JPH07120403A - Defect inspection method and device therefor - Google Patents

Abstract

PURPOSE:To shorten the inspection time as a whole, in a defect inspection method for carrying out defect detection and defect observation. CONSTITUTION:The surface of a reticle 1 is divided into many cells E(1,1), E(1,2),..., the sample value Vij1 of the detected signal of the maximum defect in every cell E(i,j) is obtained, and the center coordinates and the sample value of the maximum deffect of each cell are stored in a memory in order of detection. Thereafter, in order of the cell having larger maximum defect in the interior, the defect observation is carried out by the observation visual field 15a of an observation part on each cell unit. If a defective part is observed, the defect inspection of the reticle 1 is completed at the time point.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention provides a surface of a mask or reticle (collectively referred to as a "reticle" hereinafter) used in an exposure process for manufacturing a semiconductor device or the like, or a predetermined interval on the surface of the reticle. The present invention relates to a defect inspection method and apparatus suitable for application when detecting a defect such as a foreign substance attached on the surface of a pellicle (dustproof film) that is opened and stretched.

[0002]

2. Description of the Related Art For example, if a surface of an exposure reticle used for manufacturing a semiconductor device or the like using a photolithography technique has a defect such as a foreign substance, the yield of the manufactured semiconductor device or the like decreases. I will end up. To avoid it
Conventionally, a defect inspection apparatus is used to inspect the state of defects on the reticle surface before exposure. Further, in order to prevent foreign matter from directly adhering to the pattern forming surface and the back surface of the reticle, a dustproof film called a pellicle may be stretched over the reticle. Since such a pellicle is in a defocused state from the reticle pattern forming surface which is conjugate with the exposure surface of the photosensitive substrate (wafer, etc.), if foreign matter of the same size is directly attached to the reticle. When attached to the pellicle, the effect is smaller.

However, if the predetermined limit is exceeded, even foreign matter attached to the pellicle will affect the exposure result, so the front and back surfaces of the pellicle (reticle side surface).
Also, the defect inspection apparatus inspects the state of defects such as foreign matter. In the following, a reticle is used as a target for defect inspection, but it is assumed that this also includes a reticle in which a pellicle is stretched.

FIG. 6 shows an example of a conventional defect inspection apparatus,
In FIG. 6, the reticle 1 is placed on the stage 2, and the stage 2 is configured to be movable in the Y direction by the drive unit 3. The amount of movement of the stage 2 in the Y direction is constantly measured by the length measuring device 4 such as a linear encoder, and the position signal S1 indicating the length measured value of the length measuring device 4 is supplied to the signal processing circuit 5. On the other hand, a light beam L1 emitted from a light source (not shown) (for example, a laser light source) is reflected and deflected by a galvano scanner 6 (or a polygon scanner, etc.) vibrating by a driving unit 7, and then is scanned by a scanning lens 8 on the reticle 1. The light beam L2 converges on the scanning line 10 which is substantially parallel to the X direction on the reticle 1 and is scanned in the X direction. By scanning the light beam L2 in the X direction and moving the reticle 1 in the Y direction by the driving unit 3 at a speed slower than the scanning speed, the light beam scanning of the entire surface of the reticle 1 can be performed.

If there is a defect such as a foreign substance 11 on the surface of the reticle 1, the foreign substance 1 is removed by the light beam L2.
The scattered light L3 is generated from 1. The scattered light L3 is condensed by the light receiving lens 12 on the light receiving surface of the photoelectric detector 13 such as a photomultiplier, and the detection signal V obtained by photoelectrically converting the condensed light in the photoelectric detector 13 is processed. It is supplied to the circuit 5. The signal processing circuit 5 is also supplied with the deflection angle signal S2 supplied to the drive unit 7 for the galvano scanner 6, and the signal processing circuit 5 can determine the presence of the foreign matter 11 from the detection signal V. In parallel with this, the signal processing circuit 5 receives the position signal S1 of the length measuring device 4 and the deflection angle signal S2 for the drive unit 7 of the galvano scanner 6 when a signal indicating the foreign matter 11 is obtained in the detection signal V. Based on and
The existence position of 1 can be recognized. That is, the X coordinate of the foreign matter 11 can be known from the deflection angle signal S2, and the Y coordinate of the foreign matter 11 can be known from the position signal S1.

Since the amount of scattered light L3 increases as the size of the foreign matter increases, the magnitude of the detection signal V of the photoelectric detector 13 indicates the size of the foreign matter. Therefore, the signal processing circuit 5 can display the adhered coordinates (X, Y) of the foreign matter and the size of the foreign matter on the CRT display 14 as, for example, a table. Alternatively, the CRT display 14 displays the coordinates (X, Y) and the size of the foreign matter detected while scanning the reticle 1 with the light beam.
It is also possible to display it as a two-dimensional map on the display screen. Further, after storing the detected position (X, Y) and size (value of the detection signal V) of the foreign matter in a storage unit such as a memory in the signal processing circuit 5, after the inspection is finished, those positions and The size is read and the CRT display 14
It is also possible to display it as a two-dimensional map or table on the top, or to print it out by a printer (not shown).

FIG. 7 shows a map display example on the CRT display 14. In this display example, the surface of the reticle 1 of FIG. 6 is provided with a large number of small rectangular areas (hereinafter referred to as “cells”).
The foreign substance information of the entire surface of the reticle 1 is displayed in cell units in a rectangular window 17 on the display screen of FIG. That is, as shown in FIG. 6, by dividing the surface of the reticle 1 in the X direction and the Y direction at a predetermined pitch, the surface of the reticle 1 is divided into a large number of small cells C (1, 1) such as 1 mm square or 5 mm square. , C (1, 2), C (1,
3), ... Then, the windows 17 on the display screen of FIG. 7 are made to correspond to the cells on the reticle 1 of FIG. 6 in a 1: 1 relationship, and display cells P (1,1), P (1,2), P (1,
3), ..., And the symbols A, B, and C indicating defects such as foreign matter are displayed for each display cell.

