JP4909215B2 - Inspection method and apparatus - Google Patents

Description

  The present invention relates to a foreign substance inspection technique, and more particularly to a technique for detecting a minute foreign substance on a surface to be inspected with high sensitivity. For example, a semiconductor wafer (hereinafter referred to as a wafer) is detected by detecting the foreign substance adhering to the surface. It is related to effective technology.

Today, semiconductor integrated circuit devices (hereinafter referred to as ICs) are highly integrated and circuit patterns are miniaturized, and the line width of circuit patterns is about 1 μm or less. In order to manufacture such an IC at a high yield, foreign matter adhering to the surface of the wafer is detected, its size, shape, physical properties, etc. are inspected, and the cleanliness of various semiconductor manufacturing apparatuses and parts is quantitatively determined. It is necessary to grasp and accurately manage the manufacturing process.
Therefore, conventionally, in an IC manufacturing factory, a wafer foreign matter inspection method using a wafer foreign matter inspection apparatus is performed on a wafer as a workpiece in order to accurately manage a manufacturing process.

Conventional wafer foreign matter inspection apparatuses are roughly divided into two categories.
The first is an image comparison type wafer foreign matter inspection apparatus (hereinafter referred to as an appearance inspection apparatus) that compares an image in a bright field by vertical epi-illumination with a standard pattern stored in advance.
Second, a wafer foreign matter inspection apparatus (hereinafter referred to as foreign matter inspection) that detects scattered light in a dark field by oblique illumination and recognizes the presence or absence of foreign matter, the position coordinates and the number of foreign matters based on the coordinates when the scattered light is detected. It is called a device.)

  An example of a wafer foreign matter inspection apparatus is “Nikkei Micro Devices March 1997 issue” P97 to P116 issued by Nikkei BP Co., Ltd.

The appearance inspection apparatus described above has an advantage of high inspection accuracy, but has a disadvantage of low throughput and high price. Since the image data is obtained according to the appearance inspection apparatus, the appearance inspection apparatus can execute a so-called review (confirmation or verification work with an image).
However, the present inventor has revealed that the appearance inspection apparatus has a problem that the review defect efficiency is extremely low because the review defect efficiency is very low because there is a lot of unnecessary information such as minute defects for a small number of inspection wafers. It was done.

  The foreign matter inspection apparatus described above has the disadvantage that the inspection accuracy is lower than that of the appearance inspection apparatus, but has the advantages of higher throughput and lower price than the appearance inspection apparatus. Since the data obtained from the foreign matter inspection apparatus is the position coordinates of the foreign matter within the wafer and the intensity of the scattered light, the foreign matter inspection apparatus cannot obtain information on the size (particle size) and shape of the foreign matter. Therefore, in order to obtain such information, it is necessary to use an analysis system wafer foreign matter analysis apparatus having a long inspection time, such as an appearance inspection apparatus or an SEM (scanning electron beam microscope).

  An object of the present invention is to provide an inspection technique capable of efficiently capturing a fatal defect and inspecting the size and shape of a foreign substance.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  An outline of typical inventions among inventions disclosed in the present application will be described as follows.

That is, the inspection target image acquisition step of capturing the inspection target position in the inspection target with the pattern formed under bright field illumination to acquire the inspection target image, and the inspection target position in the inspection target A comparison target image acquisition step of capturing a corresponding comparison target position under bright field illumination to acquire a comparison target image; a sum image acquisition step of acquiring a sum image of the inspection target image and the comparison target image; About the sum image in which the image to be inspected and the image of the pattern are added among the sum images, the brightness of the image position of the pattern of the sum image and the brightness of the image position of the pattern of the comparison target image as the threshold brightness between the brightness executes threshold processing of the image and brightness is configured to divide the image composed of pixels highly than the threshold value at a lower pixel than said threshold value, said threshold value processing You characterized in that it and a determination step of determining the sum image of a single foreign substance in the case where the image of the pattern of the sum image is recognized the divided image after.

  According to the above-described means, it is possible to efficiently catch a fatal defect and inspect the size and shape of the foreign matter.

  FIG. 1 is a flowchart showing a foreign substance inspection method according to an embodiment of the present invention. FIG. 2 is a perspective view showing the foreign substance inspection apparatus. FIG. 3 and subsequent figures are explanatory diagrams for explaining the operation.

