JP2000081324A - Method and equipment for inspecting defect - Google Patents

Abstract

PROBLEM TO BE SOLVED: To set a defect candidate easily in an observation field of view at the time of observing a temporarily detected defect candidate in detail by storing information concerning to the three-dimensional coordinates of a detected defect and locating the detected defect in the observation field of view based on the stored information. SOLUTION: The image of a wafer 1 in the inspection field of view is focused on the detection face of a defect detecting sensor 12 through a detection optical system 24. Output from the sensor 12 is inputted to a defect detection processing system 13 and a pattern defect is detected utilizing iteration of pattern on the wafer 1. X, Y, Z coordinates and feature amount of a defect are stored in the defect memory at a work station 14. An observation focal point mechanism control system 18 controls an observation focal point drive mechanism 21 to shift the focal point of an observation optical system 22 based on the Z-axis height stored at the time of detecting the defect. An observation image picked up by means of an observation image sensor 19 is used for focusing through the control system 18 or display on a defect confirmation display 16.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for detecting defects on a sample related to a semiconductor, and more particularly to a method suitable for improving workability when calling out a detected defect on an observation device.

[0002]

2. Description of the Related Art Conventionally, in the field of semiconductor development and manufacturing, it has been widely practiced to detect a defect occurring on a product during or after manufacturing, to analyze and analyze the defect. In particular, feeding back these results to the manufacturing and design processes is an indispensable process for improving product quality and manufacturing yield.

[0003] For this reason, it is extremely routine to detect a defect on a product, apply the detected defect to an observation device (review device), and call and observe a coordinate position at which the defect is detected (hereinafter, review). Business.

In this case, since the sample to be inspected is a two-dimensional planar substrate, the coordinates of the detected position are stored as two-dimensional coordinates (in many cases, X and Y coordinates) and used for review.

Further, many products in the field of semiconductors are manufactured in pursuit of the limit of miniaturization, and the boundary between a normal portion and a defective portion is delicate particularly in the case of visual inspection. For this reason, the result detected by the automatic visual inspection device includes a considerable amount of false information (a normal portion determined as a defect by the inspection device). Therefore, it is an indispensable procedure to review the detection results and confirm the status of the correctly detected defect.

In some cases, the calling of the defect position is performed by indicating not only the detected coordinates but also the characteristics of the pattern on the sample. This is a semiconductor wafer for analysis (hereinafter,
This method is used when the continuity of the coordinate system cannot be maintained, for example, when a part of the wafer is broken and applied to an analyzer, or a position corresponding to a defect detected on the wafer is called on a photomask.

In many cases, this is done by designating the number of repetitive circuit patterns, called cells, corresponding to one storage unit (one bit) in the semiconductor memory from the end of the LSI chip. It is quite difficult to accurately count the number of pieces, but an example in which this is automated on an ion beam analyzer is disclosed in JP-A-5-62638.

In order to realize the above functions, a function of observing a detection result is realized on many automatic appearance defect inspection apparatuses (hereinafter, inspection apparatuses). For example, JP-A-7-7
No. 2093 discloses that defects are detected by optical means for a photomask or a reticle used in semiconductor manufacturing.
One embodiment of a device having a function of observing and confirming the result with the same device and the same optical system is shown. Japanese Patent Application Laid-Open No. Sho 60-218845 discloses an apparatus having a function of detecting a defect by optical means for a semiconductor LSI and observing and confirming the result by another electron optical system in the same apparatus. An example is shown. Further, Japanese Patent Application Laid-Open No. 58-33154 discloses that a defect is detected by optical means for a semiconductor LSI, and the result is transferred to the same device or the detected information to another device and observed by another electron optical system. One embodiment of the device having the function of confirming the status is shown.

[0009]

Even when the inspection and the review are performed by the same optical system in the same apparatus, there is a problem in that the position in the height direction is adjusted in addition to the calling of the detection position coordinates on the substrate plane. There is.

This is because an inspection apparatus having a high-resolution detection system and observation system has a narrow depth of focus due to the principle of the configuration of the optical system, so that it is necessary to precisely focus. For this purpose, an automatic focus function is generally used from the viewpoint of workability. However, due to differences in lighting conditions between detection and observation,
It is difficult to automatically focus on the same height during detection and observation. In the worst case, the autofocus mechanism may not work during observation.

If the observation optical system cannot be used due to the structure of the detection optical system, the observation optical system must be provided at another position. In this case, an expensive automatic focusing mechanism is provided for both optical systems. In this situation, it is necessary to keep the camera at the same height during detection and observation, as described above.