In this case, the X direction and the Y direction in FIG. 6 correspond to the X1 direction and the Y1 direction in FIG. 7, respectively, and correspond to the cell C (i, j) to which the coordinate (X, Y) in which the defect is detected belongs. Defects are displayed in the display cells P (i, j) by ranking them as indicated by symbols A, B, and C according to the signal intensity of the detection signal V. Reference numerals A, B, and C are ranks indicating the sizes of defects. For example, when the detection signal V is small, the A rank indicating a small defect is displayed, and when the detection signal V is large, a large defect is displayed. The C rank is displayed.

Of course, it is possible to apply the value of the detection signal V to the detected cell as it is, but it is complicated, so that it is easier to see the rank or color display.

Further, the defect inspection apparatus is often provided with an observation section 15 shown in FIG. 6 so that the defect can be observed. Figure 6
In, the observation unit 15 is attached to the slider 16 slidably in the X direction. As the observation method, the observation unit 1
Visual observation of observing a defective portion on the surface of the reticle 1 by an optical microscope as 5, or an intermediate image of the observation portion 15 is picked up by a two-dimensional image pickup device (CCD or the like), and the picked-up image is T
There is an image observation displayed on the V monitor. In general, even if a foreign matter of a predetermined size is attached to the surface of the reticle 1, if the attachment position is a light-shielding portion to which a chrome film or the like is attached, the influence on the exposure result is small, but the attachment is small. When the position is the transmissive portion, the influence on the exposure result is large. That is, even if foreign matter of the same size is attached on the reticle 1, the reticle 1 may be disabled or enabled depending on the position where the foreign material is attached. Therefore, by observing the defect detected by the light beam scanning using the observation unit 15, it is finally determined whether or not the reticle 1 can be used.

[0011]

As described above, in the conventional defect inspection apparatus, the defect detection by the light beam scanning and the defect portion observation by the observation unit 15 are performed, but the defect detection by the light beam scanning is performed. And the observation of the defective portion cannot be performed at the same time. There are two main reasons for this.

Defect detection is performed by an oblique incidence light system in which the light beam L2 is obliquely incident as shown in FIG.
Since the observation of the defect portion by 5 is performed by epi-illumination or transmitted illumination, it is not possible to make both light sources common. When a defect is detected,
Light other than the weak scattered light from defects such as foreign matter interferes with the inspection. Therefore, when the observation unit 15 is arranged near the scanning line 10, the oblique incident beam (light beam L2) or the scattered light from the surface of the reticle 1 due to the incident beam hits the observation unit 15 and generates extra light. It must be separated from the defect detection unit.

As described above, in the conventional defect inspection apparatus, the defect detection and the defect portion observation cannot be performed at the same time, and the defect detection by the light beam scanning is performed at high speed by the scanning of the galvano scanner 6 and the movement of the stage 2. Done. Therefore, most of the inspection time is the time of defect observation using the observation unit 15, and in order to shorten the time of the inspection process, it is necessary to perform defect observation by the observation unit 15 rapidly. However, the conventional inspection method merely observes the detected defects, for example, in the order in which they are detected on the reticle 1, and does not give sufficient consideration to shortening the time of the inspection process.

In view of the above point, the present invention has an object of shortening the inspection time as a whole in a defect inspection method for performing defect detection and defect observation. Another object of the present invention is to provide a defect inspection apparatus capable of implementing such a defect inspection method.

[0015]

A first defect inspection method according to the present invention is a method of inspecting light (L) on a surface to be inspected of an object (1).
2) is irradiated, the defect on the test surface is detected based on the scattered light (L3) of the inspection light from the test surface, and the detected defect is observed at a predetermined magnification. In the inspection method, the size of the defect on the surface to be inspected is determined based on the photoelectric conversion signal V of the scattered light (L3) of the inspection light (L2) from the surface to be inspected, and the detected defect When there are a plurality of defects, the detected defects are observed at the predetermined magnification in the descending order of the defects based on the determination result, and when there is a defective portion as a result of this observation, the defects of the object (1) are detected. The inspection is completed.

In the second defect inspection method according to the present invention, the inspection light (L2) is irradiated onto the inspection surface of the inspection object (1), and the inspection light from the inspection surface is irradiated. Scattered light (L
In the defect inspection method of detecting a defect on the surface to be inspected based on 3) and observing the defect thus detected at a predetermined magnification, the surface to be inspected is divided into a plurality of cells having a predetermined size ( C (i, j)), and the maximum defect size in the plurality of cells is determined based on the photoelectric conversion signal V of the scattered light (L3) of the inspection light from the surface to be inspected. Each of the plurality of cells is discriminated, and the cell containing the defect among the plurality of cells is observed at a predetermined magnification in the order of the largest internal defects as a result of the discrimination. The defect inspection of the inspection item (1) is completed.

In the third defect inspection method according to the present invention, the inspection surface of the inspection object (1) is irradiated with inspection light (L2), and the inspection light from the inspection surface is irradiated. Scattered light (L
In the method of detecting a defect on the surface to be inspected based on 3), scattered light (L
The size of the defect on the surface to be inspected is discriminated based on the photoelectric conversion signal V of 3), and the number of detected defects exceeds a predetermined allowable number, and it is discriminated among the defects. The object to be inspected (1) when at least one of the defects whose size exceeds a predetermined allowable value is present.
This completes the defect inspection.

Further, the first and second defect inspection apparatuses according to the present invention respectively use optical scanning means (2, 3, 3) for scanning the inspection light (L2) on the surface to be inspected of the object (1). 6, 7)
A scanning position measuring means (4) for measuring the scanning position of the inspection light (L2) on the surface to be inspected, and the scattered light (L3) of the inspection light from the surface to be inspected. It has a light receiving means (13) for converting and an observing means (15) for observing the surface to be inspected at a predetermined magnification, based on the output signals of the light receiving means (13) and the scanning position measuring means (4). It is assumed that the defect inspection apparatus detects the defect on the surface to be inspected and the position of the defect and observes the defect thus detected by the observation means (15).