In this embodiment, the foreign matter inspection apparatus according to the present invention detects the scattered light in the dark field by oblique illumination on the wafer that is the object to be inspected, and the presence / absence of the foreign matter and the position coordinates based on the coordinates when the scattered light is detected. And a foreign matter inspection apparatus 10 that recognizes the number of pieces.
The wafer 1 to be inspected is in the process of forming a DRAM, which is an example of an IC, on the first main surface 2 for each chip portion 4, and the chip portion 4 is oriented flat (hereinafter referred to as “the orientation flat”). It is regularly arranged vertically and horizontally with respect to 3.
That is, the chip portion 4 having the same pattern is repeatedly patterned on the wafer 1. If the foreign matter 5 adheres to the first main surface 2 of the wafer 1, it causes a defect. Therefore, the foreign matter 5 attached to the first main surface 2 of the wafer 1 is detected by the foreign matter inspection apparatus 10, and the detected foreign matter is detected. The position, number, size, shape, color, and properties of the semiconductor are inspected, the cleanliness of various semiconductor manufacturing apparatuses and processes is quantitatively grasped, and the manufacturing process is accurately managed.
Further, it is determined based on the position and size of the foreign matter whether the foreign matter 5 causes a fatal failure such as a disconnection or a short circuit of the chip portion 4.

The foreign substance inspection apparatus 10 includes a stage apparatus 11. The stage apparatus 11 has an XY table 12 for scanning the wafer 1 as an object to be inspected, a θ table 13 rotated in the θ direction, and an automatic focusing mechanism ( (Not shown) and a controller 14 for controlling them.
Then, in order to inspect the entire surface of the wafer 1, XY scanning of the wafer 1 is executed by the stage device 11. During this scanning, the coordinate position information about the wafer 1 as the inspection object is sequentially input from the controller 14 to a foreign substance determination device described later.

  An inspection light irradiation device 20 is installed obliquely above the stage device 11. The inspection light irradiation device 20 includes a laser light irradiation device 22 that irradiates the wafer 1 with laser light 21 as inspection light, and a condensing lens 23 that condenses the laser light 21. By irradiating the wafer 1 as an object to be inspected held on the stage device 11 at a low angle, the wafer 1 is illuminated obliquely.

A scattered light detection device 30 is installed directly above the stage device 11. The scattered light detection device 30 includes an objective lens 32 that collects scattered light 31 irregularly reflected on the surface of the wafer 1 as the laser light 21 is obliquely irradiated on the surface of the wafer 1, and an objective lens 32. A relay lens 33 that forms an image of the collected scattered light 31 on the light receiving surface of the scattered light detector 34 and a scattered light detector 34 that detects the scattered light 31 are provided. That is, the scattered light detection device 30 is configured to detect the scattered light 31 in the dark field.
In this embodiment, the scattered light detector 34 is configured by a line sensor in which solid-state imaging photoelectric conversion elements are arranged in an elongated shape, and is arranged so as to be long in the Y direction orthogonal to the stage moving direction.

  A foreign matter determination device 35 is connected to the scattered light detector 34. The foreign matter determination device 35 determines the presence or absence of foreign matter on the wafer 1 based on the detection time of scattered light from the scattered light detector 34. By collating the determined data with the coordinate position data from the controller 14 of the stage apparatus 11, the coordinate position of the foreign object is specified. Further, the scattered light detector 34 transmits the scattered light intensity to the foreign matter determination device 35.

  In the present embodiment, an epi-illumination device 40 is provided directly above the stage device 11, and the epi-illumination device 40 irradiates the wafer 1 with white light 41 as bright-field illumination light, A half mirror 43 that vertically reflects the white light 41 and a lens 44 that forms the white light 41 in a spot shape are provided, and the white light 41 is applied to the wafer 1 as an inspection object held on the stage device 11. It is designed to illuminate vertically and illuminate epi-illumination.

  An imaging device 45 is installed at the reflection position of the epi-illumination device 40. That is, the imaging device 45 is configured by a line sensor in which solid-state imaging photoelectric conversion elements are arranged in an elongated shape, and is long in the Y direction orthogonal to the stage moving direction on the optical axis of the epi-illumination device 40 on the transmission side of the half mirror 43. It is arranged to be. That is, the image pickup device 45 takes an image of reflected light under a bright field. The epi-illumination device 40, the imaging device 45, and the like constitute an image acquisition unit.

  A comparison unit 47 is connected to the image processing unit 46 of the imaging device 45, and a verification unit 48 is connected to the output terminal of the comparison unit 47. A single foreign object determination and classification unit (hereinafter abbreviated as a classification unit) 49 is connected to the output end of the verification unit 48. A foreign substance determination device 35 is connected to the other input terminal of the comparison unit 47 and the other input terminal of the verification unit 48, and the foreign substance determination device 35 is a host computer 36 that controls the foreign substance inspection device 10, the comparison unit 47, and the verification unit. The determination result is transmitted to the unit 48.