Further, the detection is performed by an optical inspection device having a feature of high speed, the coordinates of the defect in the plane of the substrate are transmitted as information, and the review is performed by an electronic scanning microscope having a feature of high resolution. In the case where the detection and observation are performed by different devices, for example, when the coordinates of the defect in the substrate plane are conveyed as information, the possibility of focusing on the same height during detection and observation is further reduced. I do.

In particular, on a substrate on which a translucent multilayer thin film is formed, it is not clear at the time of inspection which inspection device focuses on which layer in the multilayer thin film and which layer detected a defect at the time of observation. The operator had to consider the possibilities of all layers, and as a result, became unable to rely on the autofocus mechanism and impaired workability.

As described above, since detection is often performed in a wide field of view, when there are regions of a plurality of heights in the detection field of view, the focus is placed on the region of the structure at any height during observation. There was a problem that I couldn't tell if it was good.

In order to cope with the above problem, there is a need for a device that reproduces the defect detection position not only in the plane of the substrate but also in the height direction.

When the inspection and the review are performed by the same optical system in the same apparatus, the coordinates of the detection position and the coordinates of the observation position in the plane of the substrate can be easily matched. However, some inspection apparatuses perform detection with a remarkably wide field of view in order to pursue high speed, and others output rough detection coordinates in the apparatus configuration. In this case, when observing the detection result, a worker must search for a detection defect from a wide range for a human, and the defect is very delicate with the miniaturization of semiconductor elements. It has become. This significantly impairs the workability at the time of observation. Therefore, there is a need for a support function capable of easily specifying a defect position from the visual field at the time of observation.

As described above, when the position of a detected defect is designated by designating the number of repetitive circuit patterns called cells on a semiconductor memory from the end of an LSI chip, it is difficult to count the number of repetitive circuit patterns. Therefore, workability is significantly impaired. Although there is an automated approach to this problem, it is not the same as automation due to the nature of the work (many analyses / analyses), but rather, it is required to devise a function that supports the manual work of the operator.

Further, according to this instruction method, not only the number of cells but also the mat to which the designated cell belongs (cells are continuously arranged and are usually surrounded by an area called a peripheral circuit where a sense amplifier and a sub-word driver exist). Unit area)
It is necessary to specify the position in the chip. For this purpose, it is important to be able to see the entire chip. However, since the observation optical system in which the resolution is regarded as important generally adopts a microscope configuration, there is a limit to widening the field of view. For this reason, to be able to easily specify the mat part,
There is a need for a device that allows the entire chip to be seen.

An object of the present invention is to provide a defect inspection method and an apparatus thereof which can be easily set in an observation visual field for observing a defect candidate in detail when the detected defect candidate is observed in more detail. .

[0020]

In order to achieve the above object, according to the present invention, in detecting a defect, in addition to recording the coordinates of the detected defect in the plane of the substrate, the position of the detected height in the height direction is also simultaneously recorded. By recording, the defect detection position is reproduced not only in the plane of the substrate but also in the height direction.

That is, in order to achieve the above object, according to the present invention, in a method for inspecting a defect of a sample, an image of the sample is obtained by imaging the sample, and the defect of the sample is detected from the image of the sample. Information on the three-dimensional coordinates of the detected defect is stored, and the detected defect is located in the observation field of view based on the stored information on the three-dimensional coordinates of the defect to observe the defect.

In order to achieve the above object, according to the present invention, a defect inspection apparatus for inspecting a defect of a sample is provided with a defect detection means for detecting a defect of the sample from an image of the sample obtained by imaging the sample. Storage means for storing information relating to the three-dimensional coordinates of the defect of the sample detected by the defect detection means; output means for outputting information relating to the three-dimensional coordinates of the defect stored in the storage means; and storage means for storing the information. And an observation unit for moving the defect of the sample into the observation field of view based on the information on the three-dimensional coordinates of the defect to obtain an enlarged image of the defect.

In order to achieve the above object, according to the present invention, in a defect inspection method for inspecting a defect at a desired position on a sample, the sample is sequentially imaged using an optical system having a first magnification, and Are obtained, a plurality of images of the sample are recorded, and the recorded plurality of images are combined to obtain an image of an area larger than the visual field on the sample by the optical system of the first magnification. A desired location on the sample is determined from the image of the region, the desired location is imaged using the optical system of the second magnification, and an image of the desired location of the sample captured using the optical system of the second magnification Output information about

In order to achieve the above object, according to the present invention, there is provided a defect inspection apparatus for inspecting a defect at a desired position on a sample, comprising:
Imaging means, storage means for storing a plurality of images of the sample imaged by the first imaging means, and combining the plurality of images stored in the storage means to obtain a plurality of images from the field of view of the first imaging means on the sample. A synthetic image creating means for creating an image of a large area, a setting means for finding a desired location on the sample from the image of the large area created by the synthetic image creating means, and an image of the desired location found by the setting means A second imaging means for obtaining an image of a desired location by using the second imaging means; and an output means for outputting information relating to the image of the desired location of the sample imaged by the second imaging means.