Then, the first defect inspection apparatus has a defect discriminating means (21) for discriminating the size of the defect on the surface to be inspected based on the output signal V of the light receiving means (13), and the defect discriminating means. A storage unit (2) that stores the size of the defect and the position on the surface to be inspected for each defect on the surface to be inspected, based on the result of the determination and the output signal of the scanning position measuring unit (4).
3) and the inspection object (1) and the observation means (15) so that the defects can be observed in the descending order based on the size of the defect stored in the storage means and the position of the defect on the inspection surface.
And a control means (22) for controlling the positional relationship with the.

On the other hand, the second defect inspection apparatus judges the size of the defect on the surface to be inspected on the basis of the output signal of the light receiving means (13) and the surface to be inspected. Divide into a plurality of cells (C (i, j)) of a predetermined size,
Based on the discrimination result of the defect discriminating means (21) and the output signal of the scanning position measuring means (4), for each cell having a defect in the plurality of cells, the maximum defect size in this cell and this cell Storage means (23) for storing the position on the surface to be inspected, cells having the defects based on the size of the defect stored in the storage means and the position of the cell on the surface to be inspected. A control means (22) for controlling the positional relationship between the inspection object (1) and the observation means (15) so that the largest defect in the cell can be observed in order.
Is to have.

Further, the third defect inspection apparatus of the present invention is the optical scanning means (2, 3, 6, 7) for scanning the inspection light (L2) on the surface to be inspected of the object (1). A scanning position measuring means (4) for measuring the scanning position of the inspection light on the surface to be inspected, and a light receiving means for photoelectrically converting scattered light of the inspection light from the surface to be inspected ( 13) and a defect inspection apparatus for detecting a defect on the surface to be inspected and the position of this defect based on the output signals of the light receiving means (13) and the scanning position measuring means (4). Based on the output signal of 13), the defect determining means (21) for determining the size of the defect on the surface to be inspected, and the determination result of the defect determining means and the output signal of the scanning position measuring means (4). , A memory for storing the size of this defect and its position on the surface to be inspected for each defect on the surface to be inspected (23), first comparing means (22) for comparing the number of defects stored in the storage means with a predetermined allowable number, and the maximum defect size among the defects stored in the storage means. Second, comparing the
And a comparison means (22).

[0022]

According to the first defect inspection method of the present invention,
When a defect such as a foreign substance on the surface to be inspected is detected, the size of the defect is discriminated, and when the defects on the surface to be inspected are then observed at a predetermined magnification, the defects having a larger size are sequentially observed.
Then, when there is a defective portion, the defect inspection of the inspection object (1) is completed, so that the inspection time is shortened. When the inspection object (1) is, for example, a reticle, an example of the defective portion is the light transmitting portion of the reticle to which foreign matter exceeding a predetermined standard is attached.

Further, according to the second defect inspection method, the surface to be inspected is composed of a plurality of cells (a large number of small areas arranged in close contact).
The internal maximum defect is extracted for each cell, and the cells with the largest internal defect are observed in order. And
Since the defect inspection of the inspection object (1) is completed when there is a defective portion, the inspection time is shortened. Moreover, since the maximum internal defect is extracted for each cell, the inspection time is shorter than that of the first defect inspection method.

Further, according to the third defect inspection method, when the number of defects exceeds the allowable value or there is a defect whose size exceeds the allowable value in the defect detection stage before observation, the inspection is performed. The object is judged to be defective, and the inspection is completed without observing even during the defect inspection. This reduces the inspection time. Further, according to the first to third defect inspection apparatuses of the present invention, the above-described first to third defect inspection methods can be carried out, respectively.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A first embodiment of the present invention will now be described with reference to FIGS.
Will be described with reference to. In this embodiment, the present invention is applied to the case where the reticle 1 (including the case where the pellicle is stretched) is inspected for defects. In FIG. 1, parts corresponding to those in FIG. The detailed description thereof will be omitted.

FIG. 1 shows the defect inspection apparatus of this embodiment. In FIG. 1, the detection signal V output from the photoelectric detector 13 is supplied to the defect discrimination circuit 21, and the defect discrimination circuit 21 detects the detection signal V. When the peak value of exceeds a predetermined threshold value, for example, it is determined as a defect, and a determination signal indicating that it is a defect,
And the sampled value V _i (i =
1, 2, ...) Is supplied to the main control system 22. Main control system 22
Is also supplied with the position signal S1 of the length measuring device 4 and the deflection angle signal S2 for the drive unit 7 of the galvano scanner 6, and the main control system 22 detects the X coordinate of the defect detected by the defect determination circuit 21 and The Y coordinate can be recognized.

Further, the slider 16 is provided with a drive unit 26 for driving the observation unit 15 in the X direction, and the main control system 22 is
By controlling the operations of the drive unit 3 and the drive unit 26 via the stage control unit 27, the position of the stage 2 in the Y direction is set to a desired position, and the position of the observation unit 15 in the X direction is set to a desired position. Set to. As a result, the observation unit 15
It is possible to quickly set a desired area on the reticle 1 to the observation area. Further, the main control system 22 includes the first memory 2
3, second memory 24, third memory 25 and keyboard 2
Connect 8. The operator can input commands and the like indicating the start and end of the defect inspection of the reticle 1 to the main control system 22 via the keyboard 28. In addition, the first memory 23 and the second memory 24 store information indicating the size and the position of the detected defect as described later, and the third memory 25 stores various other data. . The other configuration is the same as that of FIG.