  Next, a foreign substance inspection method according to an embodiment of the present invention using the foreign substance inspection apparatus 10 according to the above configuration will be described with reference to FIG.

  When laser light 21 as inspection light is irradiated onto the wafer 1 by the inspection light irradiation device 20 at a low inclination angle, the foreign matter 5 and the circuit attached to the first main surface 2 of the wafer 1 are irradiated by the laser light 21 irradiation. Scattered light 31 in the dark field is generated from the pattern. The scattered light 31 is collected by the objective lens 32 and imaged on the scattered light detector 34 through the relay lens 33.

At this time, since the scattered light 31 from the circuit pattern has regularity, a light shielding element (not shown) made of a spatial filter or an analyzer provided on the Fourier transform surface of the pattern surface of the wafer 1 causes the scattered light 31 from the circuit pattern. The scattered light 31 is shielded.
On the other hand, since the scattered light 31 from the foreign material 5 is irregular, it passes through a spatial filter or analyzer and forms an image on the scattered light detector 34. Therefore, only the foreign object 5 is detected.

  Then, the detection signal of the scattered light 31 in the dark field from the foreign object 5 detected by the scattered light detector 34 is input to the foreign object determination device 35. The foreign matter determination device 35 determines the presence / absence of the foreign matter 5 based on the detection signal, and specifies the coordinate position of the foreign matter 5 by comparing this determination data with the coordinate position data from the controller 14 of the stage device 11. To do. The coordinate position of the foreign matter 5 specified in this way is output from the foreign matter determination device 35 to, for example, a host computer 36 that comprehensively executes the foreign matter inspection device 10, and a comparison unit that is electrically connected to the imaging device 45. 47 and the verification unit 48.

  When the coordinate position of the foreign material 5 transmitted from the foreign material determination device 35 comes to the imaging position of the imaging device 45, that is, the epi-illumination spot of the epi-illumination device 40, the comparison unit 47 captures an image signal in the bright field from the imaging device 45. . According to the foreign substance determination device 35, since the foreign substance 5 should be attached to this coordinate position, the foreign substance 5 should appear in the image as shown in FIG. That is, the image to be inspected is acquired.

Thereafter, the foreign substance 5 adheres to the coordinate position shifted by exactly one chip part 4 from the coordinate position of the foreign substance 5 transmitted from the foreign substance determination device 35, that is, the chip part 4 adjacent to the chip part 4 to which the foreign substance 5 has adhered. When the coordinate position corresponding to the position comes to the imaging position of the imaging device 45, the comparison unit 47 captures the image signal under the bright field from the imaging device 45.
According to the foreign matter determination device 35, the foreign matter 5 should not be attached to this coordinate position, and therefore the foreign matter 5 should not appear in the image as shown in FIG. That is, the comparison target image is acquired.

Subsequently, as shown in FIG. 3, the comparison unit 47 has a coordinate position of the tip portion 4 to which the previously captured foreign substance 5 should be attached, that is, a test target image at the test target coordinate position ( From a), the coordinate position of the tip portion 4 to which the foreign substance 5 taken in later is not attached, that is, the comparison target image (b) at the comparison target coordinate position is subtracted.
That is, as shown in FIG. 3D, the comparison unit 47 is in a state of taking in and subtracting a pair of images of the same portion in the pair of adjacent chip units 4 and 4 respectively. FIG. 3C is a difference image between the image to be inspected with the foreign material 5 in FIG. 3A and the comparison target image without the foreign material 5 in FIG. Is formed. Since the difference between the image of (a) and the image of (b) is only the foreign object 5, only the foreign object image 6 is extracted from the difference image of FIG. 3 (c).

  Then, the comparison unit 47 transmits the extracted foreign object image 6 to the verification unit 48. When the foreign object image 6 is transmitted from the comparison unit 47 with respect to the coordinate position of the foreign object 5 transmitted from the foreign object determination device 35, the verification unit 48 determines that the determination of the presence of the foreign object by the foreign object determination device 35 is appropriate. Validate. On the other hand, when the foreign object image 6 is not transmitted from the comparison unit 47 for the coordinate position of the foreign object 5 transmitted from the foreign object determination device 35, the verification unit 48 determines that there is a foreign object in the foreign object determination device 35. Is recognized as an error. If the verification unit 48 determines that it is an erroneous determination, it transmits that fact to the host computer 36.

