in the configuration of the defect inspection apparatus for a polycrystalline silicon substrate according to Embodiment 3 of the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the description thereof is omitted here. In FIG. 6, 1-1 is the first set of X-ray sources (1), 1-2 is the second set of X-ray sources (2), and 2-1 is the first set of X-ray sources 1. X-rays irradiated from -1 2-2 X-rays irradiated from a second set of X-ray sources 1-2, 5-1 a first set of X-ray detectors (1), 5-2 A second set of X-ray detectors (2). In the present embodiment, the X-ray sources 1-1 and 1-2 and the X-ray detector 5-1 so that the X-rays emitted from the X-ray sources 1-1 and 1-2 intersect each other. And 5-2 are arranged, the subject 3 is arranged at a position where the X-rays 2-1 and 2-2 intersect, and taken in from the respective X-ray detectors 5-1 and 5-2. The X-ray transmission data is used to integrate the X-ray intensity distribution images when reconstructing an image along the surface of the subject 3 at a position where the X-rays 2-1 and 2-2 intersect.

  First, the X-ray sources 1-1 and 1-2 emit X-rays 2-1 and 2-2, respectively, and transmit the subject 3. The subject 3 is a solar cell substrate made of, for example, a polycrystalline silicon substrate. The surface of the solar cell substrate is characterized by having a structure called a texture structure in which numerous irregularities are formed on the surface. The X-ray source 1-1 and the X-ray detector 5-1 are installed along the vertical line, and the X-ray source 1-2 and the X-ray detector 5-2 are also installed along the vertical line. .

  The image capturing device 6 according to the present embodiment includes a means for reading an image by visible light and an image by X-ray, and an image by scattering of X-rays on the substrate surface from the image obtained by the image reading means by X-ray. Corrects cracks on the substrate surface based on information on the image information obtained by the means for detecting information and the visible light reading means and information on the linear image by X-rays scattered at the textured portion of the substrate surface And a correcting means. The determination of the crack and the linear image by the image capturing device 6 is made based on a threshold value, and the threshold value has means for controlling the threshold value so as to change according to the resolution of the reading means and the number of integrations. The threshold is composed of two thresholds a and b, and at least one of the thresholds changes discontinuously with respect to the reading resolution, or at least one of the thresholds is expressed by a function having the reading resolution as an argument. To. Further, the other threshold value changes discontinuously with respect to the integration function regardless of the resolution, or is expressed by a function having the number of integrations as an argument.

  A visible light source and a visible light image detector (not shown) are arranged beside the X-ray source 1-1. The visible light image detector is composed of a CCD camera or the like.

  Next, a specific detection flow will be described based on the flowcharts of FIGS. 6 and 3 to 5. First, as shown in FIG. 3, a visible light image on the surface of the subject 3 is picked up by a visible light image detector installed beside the X-ray source 1-1 in FIG. 6 (step S1). The visible light source is installed on the top of the subject 3 and a visible light image reflected from the surface of the subject 3 is photographed by a CCD camera attached at substantially the same position as the visible light source. Further, the substrate position information, resolution, and magnification at the time of photographing are captured together with the visible light image (step S2). Next, the X-ray 2-1 is irradiated from the X-ray source 1-1, the X-ray 2-1 transmitted through the subject 3 is detected by the X-ray detector 5-1, and an X-ray transmission image is captured ( Step S3).

  By the way, a metal electrode wiring mainly composed of silver is formed on the surface of the solar battery cell of the subject 3 made of a polycrystalline silicon substrate as an example. In general, the metal electrode wiring does not transmit X-rays or transmits only a small amount compared to the polycrystalline silicon portion. For this reason, in the X-ray transmission image, the X-ray transmission amount of the polycrystalline silicon portion is large, and in the metal electrode wiring, the X-ray transmission amount is zero or very small. In the case where an X-ray transmission image is expressed in 256 gradations according to the amount of X-ray transmission, if a portion through which 100% of X-rays are transmitted is 255 and expressed in white, 0%, that is, The portion where no X-rays have been transmitted is represented by 0 and expressed in black. Of course, the number of gradations may be set to 256 or more, and it is not always necessary to use black and white binary representation. Here, assuming that the amount of X-ray transmission of the polycrystalline silicon substrate is 80% and expressed in this way, the polycrystalline silicon portion is expressed in gray of about 200 gradations on the X-ray transmission image, The metal electrode wiring portion is expressed in black. On the other hand, if there are cracks on the surface and inside of the subject, X-rays are scattered or absorbed by the cracks, so that the amount of X-ray transmission increases or decreases. For this reason, a crack part will be expressed white or black. In addition, when there are innumerable irregularities called texture structures on the substrate surface, X-rays are scattered at the textured portion, so that the amount of transmission increases or decreases at the boundary between the scattered X-ray collisions. A linear white or black image (hereinafter referred to as a linear image) may be obtained at the boundary portion.