Next, the defect inspection operation of this embodiment will be described with reference to FIG.
And FIG. 3 will be described. First, in FIG. 1, the entire surface of the reticle 1 is scanned with the light beam L2, and the defect coordinates (X, Y) and the sample value V _i of the detection signal V when the defect is detected by the defect determination circuit 21 are detected. The main control system 22 stores the data in the first memory 23 in this order. FIG. 2 shows an example of distribution of defects on the reticle 1. In FIG. 2, the X coordinate is determined by the deflection angle signal S2 of the driving unit 7 of the galvano scanner 6 of FIG. 1, and the position signal of the length measuring device 4 of FIG. Y by S1
The coordinates are fixed. In this example, the coordinates (X,
Y) 10 coordinates (X ₁ , Y ₁ ), (X ₂ , Y ₂ ), ...
At (X ₁₀ , Y ₁₀ ), the sample value of the detection signal V is V
Defects of ₁ , V ₂ , ..., V ₁₀ are distributed. That is, since the value of the detection signal V at coordinates other than those is less than or equal to a predetermined threshold value determined according to the detection sensitivity, it is considered that there is no defect.

In this case, the coordinates (X _i , Y _i ) (i = 1 to 1)
0) and the sampled value V _i of the detection signal V at this coordinate are collectively represented by data (X _i , Y _i , V _i ). Then, as shown in FIG. 3A, the first memory 23 of FIG. 1 stores data (X ₁ , Y ₁ , V ₁ ) at the first address and data (X ₂ , Y ₁ ) at the second address. ₂ and V ₂ ) are stored, and data (X ₃ , Y ₃ and V ₃ ) to (X ₁₀ , Y ₁₀ and V ₁₀ ) are sequentially stored in the following addresses. Next, after scanning the entire surface of the reticle 1 with the light beam, the stored data is read out from the first memory 23. This time, the sampled values of the detection signal are rearranged in the descending order of sample values V _{1 to} V ₁₀ , Figure 1 in descending order
It is stored in the second memory 24 at a series of addresses.

For example, if the sampled values V _{1 to} V ₁₀ of the detection signal are arranged in descending order, and V ₄ > V ₇ > V ₂ > V ₉ > ..., The second memory 24 of FIG. As shown in b), the data (X ₄ , Y ₄ , V ₄ ) is stored in the first address, the data (X ₇ , Y ₇ , V ₇ ) is stored in the second address, and the following addresses are stored in the following addresses. Data (X ₂ , Y ₂ , V ₂ ), ...
Will be stored. If the sampled values V _i are the same, for example, the sampled values may be stored in the second memory 24 in the order of sampling. As a result, the ranked data is stored in the second memory 24, and the ranking is in the descending order of the sample value V _i of the detection signal V.

As described above, the larger the detection signal V, the larger the defect size. Therefore, the data in the second memory 24 are arranged in descending order of defect size. As shown in FIG. 7 of the conventional example, this ranking is not accurate after dividing into ranks A, B, and C. This is because even if there is a signal difference, they may be divided into the same rank. Therefore, in this embodiment, the sample value itself of the detection signal V is used.

After that, the main control system 22 reads the data (X ₄ , Y ₄ , V ₄ ) of the first (that is, the largest size) defect from the first address of the second memory 24, and its coordinates (X ₄ ,
The driving units 3 and 26 are operated to move the reticle 1 and the observation unit 15 so that the center of the observation visual field of the observation unit 15 in FIG. 1 is set to (Y ₄ ). As a result, in FIG. 2, the center of the observation visual field 15a of the observation unit 15 is set to the point of the coordinates (X ₄ , Y ₄ ) on the reticle 1. Then, the operator visually observes the defect at the coordinates (X ₄ , Y ₄ ) by visually observing through the observing unit 15 or by observing an image with a two-dimensional image pickup device (CCD or the like) and a TV monitor attached to the observing unit 15. .

As a result, when it is determined that the defect has an acceptable size, the second address (2
The data (X ₇ , Y ₇ , V ₇ ) of the defect having the second largest size is read out and similarly observed by the observation section 15. The reason for observing in this way is that the magnitude of the detection signal V and the magnitude of the defect observed by the observation unit 15 may not always match. However, if the largest defect observed first is an acceptable size, it is often considered that the second and smaller defects are also acceptable, so the largest defect is acceptable. When it is determined that the size has been reached, the reticle 1 can be treated as a non-defective product and the inspection process can be terminated without further observation. In this case, in FIG. 1, the operator inputs an inspection end command to the main control system 22 from the keyboard 28, so that the reticle 1 is carried out by a reticle loader system (not shown), and the reticle to be inspected next is placed on the stage 2. Set.

However, in practice, not only the size of the defect but also the adhesion position of the foreign matter as the defect is a criterion for the quality judgment. For example, in the case of a reticle, the foreign matter in the light transmitting portion poses a problem, but the foreign matter adhering to the light-shielding portion made of the chromium film is not transferred onto the wafer, which causes no problem. Therefore, if the defect of the first rank is large and is located in the light-shielding portion of the reticle 1, there is no problem with the defect, but the defect of the second rank is the defect of the first rank in the observation result. Even if it is smaller, if the second-order defect is located in the light transmitting portion of the reticle 1, the reticle 1 may be a defective product. Thus, as a result of the observation of the second-order defect, when the position is the light transmission part of the reticle 1, the data of the third-order defect is read from the second memory 24 and observed by the observation part 15. Needless to say, the reticle 1 can be judged as a defective product and the inspection process can be ended.

Obviously, when observing a defect such as a foreign matter of the first rank and the position is the light transmitting portion of the reticle 1, the inspection process is immediately continued without continuing the remaining inspection process for the reticle 1. You just have to finish. Further, even if the defect is located in the light transmitting portion of the reticle 1, if the defect has a size that causes no problem even when transferred onto the wafer, the defect is determined to be no problem.

As shown in FIG. 1, the first memory 23 and the second memory 24 are not provided, and the magnitude comparison is performed in the order in which the sample value V _i of the detection signal V is obtained, and the larger sample value V _i is determined each time. You may use the method of storing in the storage part of. Also, without rearranging the detected data,
A method may be used in which only the data indicating the order is attached to the data arranged in the order of detection in the descending order of defects. Either way,
There is no substitute for observing according to the assigned (or sorted) order.