In addition, the extracted foreign object image 6 is transferred to the classification unit 49. The classification unit 49 classifies the organic substance or inorganic substance that is the size, shape, color, and property of the foreign substance 5 by a preset algorithm.
For example, as shown in FIG. 4A, the vertical a and horizontal b dimensions of the foreign object image 6 are specified by counting the number of pixels of the foreign object image 6. The area, that is, the size of the foreign object 5 is specified as shown in FIG. 4B by the product (a × b) of the vertical a and horizontal b of the foreign object image 6. The shape of the foreign object 5 is specified by the quotient (a / b) of the vertical a and the horizontal b of the foreign object image 6.
For example, when a / b = 1, the foreign material 5 is specified as being circular as shown in FIG. When a / b> 1 or a / b <1, the foreign material 5 is specified as having an elongated shape as shown in FIG.

  Since the white light 41 is used as the epi-illumination light, the color can be specified on the basis of the hue and brightness (gradation constituting the image) of the foreign object image 6. As a result, for example, if the entire extracted foreign matter has substantially the same dark blue hue and gradation, it can be estimated that the foreign matter is a dielectric film that is flat and has a substantially uniform film thickness.

  Further, the unevenness of the surface can be recognized by using the scattered light signal of the foreign matter detected by the scattered light detector and the size and property of the foreign matter specified by the foreign matter image 6. That is, when the laser light 21 as the inspection light is irradiated onto the wafer 1 by the inspection light irradiation device 20 at a low inclination angle, the scattered light 31 in the dark field is generated from the foreign matter 5 by the irradiation of the laser light 21. . Since this scattered light intensity changes in relation to the scattering cross section and surface irregularities, the scattered light intensity recorded at the time of detection and the length of the foreign matter in the direction perpendicular to the laser irradiation direction calculated from the foreign matter image 6 are obtained. By using and analyzing, it is possible to recognize the surface irregularity state. As a result, when there are few unevenness | corrugations, for example, it specifies with inorganic substances, such as a silica (glass), a silicon piece, and a metal piece. When there are many unevenness | corrugations, it identifies with organic matters, such as dust generation and resin resulting from a human being.

  The classification result of the foreign object 5 classified as described above is transmitted from the classification unit 49 to the host computer 36. The host computer 36 uses the classification data from the classification unit 49 and the coordinate position data and the number data of the foreign substance 5 from the foreign substance determination device 35 to appropriately create various monitoring materials shown in FIG. Output in a timely manner by an output device such as a printer. The operator can accurately and promptly manage the IC manufacturing process based on the various analysis materials output.

  FIG. 5A is a foreign matter size map, which is created from foreign matter size data and foreign matter coordinate position data. FIG. 5B is a histogram for each foreign substance size. The horizontal axis indicates the foreign substance size as a variable divided into sections, and the vertical axis indicates the number of detections as the number of measurement values belonging to each foreign substance size.

  FIG. 5C is a foreign matter shape map, which is created from foreign matter shape data and foreign matter coordinate position data. FIG. 5D is a foreign object shape-specific histogram, where the horizontal axis represents the foreign object shape as a variable divided into sections, and the vertical axis represents the number of detected values as the number of measurement values belonging to each foreign object shape.

  FIG. 5E is an inspection result time-series transition graph, which is created from the inspection date, foreign matter size data, foreign matter shape data (for example, elongated), and organic matter data that is an example of foreign matter properties. Note that a lot number, a wafer number, or the like may be used for the time series.

  By the way, when the foreign object is attached so as to straddle the adjacent wiring patterns, the foreign object image is obtained by subtracting the comparison target image at the comparison target coordinate position from the inspection target image at the inspection target coordinate position. It has been discovered by the present inventor that the phenomenon of division occurs as shown in FIG. The reason why this phenomenon occurs can be considered as follows.

FIG. 6A shows an inspection target image at the inspection target coordinate position. A pair of wiring pattern images 8 and 8 laid in parallel to each other are projected on the background image 7, and both wiring patterns are shown. The foreign object image 6 is projected over the images 8 and 8.
FIG. 6B shows a comparison target image at the comparison target coordinate position. Only the wiring pattern images 8 and 8 are displayed on the background image 7.

  Assuming that the value of the binarized image signal of the background image 7 is white when subtracting in the comparison unit 47, the value of the foreign object image 6 corresponds to black, and the wiring pattern image 8 corresponds to gray. To do. Accordingly, when FIG. 6B is subtracted from FIG. 6A, the difference image 9 is a state in which the wiring pattern image 8 remains light gray in the foreign object image 6 as shown in FIG. 6D. become. However, when the difference image 9 is binarized, the binarized difference image 90 is divided into three by the two wiring pattern images 8 and 8 as shown in FIG. Become.