According to another aspect of the present invention, there is provided a defect inspection method for detecting a defect of a repetitive pattern on a sample on which a repetitive pattern is formed and observing the detected defect. Image of the sample is obtained by imaging the sample having b), b) detecting the defect candidate of the repetitive pattern from the obtained image, c) displaying the image of the detected defect candidate on a screen, and d) displaying the defect candidate on the screen And (e) enlarge the designated portion on the screen.

In order to achieve the above object, according to the present invention, there is provided a defect inspection method for detecting a defect on a substrate sample and observing the detected result. Is positioned in the field of view of the observation system,
b) fetching images at a plurality of heights into an image processing system to obtain a fetched image; c) obtaining a height at which the brightness information is maximized from the fetched image; Displayed g) A defect candidate point of the captured image displayed on the screen is designated and marked with a pointing device, and h) The marked defect candidate point is further enlarged and displayed.

In order to achieve the above object, according to the present invention, there is provided a defect inspection apparatus for detecting a defect of a repetitive pattern on a sample on which a repetitive pattern is formed and observing the detected defect. (B) a defect candidate detecting means for detecting a repetitive pattern defect candidate from the image obtained by the imaging means, and c) a defect candidate detecting means. Display means for displaying an image of the detected defect candidate on a screen;
d) enlarged display designating means for designating a portion to be enlargedly displayed from among the defect candidate images displayed on the screen of the display means, and e)
And an enlargement display means for enlarging and displaying a designated portion in the image of the defect candidate by the enlargement display designation means on a screen of the display means.

In order to achieve the above object, according to the present invention, in a method for inspecting a defect of a sample, the method comprises the steps of:
An image of the sample is obtained by using the imaging means, and a defect candidate of the sample is detected from the image of the sample to obtain positional information of the defect candidate in the sample plane. The position information of the surface is obtained, and the defect candidate detected based on the position information of the defect candidate in the sample plane and the position information of the sample surface in the normal direction is positioned in the observation field of view of the second imaging means. The second candidate picks up an image of a defect candidate in the observation field of view and displays an image of the defect candidate on the screen.

In order to achieve the above object, according to the present invention, in a defect inspection apparatus for detecting a defect candidate of a sample and observing the detected defect candidate in detail, an image of the sample is obtained by imaging the sample. A first imaging unit, a defect candidate detection unit that detects a defect candidate of the sample from an image of the sample obtained by imaging the first imaging unit, and obtains position information of the defect candidate in the sample plane; Height detection means for obtaining position information of the sample surface in the normal direction of the sample surface, and position information within the sample surface of the defect candidate obtained by the defect candidate detection means and the method of the sample surface obtained by the height detection means Second imaging means for positioning a defect candidate detected based on the position information in the line direction in the observation visual field, and imaging the defect candidate in the observation visual field with the second imaging means to display an image of the defect candidate on the screen. And a display means for displaying the above.

[0030]

FIG. 1 shows an embodiment of the present invention.

A sample 1 to be inspected on a substrate is placed on a sample holder 2. Here, a semiconductor wafer or a photomask is assumed as an example of the sample to be inspected, and is hereinafter referred to as a wafer. The sample holder 2 is positioned by a Z-axis stage 3, an X-axis stage 4, and a Y-axis stage 5 connected thereto. The X and Y axis stages are mainly used for scanning at the time of inspection (the entire surface of the wafer is inspected by performing X and Y scanning over the entire surface of the wafer with a detection optical system smaller than the wafer. In an inspection apparatus for r-θ scanning, this portion is used). The Z stage is mainly used for adjusting the plane to be inspected of the wafer to the focus of the detection optical system 24 and the observation optical system 22.

An encoder 6 for the Y-axis stage, an encoder 7 for the X-axis stage, and an encoder 8 for the Z-axis stage are attached to each stage.
You can read from. In the present invention,
The height direction is referred to as Z, the movement direction between the inspection position and the observation position in the movement within the plane is referred to as Y, and the direction perpendicular thereto is referred to as X.

The control of the Z-coordinate at the time of inspection has been described with a configuration in which the reflected light from the inspection light source 23 is read by the auto-focus sensor 10, but the control of the contrast of the wafer image or the projected stripe pattern is performed. Other methods may be used. The output of the sensor is processed by the automatic focus control system 9, and the position of the Z axis is adjusted.

In the case of a detection method which does not require automatic focusing at the time of inspection, this processing system functions not as automatic focus control but as a mechanism for measuring the height of the surface of the sample to be inspected at the time of detection.