Next, a second embodiment of the present invention will be described with reference to FIGS. Also in this embodiment, the defect inspection apparatus of FIG. 1 is used to inspect the reticle for defects. Figure 4 (a)
Represents the reticle 1 to be inspected in this embodiment, and the surface size of the reticle 1 is 100 mm × 100 mm. In FIG. 4A, by dividing the surface of the reticle 1 in the X direction and the Y direction at predetermined pitches, respectively,
The surface of the reticle 1 is a cell E which is a large number of small rectangular areas.
, (1), E (1,2), ..., E (2,1), ... And defect data is summed up for each cell.

Originally, in the case of a reticle, the number of defects such as foreign matters was small, and one defect was found in a 10 mm square on average.
Depending on whether or not there are individual pieces, there are cases where a plurality of foreign substances are adjacent to each other, or the adhesion state (on the light-shielding portion or the light-transmitting portion) is different as described above. In order to observe adjacent foreign matter with a high resolution and to judge the adhering state by observing the adhering state, a reasonably high observation magnification is required. Therefore, as shown in FIG. Cell E (i, j) larger than the size of the observation field 15a of the observation unit 15 of
The cell size must be determined so that (i = 1, 2, ...; j = 1, 2, ...) Is small. Therefore,
The cell size is preferably 0.5 mm square or 0.1 mm square.

When a plurality of defects are detected in the cell E (i, j) of 0.5 mm square or 0.1 mm square on the reticle 1 when detecting defects by light beam scanning in this embodiment. Handles the larger sample value of the detection signal V as the sample value representing the cell. That is, for example, in FIG.
As shown in (b), it is _assumed that the cell E (i, j) has three defects with sample values of the detection signal V being V _ij1 , V _ij2 , and V _ij3 . The maximum value is V _ij1
, Cell E as shown in FIG.
V _ij1 is assigned to (i, j) as a sample value of the detection signal indicating a defect. Then, in the first memory 23 of FIG. 1, the cell center coordinates (X
_ij , Y _ij ) and the sample value V _ij of the maximum defect detection signal in the cell are stored as data (X _ij , Y _ij , V _ij ). At the same time, the defects in the cells are classified into, for example, A, B, and C according to the sample value, the number of defects of each rank is integrated for each cell, and the integrated result is stored in the third memory 25. .

It should be noted that this method is effective for an apparatus for detecting a foreign substance by scanning the reticle with a light beam even without the foreign substance observation unit 15. That is, when scanning the light beam, the size of the light beam on the reticle is larger than the size of the foreign matter, and on the reticle, the brightness of the light beam is a Gaussian distribution (the brightness is higher at the center of the beam and lower at the periphery). distribution)
Therefore, the light beam scanning is performed so that the tails of the Gaussian distribution always overlap. Usually, when the size of the Gaussian-distributed beam is W (the width at which the brightness is 13.5% of the peak value), the second scan is shifted by about W / 4 in the distance on the reticle from the first scan. To do so. Therefore, a scattering signal is obtained for one foreign matter over several times of beam scanning, and the amount of these signals is different when the center of the beam hits the foreign matter and when the tail of the beam hits. Therefore, if the signal value and the X and Y coordinate values are captured every time a scattered signal is obtained by scanning the light beam on the entire surface of the reticle, the data is captured with the X and Y coordinate values that are almost the same even though the foreign matter is the same. There will be a problem that it will happen.

On the other hand, if the unit cell (data unit) for capturing is set to 0.1 mm square and the maximum signal thereof is set to the representative value of the cell, such a problem is solved. This operation may be performed by software, but is preferably performed by hardware. That is, 0.1 mm square is taken as a unit of capture, and if a defect signal is detected within this 0.1 mm square, the signal value is saved, and when a larger defect signal is received, the previous signal is erased and updated. Go. Finally, only the signal of the largest defect in the cell (0.1 mm square) remains, and this signal and cell coordinates are used as data (X _ij , Y _ij , V _ij ).
Save as. X _ij and Y _ij are the center coordinates of the cell.

In order to display the data on a display such as a CRT, this data is transferred. There are two methods for transferring the data. One is a method of transferring all data of cells divided on the reticle. For example, 100 ×
When transferring a 100 mm area in units of 0.1 mm square cells, if no foreign matter is detected, zero data and detection are performed according to a determined order (for example, 100 × 100 mm is determined as the first address from the upper left to address 100000). Then, the sample value V _ij is sent. When displaying, in this order, when there is zero data, it is displayed that there is no foreign matter, and when the sample value V
_The detection display and the like classified according to _ij are performed. But,
This method is not preferable because the amount of data to be transferred is large.

More preferably, only the detected data (X _ij , Y _ij , V _ij ) is sent by the following second method, and detection display is performed on the cell corresponding to the coordinates X _ij , Y _ij . This method is very effective when the amount of foreign matter is small, the amount of transfer data is small (usually because the amount of foreign matter is small on the reticle), and the transfer time is not noticeable at all. In this way, 100
When the light beam inspection on the entire surface of the reticle 1 of mm × 100 mm is completed, the main control system 22 of FIG. 1 displays a defect inspection defect on the CRT display 14.

FIG. 5A shows a display example on the CRT display 14. In FIG. 5A, the window 19 in the large window 18 corresponds to the area of 100 mm × 100 mm of the reticle 1 of FIG. 4A. To do. The window 19 is divided by a predetermined pitch in the X1 direction corresponding to the X direction in FIG. 4A and the Y1 direction corresponding to the Y direction, and a large number of display cells P (1,1), P are displayed.
(1,2), ..., P (2,1) ,.
Then, in the display cells corresponding to the defective cells in FIG. 4A in each display cell, A, B, and B are displayed according to the magnitude of the sample value of the detection signal of the maximum internal defect as a two-dimensional map.
Defects ranked into C are displayed.