If the foreign object image 6 is recognized in a divided state, the classification unit 49 may classify the foreign object image 6 as a continuous small foreign object. Therefore, the foreign object 5 has a fatal failure such as a disconnection failure or a short circuit failure. In spite of this, there is a possibility that it is erroneously determined that it is a minute foreign matter that does not cause a fatal defect.
For example, in the foreign substance inspection process after the wiring pattern formation process, if it is erroneously determined that a metal foreign substance straddles between the wiring patterns is a minute foreign substance existing between the wiring patterns, a short circuit failure Will miss the deadly countermeasures. For this reason, a yield will fall. Therefore, even when the difference image 90 is recognized to be minute, it is necessary to determine whether the foreign image is divided by the wiring pattern image or the minute foreign matter existing between the wiring patterns.

In the present embodiment, the classification unit 49 determines whether the difference image is a divided difference image by the following algorithm.
First, when the difference image is divided into three or more by the adjacent wiring pattern, as shown in FIG. 6C, the difference image 90 is divided into the left divided image 91, the central divided image 92, and the right divided image. A description will be given by taking the case of being divided into 93 as an example.

As shown in FIG. 6C, the wiring pattern image 8 is calculated from the distance d between the left divided image 91 and the central divided image 92 and the distance d between the central divided image 92 and the right divided image 93. Is subtracted. If the difference value obtained by the subtraction is smaller than a preset value ε, the difference image 90 of a single foreign object is recognized.
6D, the width X of the difference image 90 in the direction orthogonal to the wiring pattern image 8 is the distance between the left end of the left divided image 91 and the right end of the right divided image 93. Measured by. Thus, it is determined that the foreign matter recognized as the single foreign matter difference image 90 causes a fatal defect.

When the difference value by subtraction is larger than a preset value ε, the left divided image 91, the central divided image 92, and the right divided image 93 are recognized as separate minute foreign matters, and each width x is set to be equal to each foreign matter. Measured as size.
When the size of the left divided image 91, the central divided image 92, and the right divided image 93 is larger than the distance S between the wiring pattern images 8, 8, there is a possibility of causing a fatal failure, that is, a defect candidate. If it is small, it is determined that there is no possibility of causing a fatal failure.

As shown in FIG. 7C, when the left divided image 94 and the right divided image 95 are divided into two, the pattern is determined from the distance d between the left divided image 94 and the right divided image 95. The width L is subtracted. When the difference value by subtraction is smaller than a preset value ε, it is determined that the difference image 90 is a single foreign object.
That is, as shown in FIG. 7C, the width X of the difference image 90 in the direction orthogonal to the wiring pattern image 8 is the distance between the left end of the left divided image 94 and the right end of the right divided image 95. Measured by. Then, this foreign substance is determined to have a possibility of causing a fatal defect, that is, a defect candidate.

  When the difference value by subtraction is larger than a preset value ε, the left divided image 94 and the right divided image 95 are recognized as separate minute foreign matters, and each width x is measured as each foreign particle size. The If each size of the left divided image 94 and the right divided image 95 is larger than the distance S between the wiring pattern images 8 and 8, it is determined that the fatal defect may be caused, that is, a defect candidate, and is small. In this case, it is determined that there is no possibility of causing a fatal failure.

  Information on all the foreign substances detected and determined as described above and the determination results of fatal defects or defect candidates are collectively displayed as shown in Table 1 below.

Further, wafer inspection results such as wafer inspection information, the number of foreign matters, the number of foreign matter attached chips, and the wafer yield by process are collectively displayed as shown in Table 2.

Here, the chip non-defective rate yi is the ratio of the total of the chips having no fatal defects and no defect candidates to the total chips, and the chip defective product ratio fi is the ratio of the total of the chips having the fatal defects to all the chips. The defect density is the number of foreign matters / defects per unit area. According to the following definition, the defect density is displayed in two types, the maximum and minimum of the estimated range, respectively.
The defect density Di is
Di (max) = − (1 / A) × Ln (yi)
Di (min) = − (1 / A) × Ln × (1-fi)
In the formula, A is a chip area and Ln is a natural logarithm.
The defective wafer density Wi is
Wi (max) = (number of fatal defects + number of defect candidate defects) / (A × number of product chips)
Wi (min) = (number of fatal defects) / (A × number of product chips)
It is.

  Further, as shown in FIG. 8, fatal defects and defective chips are displayed during the wafer map display.