  Therefore, based on the position information of the metal electrode wiring part of the visible light image obtained in steps S1 and S2, from the X-ray transmission image obtained in step S3, image processing of only this metal electrode wiring part is performed, Wiring position information is removed from the X-ray transmission image (step S4). Note that, as the removal processing, here, the image of the metal electrode wiring portion is converted into gray having the same gradation 200 as that of the polycrystalline silicon portion. Further, when acquiring the X-ray transmission image information in step S3, the substrate position information, the resolution, and the magnification are captured in the same manner as when the visible light image is acquired. Naturally, when image processing is performed only on the metal electrode wiring portion, information relating to various image captures, such as substrate position information, image resolution, and magnification, is consistent between the visible light image and the X-ray transmission image. Or it needs to be adjusted. In addition, a metal electrode mainly composed of aluminum is formed on the back surface of the solar cell as an example, but unlike the front surface, it is uniformly formed on the entire back surface, so that the amount of X-ray transmission is uniformly reduced. Just do it.

  Moreover, in order to prevent the electrical short circuit of a photovoltaic cell, the part which does not form the metal electrode which has the said aluminum as a main component may exist in a back surface side edge part. Since the X-ray transmission state is different in the portion where the back electrode is not formed as described above, a line image similar to a crack is obtained at this boundary portion. Such a line image can be removed by applying steps S1 to S4 to the back side.

  Furthermore, when assembling a solar cell module using solar cells, an electrode pad may be formed on a part of the back surface of the solar cells in order to connect module electrodes for connecting the solar cells. As an example of this electrode pad, an electrode mainly composed of silver is formed. In this way, when an electrode made of a different material is formed as the back electrode, the X-ray transmission state is different, so that a line image similar to a crack is obtained at the boundary portion of the different material. This line image can also be removed by applying steps S1 to S4 to the back side.

  Next, whether or not there is a crack using information obtained by excluding information on the surface metal electrode wiring portion from the X-ray transmission image, and information such as resolution and magnification of the image when the X-ray transmission image is captured For the location assumed to be a crack portion, the coordinate position, area, length, etc. in the substrate surface are sequentially specified and written to the crack detection location specification table (1), and the crack detection location is determined. The specific table (1) is completed, and finally, the quantity thereof is calculated (step S5). This determination process is performed according to the flow procedure shown in FIG. This will be described later.

  As a result of the determination in step S5, if a crack exists, the process proceeds to step S7. On the other hand, if there is no crack, the process proceeds to step S13, the substrate position is moved or the substrate is replaced, and then the process returns to step S1.

  The case where it progresses to step S7 is demonstrated. First, the X-ray 2-2 is irradiated from the X-ray source 1-2 in FIG. 6, the X-ray 2-2 transmitted through the subject 3 is detected by the X-ray detector 5-2, and an X-ray transmission image is taken. (Step S7). Thereafter, the same processing as in steps S4 to S6 is performed (steps S8 to S11), the crack detection position specifying table (2) is completed, and finally the quantities thereof are calculated. The crack detection position specification table (1) and the crack detection position specification table (2) thus obtained are output (step S12).

  Here, the X-ray source 1-1 and the X-ray source 1-2 are arranged so that the intersection angle between the X-ray 2-1 and the X-ray 2-2 is about 30 degrees. It is preferable to select an optimum angle that is not easily affected by the crystal grain boundary due to the texture size formed on the substrate surface. Of course, an apparatus configuration having an angle adjustment function capable of arbitrarily setting this angle may be used.

  Subsequently, the process proceeds to the process of FIG. 5 to identify the crack and the linear image. First, the crack information written in the crack detection position specifying table (1) is read (step S41). Next, the crack information written in the crack detection position specifying table (2) is read (step S42). Next, the crack information read from the crack detection position specification table (1) and the crack detection position specification table (2) is compared (step S43).

  A linear image of scattered X-rays generated due to the texture structure on the substrate surface has a feature that the appearance position thereof changes when the X-ray incident angle is changed. This is a unique phenomenon that occurs only in a polycrystalline silicon substrate having a texture structure formed on the substrate surface. Therefore, in the crack detection position specification tables (1) and (2), if a crack of the same size is detected on the same coordinate, it can be determined as a true crack, and the crack detection position specification table If the position coordinates of the cracks written in (1) and (2) are different, it can be determined that it is a fake crack, that is, a linear image.

  Therefore, based on the comparison result in step S43, it is determined whether or not the position coordinates of the cracks written in the crack detection position specification tables (1) and (2) are the same (step S44). In this case, it is determined as a crack (step S45), and when they are not the same, it is determined as a linear image (step S48), and the process proceeds to step S49.