However, even if the cell E (i, j) in FIG. 4A is, for example, 0.1 mm square, the display cell P in FIG.
(I, j) corresponds to a 5 mm square area on the reticle 1,
The window 19 is composed of a 20 × 20 matrix. Therefore, in the display cell P (i, j), the sample values of the detection signal of the largest defect in the corresponding area of 5 mm square on the reticle 1 are ranked and displayed. Therefore, since the exact number of defects in a unit of 0.1 mm square is buried and is not known, the number of defects in each rank A, B, C is displayed in the display cell unit ( That is, 5 mm square unit on reticle 1)
Are displayed separately on the reticle 1 in units of 0.1 mm square.

Of course, in addition to the display using letters such as alphabets, the size of the defect is displayed as blue when it is determined as a small size defect (for example, foreign matter), green when the size is medium, and red when the size is large. As described above, a defect may be represented for each display cell by the color-coded display. Next, reticle 1
When the operator wants to observe the defect in each cell above using the observing section 15 of FIG. 1, the operator inputs a command to start observation to the main control system 22 via the keyboard 28. In response to this, the main control system 22 controls the cell E (i, j) of 0.1 mm square.
The cells are moved to the observation visual field 15a of the observing unit 15 based on the defect data for each of the cells and in the order of the largest defects in the cell (the order of the largest defect size). The position of the entire defect currently being observed is indicated by the arrow cursor 21 in FIG. Further, near the large window 18 on the display screen, as shown by a window 22 in FIG. 5B, a defect display is also performed in units of 0.1 mm square cells on the reticle 1. Also in the cell-based window 22 on the reticle 1, the cursor 23 of the arrow indicates the cell currently being observed.

When further observing, it is preferable to know how much (up to how many) the observation has been completed. For example, the observation end cell is deleted from the window 19 in FIG. When observing in the descending order, for example, as shown in FIG. 5 (b), even if the C rank display is 5 mm square display, if the display is 0.1 mm square, two defects of C rank and one B rank are found in 5 mm square. If yes, C rank 2
After observing the cells, the remaining cells of the C rank of the window 19 are changed to the B rank display because the remaining one is the B rank, and the display cell having the defect still does not disappear from the window 19, but the C rank display from the window 22. The cell disappears. When all the display cells having a defect disappear from the window 19, it means that the observation is completed. It should be noted that even if the cell display is not erased, the color of the cell that has been observed may be changed or blinked to leave it in an identifiable form. Alternatively, instead of erasing the cell (or at the same time as erasing the cell), the number of A, B, and C ranks of 5 mm square and 0.1 mm square of the window 20 is reduced in the order in which they have been observed, and the remaining number remains. You may be able to know whether or not.

Instead of the cursors 21 and 23, the observation position may be displayed in different colors, or the observation position may be indicated by blinking the display. Alternatively, FIG.
When the defects are displayed or output in a table format rather than the display formats as in (a) and (b), the observation position is shown near the table so that the defect under observation can be identified. You may display or output a figure.

Further, in this embodiment, as shown in FIG.
A cell E (i, j) on the reticle 1 may include a plurality of defects of different sizes. Even in such a case, the size of the cell is smaller than the observation visual field 15a of the observation unit 15, and the entire area in the cell is one observation visual field 1a.
Since it can be observed with 5a, it is possible to observe a plurality of foreign substances at the same time.

Obviously, the observer cannot know how many foreign matters are detected in the cell until he / she actually observes them, but at a relatively high observation magnification as described above, a plurality of foreign matters are detected. Is extremely easy to find. At this time,
The movement of the reticle 1 with respect to the observation visual field 15a is performed by each cell E
It is based on the center coordinates of (i, j). That is, since the X coordinate and the Y coordinate are determined in units of cells,
The drive units 3 and 26 of FIG. 1 are controlled so that the center coordinates of the cell E (i, j) match the center of the observation visual field 15a.

A series of defect inspection operations in this embodiment are summarized in the following steps (1) to (3). Defect detection is performed by light beam scanning. The largest defect is determined for each 0.1 mm square (or 0.5 mm square) cell. The defective data is stored in the first memory 23 for each cell. Ranking by defect size (ranking in descending order)
I do. A map is displayed on the CRT display 14. When the observation is designated, the center of the observation visual field 15a is moved to the center of the corresponding cell according to the order in the step. The defect is judged to be good or bad by visual observation or image observation. If it is defective, the subsequent processing can be stopped.

In the above steps, it is preferable that the observation designation in each step be selectable between automatic and manual. For example, when automatic is designated, observation is performed sequentially according to the ranking, and when manual is designated, a method of particularly preferentially observing a cell using a mouse or a cursor is used. After the designation, the reticle 1 and the observation unit 15 are moved so that the center coordinates of the cell coincide with the center of the observation visual field 15a. This manual designation method is effective when observing other adjacent defects. In that sense, the raster scan may be sequentially performed regardless of the size of the defect.

Further, the reticle 1 has a transfer area (pattern area), and there is no problem even if there is a defect such as a foreign substance in an area other than the transfer area. However, the transfer area and the defect inspection area do not always match. For example, in FIG.
In (a), the transfer area has a size of 80 mm × 90 mm, but since it is desired to inspect as wide an area as possible at the time of inspection, the inspection area is designated as an area of 100 mm × 100 mm. However, what is important for defect management is the presence / absence of defects in the transfer area of the reticle 1. Therefore, in some cases, the transfer area may be preferentially observed by the observation unit 15 in FIG. In this case, first, the defect in the transfer area is selected, and the observation section 15 observes the defects in the order of the larger size, and then the defects detected outside the transfer area are observed in the observation section 15 in the order of the larger size. The order of observation may be prioritized.