When the inspection of one lot is completed, inspection result information for each lot is displayed as shown in Tables 3 and 4. Further, the inspection result information is transmitted to an external system such as an inspection information analysis system using communication means.

The scattered light detection type foreign matter inspection apparatus detects only the foreign matter on the outermost surface of the wafer. Therefore, if there is a foreign matter on the pattern, the product becomes defective. By managing the non-defective product rate and defective product rate in each inspection process, the yield in the final inspection can be predicted. In addition, from the distribution of defective chips, it can be determined whether the foreign matter is a random foreign matter, in the case of a random foreign matter, assuming that the distribution of foreign matter is Poisson distribution,
yi = exp (−A × Di) (where A is the chip area)
And the defect density Di (subscript i indicates a process) of each process can be calculated. Accordingly, in the case of a defect, measures against foreign matter / defects in the process are performed based on the deviation from the defect density management value allocated to each process.
The foreign matter / defect management method in these intermediate processes is different from the defect density calculated from the number of foreign matter / defects and the inspection area, which is the conventional method,
y = exp (−A × D) (where y is the yield, A is the chip area, and D is the defect density)
Since the defect density Di of each process is calculated by a method similar to the above, it is possible to extract a defective process that is directly effective in improving the yield of the final process.

  According to the embodiment, the following effects can be obtained.

1) Since the size of foreign matter can be output without using a long inspection device or multiple inspection devices, the time required for inspection per wafer can be greatly reduced. , Can contribute to early yield improvement.

2) By providing inspection data directly related to the reduction of defect density in the yield model in probe inspection, it is possible to quickly determine the points for improving yield and the pros and cons of measures.

3) By verifying the determination of the foreign matter determination unit based on the scattered light detector by the verification unit connected to the imaging device under the bright field, the inspection accuracy in the foreign matter inspection device can be increased. The quality and reliability of the foreign substance inspection apparatus can be improved.

4) The size of the foreign matter can be measured more accurately by determining whether or not the divided portion is a single foreign matter and measuring the particle size.

5) Since the lethality of the foreign matter can be determined from the ground pattern and the size and properties of the foreign matter, the efficiency of the foreign matter inspection method can be further enhanced.

6) Since the non-defective product rate and defect density of the inspection wafer can be calculated, the utilization efficiency of the foreign matter inspection method can be further increased.

  FIG. 9 is an explanatory diagram for explaining the recognition operation of a single foreign object in the foreign object inspection method according to the second embodiment of the present invention.

  The present embodiment is different from the above embodiment in a method for recognizing whether or not a foreign object image is a single foreign object when divided by a wiring pattern image. That is, as shown in FIGS. 9A, 9B, and 9C, when the difference image 90 is divided into three by subtraction processing, color determination of the divided image is executed. .

When the image from the imaging device 45 is a color image and is output as an NTSC (National Television System Committee) signal, color determination using the saturation hue distribution diagram shown in FIG. Is executed.
When the left divided image 91, the central divided image 92, and the right divided image 93 shown in FIG. 9C are determined to have the same color in the areas Δγ and Δθ that are predetermined in the saturation hue distribution diagram, It is recognized as a difference image 90 of a single foreign object, and the size of the foreign object is measured by grouping the divided images, and is determined to be fatal. If it is determined that the colors are not the same, they are recognized as different foreign substances, and the foreign substance size is measured for each. If the size of the foreign material is larger than the distance between the patterns, it is determined as fatal failure, and if it is smaller, it is determined as non-fatal.

  When the image from the imaging device 45 is a color image and is output as an RGB signal, the color is expressed by αR + βG + γB, and | Δα | + | Δβ | + | Δγ with respect to a preset value η. If | ≦ η, the same color is determined. If it is determined that the colors are the same, it is recognized as a single foreign object, foreign particle size measurement is performed by grouping the divided images, and it is determined that the defect is fatal. If it is determined that they are not the same color, they are recognized as different foreign matters, and the foreign matter size is measured for each. If the size of the foreign material is larger than the distance between the patterns, it is determined to be fatal.

  As shown in FIG. 10, even when the number of divisions is two, the color determination between the divided images is executed in the same manner as when the color determination is divided into three or more. If the same color is determined, it is recognized as a single foreign object, foreign particle size measurement is performed by grouping the divided images, and this foreign object is determined as a defect candidate. The inspection result is processed in the same manner as in the above embodiment.

  FIG. 11 is an explanatory diagram for explaining the recognition operation of a single foreign object in the foreign object inspection method according to the third embodiment of the present invention.

  The present embodiment is different from the above embodiment in a method for recognizing whether or not a foreign object image is a single foreign object when divided by a wiring pattern image. That is, as shown in FIG. 11, the recognition method is executed by the expansion processing of the divided images.