  If it is determined as a crack in step S45, then the position coordinates and size of the crack are specified (step S46), and the crack is written in the fixed crack detection position specification table (step S47), and the process proceeds to step S49.

  It is determined whether all cracks written in the crack detection position specifying table (1) have been read (step S49). If all have been completed, the process is terminated as it is (step S50), and if not completed, the process returns to step S41 to repeat the process for all the crack information (step S51).

  In the above processing, in order to further improve the detection accuracy, it is only necessary to capture images when the X-ray incident angle is further changed, create more crack detection position specification tables, and use them for determination. It is also possible to improve the detection accuracy by comprehensively determining image information captured at a plurality of image resolutions.

  As described above, it is possible to specify a solar cell substrate having a crack determined to be a true crack.

  Here, the crack determination process shown in FIG. 4 (steps S5 and S9 in FIG. 3) will be described. FIG. 4A is the process of step S5 in FIG. 3, and FIG. 4B is the process of step S9 in FIG. First, FIG. 4A will be described. First, information such as the resolution and magnification of the image when the X-ray transmission image is taken in is read (step S21), and the information and the X-ray transmission image obtained by the processing of steps S1 to S4 in FIG. Based on the information on the metal electrode wiring portion and the image excluding the information on the influence portion that discharges the line image by the metal electrode on the back surface, the size information of the crack to be determined is read from the crack determination size threshold table (step S22). A detection determination size is determined (step S23). The threshold value information of the crack size to be determined is set so that the grain boundary of the polycrystalline silicon substrate and the crack having the same order size as the grain boundary can be distinguished. That is, the threshold value is set to the same value as the grain boundary size of the polycrystalline silicon substrate, and if there is a defect larger than the threshold value, it is determined as a crack. Further, the threshold information of the crack size is enhanced by feeding back information such as the size of the influence of the substrate crack or chipping in the manufacturing process after the inspection. Alternatively, a crack size that causes significant deterioration of the product due to aging can be calculated and fed back from a quality inspection process such as an accelerated test of the product. When the crack size of the substrate having cracks that are to be excluded as defective products from the quality inspection process of the solar battery cell is determined, the crack size to be judged is determined in consideration of the resolution, magnification, etc. of the image in the crack inspection process, and cracks are determined. It is previously written in the judgment size threshold table.

  Then, the crack judgment size threshold value information changes sharply than the reference gradation of the polycrystalline silicon portion, and any one of the length, width and area of the crack information is larger than the threshold information. (Step S24), if such a thing is detected (Step S25), the coordinate position, area, length, etc. in the substrate surface are specified (Step S26), and the crack detection position Write to the specific table (1) (step S27). Furthermore, crack determination is repeated, information such as the coordinate position in the substrate surface, area, and length is written to the detection position specifying table (1), and finally the quantity is calculated. In this way, crack determination is performed, and the detection position specifying table (1) is completed.

  Further, depending on the size and quantity of the determined crack, in the case where the solar cell itself is excluded as a defective product, the writing to the detection position specifying table (1) is completed in the middle without completing the measurement time. You may make it shorten.

  In FIG. 4B, the same process as in FIG. 4A is performed to complete the crack detection position specifying table (2). Since the process of FIG. 4B is basically the same as that of FIG. 4A, detailed description thereof is omitted here.

  Since other operations are the same as those in the first embodiment, description thereof is omitted here.

  As described above, the defect inspection apparatus according to this embodiment includes the X-ray source 1-1 and the X-ray detector 5-1, and the X-ray source 1-2 and the X-ray detector 5-2. The X-ray transmission data acquired for each combination of the X-ray source and the X-ray detector is arranged using the X-ray transmission data acquired for each combination of the X-ray source and the X-ray detector. Since the spatial variation data of the X-ray absorption rate in each part of the subject 3 such as the difference value of the data) is obtained and the planar image of the subject 3 is constructed using the spatial variation data, A planar image is obtained in which a region where the spatial change of the linear absorptance is large is emphasized. Further, by using the difference value of the X-ray intensity data, it is possible to obtain a planar image excluding a specific scattered image that appears when the solar cell substrate is observed by X-ray. As described above, according to the present embodiment, X-rays are scattered due to the texture structure formed on the substrate surface of the polycrystalline silicon substrate, particularly the solar cell substrate, as if minute cracks exist. It is possible to properly inspect minute cracks and the like on the visible substrate.

  In the third embodiment, an example in which two sets of X-ray sources and X-ray detectors are provided has been described. However, the present invention is not limited to this, and more may be provided. In this case, it goes without saying that the same effect can be obtained.