As a modification of the above embodiment, FIG.
The cell E (i, j) in (a) is made relatively large, such as 5 mm square or 1 mm square, and at the time of map display, only the maximum sample value of the detection signal in each cell is represented as the representative value of that cell. As an example, it may be displayed as in the window 19 of FIG. In this case, the actual data is
Defect data is stored for each central coordinate of a region (0.1 mm square or the like) finer than the size of the cell on the reticle 1 corresponding to the display cell in FIG. 5A, and the fine data is specified by the operator. Make it possible to observe defects in area units.

As a specific application example thereof, there is a defect inspection by an image processing method. In this image processing method, an image pickup apparatus having a predetermined observation field of view (hereinafter, referred to as “image field of view”) is used to sequentially capture images of the entire surface of the reticle 1 as a two-dimensional image, and a defect such as a foreign substance is determined by image processing. To do. Each image in the image field corresponds to each cell on the reticle 1 in FIG. The position and size of the defect in each image field of view are obtained in advance, and when observing the defect in more detail, the position of the defect in each image field of view is set at an observation magnification higher than the observation magnification of the image field of view. The observation visual field of the observation unit 15 is moved.

In addition, image data when a defect is detected is stored in an image storage unit such as a frame memory, the entire surface of the reticle 1 is inspected, and then the image data is read from the image storage unit and displayed on a TV monitor or the like. The image of the reticle 1 may be displayed. With this method, when it is desired to observe a defect in the image field of view at a higher observation magnification, the position of the defect in the image field of view is designated by a cursor or the like, and the position is moved into the observation field of the observation unit 15 to obtain high magnification. Observe. When the image data including the defect is read, T
A mark (+ mark or arrow) indicating the presence of a defect on the image display unit such as a V monitor, or a display that can be distinguished from others by color may be attached to the defect position. At this time, similarly to the above-described embodiment, it is preferable to read the image fields of view including the defects from the storage unit in the descending order of the defect size and display them.

There are various image processing methods other than the above-mentioned image processing method. For example, the simplest method is to perform a defect inspection on a reticle (glass blanks) having no pattern by dark-field illumination. In this case, since only the defect such as a foreign substance is illuminated by the dark field illumination, only the defect can be easily extracted by binarizing the dark field image.

Next, a third embodiment of the present invention will be described. Also in this embodiment, the defect inspection apparatus of FIG. 1 is used. The defect inspection method of this embodiment is the most efficient method for shortening the time of the inspection process. That is, in this embodiment, a quality determination standard is set in advance before the inspection. For example, the number of detected defects is limited by rank, and if there is one C-rank defect and two or more B-rank defects, a pass / fail judgment criterion is set so that the reticle is defective. .
It is the main control system 22 in FIG. 1 that determines whether or not this quality determination criterion is satisfied.

Then, in FIG. 1, when the inspection of the reticle 1 by the light beam scanning is started and the light beam L2 is scanned in the X direction as much as possible, or when the reticle 1 is sent in the Y direction as little as possible, the quality is good or bad. If the criterion is exceeded (not satisfied), the inspection by the light beam scanning is stopped at that point and the observation by the observation unit 15 is not performed. However, since it may be better to observe with the observing unit 15 as in the case of analyzing the cause of the defect for the future, it is desirable to be able to select the above-mentioned automatic determination or manual determination. Specifically, in the automatic judgment mode, the inspection is ended when the quality criterion is exceeded, and in the manual judgment mode, the inspection is ended in the same manner as in the automatic judgment mode, or the inspection up to that point is detected. It is advisable to move the reticle 1 to the observation visual field of the observation unit 15 so that the observed defects can be observed.

The defect inspection apparatus of the above-described embodiments often includes a case for accommodating the reticle and a transfer section for transferring the reticle from the reticle case to the foreign matter inspection section.
In particular, the reticle case is often configured so that a plurality of reticle cases can be set, and for example, 5 sets (5 reticle cases) can be set. At this time, when the reticle is housed in all of the five cases and five sheets are continuously inspected, the automatic determination mode described above is selected. 5 after inspection
A good reticle is “good” by comparing it with the pass / fail judgment standard
Label (processing on the data), label the defective reticle as “defective” and return all to the reticle case. When the operator wants to observe the inspection data (label) and observe the data of the reticle labeled as "defective" after the continuous inspection of five sheets to observe the foreign matter,
Based on the display of, the reticle may be directly conveyed from the reticle case to the foreign matter observing section so that the foreign matter can be observed by the method described above. Of course, you may want to observe foreign matter on the reticle labeled as “good”. Therefore, it is not limited to continuous processing, but after inspecting once, the reticle returned to the case can be transported again from the case and arbitrarily selected for foreign matter observation. I hope I can do it. This is also very advantageous. This is because it is natural that no foreign matter is present in the reticle at a level where zero foreign matter is attached, and any observation of zero foreign matter is unnecessary. This is because it would be very effective if the observations are made collectively only when there is a foreign matter, since the observations can be made at a time convenient for the operator.

Although the target of the defect inspection is the reticle in the above-mentioned embodiment, the present invention can be similarly applied even when the target of the defect inspection is a wafer or a printed circuit board. As described above, the present invention is not limited to the above-described embodiments, and various configurations can be taken without departing from the gist of the present invention.

[0062]

According to the first defect inspection method of the present invention,
Since the detected defects are observed in descending order and the defect inspection is finished when there is a defective portion, there is an advantage that the inspection time can be shortened as a whole. Further, according to the second defect inspection method, the largest defect is detected for each cell, the cells are observed in the unit of the largest largest defect, and the defect inspection is completed when there is a defective portion. Therefore, there is an advantage that the inspection time can be shortened as a whole.

According to the third defect inspection method, the defect inspection is completed when the number of defects exceeds the allowable number or the size of the defects exceeds the allowable value. There is an advantage that it can be shortened as a whole. Further, according to the first to third defect inspection apparatuses of the present invention, the above-mentioned first
~ The third defect inspection method can be implemented.

[Brief description of drawings]

FIG. 1 is a configuration diagram including a partial perspective view showing a defect inspection apparatus used in an embodiment of the present invention.