As shown in FIG. 11, when the left divided image 91, the central divided image 92, and the right divided image 93 are divided, the divided elements are divided by the numerical value ζ inputted in advance, the left divided image 91, The center divided image 92 and the right divided image 93 are expanded as shown in FIG.
In this embodiment, the numerical value ζ is set to the space width of the pattern. As a result of the expansion process, when the number of divisions after the expansion process is smaller than the original number of divisions, the divided elements combined in each expansion process are combined into one and recognized as a single foreign object. The foreign object size measurement is executed by measuring the distance between the left end of the original left divided image 91 and the right end of the right divided image 93, and a single foreign object is determined to be fatal. When the number of division elements after the expansion process is the same as the original number of divisions, this foreign substance group is recognized as a separate foreign substance and determined as a defect candidate. The inspection result is processed in the same manner as in the above embodiment.

  FIG. 12 is an explanatory diagram for explaining the recognition operation of a single foreign object in the foreign object inspection method according to the fourth embodiment of the present invention.

  The present embodiment is different from the above embodiment in a method for recognizing whether or not a foreign object image is a single foreign object when divided by a wiring pattern image. That is, as shown in FIG. 12, the recognition method is executed using the sum image.

FIG. 12A shows an inspection target image at the inspection target coordinate position, and a pair of wiring pattern images 8 and 8 laid in parallel with each other are projected on the background image 7. The foreign object image 6 is projected over the images 8 and 8.
FIG. 12B shows a comparison target image at the comparison target coordinate position, and both wiring pattern images 8 and 8 are projected on the background image 7. The inspected image and the wiring pattern images 8 and 8 are added to form a sum image 9 ′ as shown in FIG.
Threshold processing is executed for this sum image 9 ′, and wiring pattern images 8 ′ and 8 ′ divided by the foreign object image 6 shown in FIG. 12D are formed. When the wiring pattern images 8 ′ and 8 ′ divided in this way are recognized, the foreign object is determined to be fatal. If not recognized, it is determined as non-fatal.
The inspection result is processed in the same manner as in the above embodiment.

  Although the invention made by the present inventor has been specifically described based on the embodiment, the present invention is not limited to the embodiment, and various modifications can be made without departing from the gist of the invention. Needless to say.

  For example, the specification of the position of the foreign matter by the scattered light detection in the dark field is not limited to be executed by the light shielding element, but may be executed by comparing the detection data at the same position in the repeated pattern. Good. In this case, the comparison data may be detection data of adjacent chips, or may be design pattern data or standard pattern data stored in advance.

  As an imaging device, in the example using a line sensor, not only a line sensor but an area sensor, an imaging tube, etc. can be used.

  In the above description, the case where the invention made mainly by the present inventor is applied to the wafer foreign matter inspection technology, which is the field of use behind the invention, has been described. However, the present invention is not limited to this. The present invention can be applied to all foreign matter inspection techniques for plate-like objects.

  The effects obtained by the representative ones of the inventions disclosed in the present application will be briefly described as follows.

  Since the size of the foreign matter can be output without using an apparatus having a long inspection time or a plurality of inspection apparatuses, the time required for the inspection per inspection object can be greatly shortened. By determining whether or not the divided portion is a single foreign substance and measuring the size, the size of the foreign substance can be measured more accurately. By determining the lethality of the foreign matter from the pattern and the size and properties of the foreign matter, the efficiency of the foreign matter inspection method can be further enhanced.

  By calculating the non-defective product rate and defect density of the inspection wafer, the utilization efficiency of the foreign matter inspection method can be further enhanced.