It is the block diagram which showed the structure of the defect inspection apparatus which concerns on Embodiment 1 of this invention. It is the block diagram which showed the structure of the defect inspection apparatus which concerns on Embodiment 2 of this invention. It is the flowchart which showed the flow of the process of the defect inspection method which concerns on Embodiment 2 of this invention. It is the flowchart which showed the flow of the process of the defect inspection method which concerns on Embodiment 2 of this invention. It is the flowchart which showed the flow of the process of the defect inspection method which concerns on Embodiment 2 of this invention. It is the block diagram which showed the structure of the defect inspection apparatus which concerns on Embodiment 3 of this invention.

Explanation of symbols

  1,1-1,1-2 X-ray source, 2,2-1,2-2 X-ray, 3 subject, 4 sample stage, 5,5-1, 5-2 X-ray detector, 6 image acquisition Insertion device, 7 arithmetic device, 10 fixed rotation jig, 11 rotation drive motor.

Claims (5)

A polycrystalline silicon substrate is disposed between the X-ray source and the X-ray detector, and the polycrystalline silicon substrate is provided with relative rotation between the X-ray source and the pair of X-ray detectors and the polycrystalline silicon substrate. Polycrystalline silicon that inspects the presence or absence of defects by irradiating a crystalline silicon substrate with X-rays and using the X-ray transmission data taken at every predetermined rotation angle to distinguish the grain boundaries and cracks of the polycrystalline silicon substrate. A substrate defect inspection apparatus,
Spatial variation data calculation means for calculating spatial variation data related to the spatial variation of the X-ray absorption rate of each part of the polycrystalline silicon substrate based on the X-ray transmission data;
Emphasizing the use of the spatial variation data to construct a planar image of the polycrystalline silicon substrate, and generating a planar image in which the portion where the spatial variation of the X-ray absorption rate is large is enhanced in the polycrystalline silicon substrate Image reconstruction calculation means;
Threshold setting means for setting a threshold for the crack based on the size of the grain boundary of the polycrystalline silicon substrate,
Comparing the size of the part in the planar image generated by the enhanced image reconstruction calculation means with the threshold value, if there is a part larger than the threshold value, a defect detection means for detecting it as a defect. A defect inspection apparatus for a polycrystalline silicon substrate, comprising:   The spatial variation data calculating means uses the spatial variation data of the X-ray absorption rate of each part of the polycrystalline silicon substrate as an X-ray intensity distribution image for each rotation angle based on the output of the X-ray detector. 2. The defect inspection apparatus for a polycrystalline silicon substrate according to claim 1, wherein a difference value of density data between adjacent pixels is calculated.   The enhanced image reconstruction calculation means rotates based on the output of the X-ray detector when constructing the planar image using the spatial variation data of the X-ray absorption rate of each part of the polycrystalline silicon substrate. The defect inspection apparatus for a polycrystalline silicon substrate according to claim 1 or 2, wherein X-ray intensity distribution images for each angle are integrated. There are two or more sets of the X-ray source and the X-ray detector,
Each set of the X-ray source and the X-ray detector is arranged so that the X-rays irradiated from the respective X-ray sources intersect, and the polycrystalline silicon substrate is located at the position where the X-rays intersect. Are used to reconstruct an image along the plane of the polycrystalline silicon substrate at the position where each X-ray intersects using the X-ray transmission data acquired from each X-ray detector. The defect inspection apparatus for a polycrystalline silicon substrate according to any one of claims 1 to 3, wherein the X-ray intensity distribution images are integrated. A polycrystalline silicon substrate is disposed between the X-ray source and the X-ray detector, and the polycrystalline silicon substrate is provided with relative rotation between the X-ray source and the pair of X-ray detectors and the polycrystalline silicon substrate. Polycrystalline silicon that inspects the presence or absence of defects by irradiating a crystalline silicon substrate with X-rays and using the X-ray transmission data taken at every predetermined rotation angle to distinguish the grain boundaries and cracks of the polycrystalline silicon substrate. A defect inspection method for a substrate,
Spatial variation data calculation step for calculating spatial variation data related to the spatial variation of the X-ray absorption rate of each part of the polycrystalline silicon substrate based on the X-ray transmission data;
Emphasizing the use of the spatial variation data to construct a planar image of the polycrystalline silicon substrate, and generating a planar image in which the portion where the spatial variation of the X-ray absorption rate is large is enhanced in the polycrystalline silicon substrate An image reconstruction calculation step;
A threshold setting step for setting a threshold for the crack based on the size of the grain boundary of the polycrystalline silicon substrate;
Comparing the size of the part in the planar image generated by the emphasized image reconstruction calculation step with the threshold value, and detecting a defect larger than the threshold value as a defect; A defect inspection method for a polycrystalline silicon substrate, comprising:

Images