FIG. 2 is a plan view showing an example of defect distribution of the reticle 1 according to the first embodiment of the present invention.

3A is a diagram showing defect data stored in a first memory 23 in the first embodiment, and FIG. 3B is a diagram showing defect data stored in a second memory 24. FIG.

FIG. 4A is a plan view showing an example of cell division of the reticle 1 in the second embodiment, and FIG. 4B is an enlarged plan view showing an example of defect distribution in the cell E (i, j). (C) is cell E
It is a figure which shows the detection signal of the defect assigned to (i, j).

5A is a diagram showing a part of a defect display map in the second embodiment, and FIG. 5B is a diagram showing another part of the display map.

FIG. 6 is a configuration diagram including a partial perspective view showing an example of a conventional defect inspection apparatus.

FIG. 7 is a diagram showing an example of a conventional defect display map.

[Explanation of symbols]

 1 Reticle 2 Stage 4 Length Measuring Device L1, L2 Light Beam 6 Galvano Scanner 8 Scanning Lens 10 Scanning Line 12 Photosensitive Lens 13 Photoelectric Detector 15 Observing Section 21 Defect Discriminating Circuit 22 Main Control System 23 First Memory 24 Second Memory 27 Stage Control unit

Claims (6)

[Claims] 1. A test surface of a test object is irradiated with inspection light, and a defect on the test surface is detected based on scattered light of the inspection light from the test surface, In a defect inspection method for observing the detected defect at a predetermined magnification, the size of the defect on the surface to be inspected is determined based on a photoelectric conversion signal of scattered light of the inspection light from the surface to be inspected. Then, when there are a plurality of the detected defects, the detected defects are observed at the predetermined magnification in the descending order of the defects based on the determination result, and when there is a defective portion as a result of the observation, the defect is detected. A defect inspection method characterized by terminating a defect inspection of an object to be inspected. 2. The inspection surface of the inspection object is irradiated with inspection light, and a defect on the inspection surface is detected based on scattered light of the inspection light from the inspection surface, In the defect inspection method of observing the detected defect at a predetermined magnification, the surface to be inspected is divided into a plurality of cells having a predetermined size, and the scattered light of the inspection light from the surface to be inspected is divided. The size of the largest defect in each of the plurality of cells is determined based on the photoelectric conversion signal, and the cells including the defect in the plurality of cells are arranged in the descending order of the largest defects in the determination result. A defect inspection method characterized by observing at a predetermined magnification and terminating the defect inspection of the inspection object when there is a defective portion as a result of the observation. 3. A method of irradiating a test surface of a test object with inspection light, and detecting a defect on the inspection surface based on scattered light of the inspection light from the inspection surface. The size of a defect on the surface to be inspected is determined based on a photoelectric conversion signal of scattered light of the inspection light from the surface to be inspected, and the number of the detected defects exceeds a predetermined allowable number. And that at least one of the defects having the size of the determined defect exceeding a predetermined allowable value exists, the defect inspection of the inspection object is terminated. Characteristic defect inspection method. 4. Optical scanning means for scanning inspection light on a surface to be inspected of an object to be inspected, and scanning position measuring means for measuring a scanning position of the inspection light on the surface to be inspected, Light receiving means for photoelectrically converting scattered light of the inspection light from the surface to be inspected,
An observation means for observing the surface to be inspected at a predetermined magnification, and detecting a defect and a position of the defect on the surface to be inspected based on output signals of the light receiving means and the scanning position measuring means, In a defect inspection apparatus for observing the detected defect by the observing means, a defect determining means for determining a size of the defect on the surface to be inspected based on an output signal of the light receiving means, and a determining operation of the defect determining means. Storage means for storing the size of the defect and the position on the surface to be inspected for each defect on the surface to be inspected, based on the result and the output signal of the scanning position measuring means, and stored in the storage means. Based on the size of the defect and the position on the surface to be inspected, control means for controlling the positional relationship between the object to be inspected and the observation means so that the defects can be observed in descending order. Characteristic defect inspection equipment. 5. Optical scanning means for scanning inspection light on a surface to be inspected of an inspection object, and scanning position measuring means for measuring a scanning position of the inspection light on the inspection surface, Light receiving means for photoelectrically converting scattered light of the inspection light from the surface to be inspected,
An observation means for observing the surface to be inspected at a predetermined magnification, and detecting a defect and a position of the defect on the surface to be inspected based on output signals of the light receiving means and the scanning position measuring means, In the defect inspection device for observing the detected defect by the observing means, a defect determining means for determining a size of a defect on the surface to be inspected based on an output signal of the light receiving means, and the surface to be inspected are predetermined. A plurality of cells having the same size, and based on the determination result of the defect determining means and the output signal of the scanning position measuring means, for each cell having a defect in the plurality of cells, the largest defect in the cell And a storage unit that stores the size of the cell and the position of the cell on the test surface, and the defect based on the size of the defect stored in the storage unit and the position of the cell on the test surface. The cell with the largest of the largest defects in the cell Defect inspection apparatus characterized by and a control unit operable to control the positional relationship between the test object and the observation means so as to be observed. 6. An optical scanning means for scanning inspection light on a surface to be inspected of an inspection object, and a scanning position measuring means for measuring a scanning position of the inspection light on the inspection surface, Light receiving means for photoelectrically converting scattered light of the inspection light from the surface to be inspected,
In the defect inspection apparatus for detecting a defect on the surface to be inspected and the position of the defect based on the output signals of the light receiving means and the scanning position measuring means, the defect inspection apparatus based on the output signal of the light receiving means. Defect discriminating means for discriminating the size of the defect on the inspection surface, and the defect size for each defect on the inspection surface based on the discrimination result of the defect discriminating means and the output signal of the scanning position measuring means. And storage means for storing the position on the surface to be inspected, first comparing means for comparing the number of defects stored in the storage means with a predetermined allowable number, and of the defects stored in the storage means And a second comparing means for comparing the maximum defect size in the inside with a predetermined allowable value.