It is a flowchart which shows the foreign material inspection method which is one Embodiment of this invention. It is a perspective view which similarly shows a foreign material inspection apparatus. FIG. 4A illustrates the action of extracting a foreign object, where FIG. 5A is an image diagram with a foreign object attached, FIG. 5B is an image diagram with no foreign object attached, FIG. It is a perspective view of an imaging situation. It is each explanatory drawing which shows the classification | category effect | action, (a) is the specific effect | action of the vertical and horizontal dimension of a foreign material, (b) is a specific operation of a foreign material size, (c) is a specific operation of a circular foreign material, (d) is an elongated foreign material Each shows a specific action. It is each explanatory drawing which shows various analysis data, (a) is a map according to foreign material size, (b) is a histogram according to foreign material size, (c) is a map according to foreign material shape, (d) is a histogram according to foreign material shape, (e) Is a test result time series transition graph. It is each explanatory drawing for demonstrating the recognition effect | action of the single foreign material when a foreign material image is divided into three. It is each explanatory drawing for demonstrating the recognition effect | action of the single foreign material when a foreign material image is divided into two. It is a wafer map figure on which a fatal defect and a defective chip are displayed. It is explanatory drawing for demonstrating the recognition effect | action of the single foreign material of the foreign material inspection method which is Embodiment 2 of this invention, (a), (b), (c) is each image figure, (d) is saturation. It is a hue distribution diagram. It is each explanatory drawing for demonstrating the recognition effect | action of the single foreign material when the division | segmentation number is two similarly. It is each explanatory drawing for demonstrating the recognition effect | action of the single foreign material of the foreign material inspection method which is Embodiment 3 of this invention. It is each explanatory drawing for demonstrating the recognition effect | action of the single foreign material of the foreign material inspection method which is Embodiment 4 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Wafer (inspection object), 2 ... 1st main surface, 3 ... Orientation flat, 4 ... Chip part, 5 ... Foreign material,
6 ... Foreign object image, 7 ... Background image, 8 ... Wiring pattern image, 9 ... Difference image, 9 '... Japanese image,
DESCRIPTION OF SYMBOLS 10 ... Foreign object inspection apparatus, 11 ... Stage apparatus, 12 ... XY table, 13 ... (theta) table, 14 ... Controller,
DESCRIPTION OF SYMBOLS 20 ... Inspection light irradiation apparatus, 21 ... Laser light (inspection light), 22 ... Laser light irradiation apparatus, 23 ... Condensing lens,
DESCRIPTION OF SYMBOLS 30 ... Scattered light detection apparatus, 31 ... Scattered light, 32 ... Objective lens, 33 ... Relay lens, 34 ... Scattered light detector, 35 ... Foreign substance determination apparatus, 36 ... Host computer,
40 ... Epi-illumination device, 41 ... White light (bright-field illumination light), 42 ... White light irradiation device, 43 ... Half mirror, 44 ... Lens, 45 ... Imaging device, 46 ... Image processing unit, 47 ... Comparison unit, 48 ... Verification unit, 49 ... Single foreign matter determination and classification unit,
90: binarized difference image, 91: left divided image, 92 ... center divided image, 93 ... right divided image, 94 ... left divided image, 95 ... right divided image.

Claims (7)

Inspected image acquisition step of acquiring an image to be inspected by taking an image of the position to be inspected under bright field illumination in the inspected object in which the pattern is formed,
A comparison target image acquisition step of capturing a comparison target position corresponding to the inspection target position under bright field illumination and acquiring a comparison target image in a state where no foreign matter exists in the inspection target;
A sum image acquisition step of acquiring a sum image of the image to be inspected and the image to be compared;
The image position of the pattern of the comparison target image that is a comparison target position corresponding to the pixel value of the image position where no foreign matter is present on the pattern in the inspection target image with respect to the sum image When a threshold value process is performed to leave an image area having a value obtained by adding the pixel values of the pixel values, and it is recognized that the image area after the threshold value process is obtained by dividing the image of the pattern, A determination step of determining an image;
An inspection method comprising:   The pixel value is a hue or brightness
The inspection method according to claim 1. 3. The inspection method according to claim 1, wherein the determination step determines that the fatal defect is determined when the sum image of the single foreign matter is determined. An inspection target image acquisition unit that acquires an inspection target image by capturing an image of an inspection target position specified in advance in a bright field illumination in the inspection target formed with a pattern;
A comparison target image acquisition unit that captures a comparison target position corresponding to the inspection target position under bright field illumination to obtain a comparison target image in a state in which no foreign matter exists in the inspection target;
A sum image acquisition unit for acquiring a sum image of the inspection target image and the comparison target image;
The image position of the pattern of the comparison target image that is a comparison target position corresponding to the pixel value of the image position where no foreign matter is present on the pattern in the inspection target image with respect to the sum image When a threshold value process is performed to leave an image area having a value obtained by adding the pixel values of the pixel values, and it is recognized that the image area after the threshold value process is obtained by dividing the image of the pattern, A determination unit for determining an image;
An inspection apparatus comprising:   The pixel value is a hue or brightness
The inspection apparatus according to claim 4. The inspection apparatus according to claim 4 , wherein the determination unit determines that the fatal defect is determined when the sum image is determined to be a sum image of the single foreign matter. The inspection according to any one of claims 4 to 6, wherein the inspection target image acquisition unit and / or the comparison target image acquisition unit includes a line sensor, an area sensor, or an imaging tube. apparatus.

Images