In this embodiment, a method is described in which the entire inspection visual field is treated as one Z-axis height. However, some wafers have structures with significantly different heights (several μm) in the inspection visual field. In order to accurately record the height at the time of detection, it is desirable that the configuration be such that the Z height can be measured at a plurality of points in the visual field.

An image in the inspection field of the wafer 1 is formed on the detection surface of the defect detection sensor 12 by the detection optical system 24, and an output from the defect detection sensor 12 is input to the defect detection processing system 13. You. The defect detection processing system 13 detects pattern defects using the repetition of the pattern on the wafer 1. That is, by comparing images of patterns that should have the same shape on the wafer, a mismatch between the images is determined as a defect. If it is determined to be defective, X
The Y and Z defect coordinates and the feature amount (for example, size) of the defect are sent to the workstation 14 having the defect memory and stored.

The workstation 14 is connected with a pointing device 15 such as a mouse or a trackball, and is used for various setting functions described later. Further, a display 16 for detecting a detected defect and a display 17 for displaying / inputting an operation condition are connected. These displays may be used by dividing the screen of one display by windows or the like and sharing the display. On the other hand, the inspection stage moves to the observation position 25 during observation. This is because of the configuration of each optical system, the detection optical system 24 and the observation optical system 2
2 is a unit that is used when the same optical system cannot be used or can not be arranged at the same position. If the detection optical system 24 and the observation optical system 22 can be used together or can be arranged at the same position, such means is adopted. It is desirable to do. For more detailed observation, it is preferable to use another optical review device or a review device using a scanning electron microscope (hereinafter, SEM). In this case, however, it is interpreted that the observation position 25 is located on another device. It does not change the essence of the present invention.

The observation focus mechanism control system 18 controls the observation focus system drive mechanism 21 to move the focus position of the observation optical system 22 based on the Z-axis height stored at the time of detecting a defect. Although the type in which the lens of the observation optical system is moved is shown in the figure, when the sample stage has a Z-axis mechanism,
The axis stage may be directly controlled.

Epi-illumination is generally used for the observation illumination system 20, but dark-field illumination or oblique illumination by a laser may be used to make it easy to confirm a defect. Although the observation image may have a function of directly observing the image by branching or switching the optical path, the observation image sensor 19 (for example, T
V camera).

The captured observation image is used for focusing (which is an auxiliary in the present invention) by the observation focusing mechanism control system 18 and for display on the display 16 for detecting a detected defect. In addition, it is sent to the image processing system on the workstation 14.

FIG. 2 shows a detection state when a multilayer thin film is present on the sample substrate 26 to be inspected. In FIG. 3A, the upper thin film layer 28 of the sample to be inspected is located on the lower thin film layer 27 of the sample to be inspected, and a defect 29 such as a foreign substance is present on the surface thereof. On the other hand, in b) in the figure, the lower thin film layer 27 and the upper thin film layer 2
There is a defect 30 such as a foreign substance between the layers 8 and 8. Also, each thin film is often transparent at the wavelength of the detection light.

By the way, it is impossible to know whether the defect is detected by focusing on the surface layer or between the layers during the defect inspection by the conventional apparatus. Even if the defect is called to the correct observation position, the normal part is not noticed. It must be treated as an erroneous detection of, or observed by changing the Z height to search for defects such as foreign matter. According to the present invention, Z
Since the height is also recorded, the position of the defect such as a foreign substance detected at the time of observation is positioned including the Z height, so that erroneous determination and search in the height direction to prevent erroneous determination are omitted, and review efficiency is improved. Is significantly improved.

There is another problem in observing foreign matter / defects in the multilayer film. This occurs when observation with an electron microscope is performed. In an electron microscope, since the shape data of the outermost layer can be obtained as an image, foreign matter in the layer
No image can be obtained to observe the defect. Therefore, subjecting such non-observable foreign matter / defects in the layer to time-consuming electron microscopic observation causes a significant decrease in productivity. For this reason, the efficiency of observation is improved if it is determined in advance whether the surface layer is inside the layer or not.

A method for obtaining the Z coordinate of a foreign substance / defect from the Z coordinate at the time of detection was proposed by the present invention.
Another means for obtaining height information at the time of detection by an optical system or at the time of observation by an optical observation system. The optical path on the image side is divided by an optical path dividing unit 203 with respect to the object-side reference focus position 202 of the optical system 201, and the image detection unit 2 disposed on the back focus side with respect to the image-side reference focus position 206
04 and the image detection means 205 arranged on the front focus side respectively obtain images. If the position of the object is on the rear focus side, the contrast in the image detection means 204 is high. Conversely, if the position of the object is on the front focus side, the contrast in the image detection means 205 is high. In the middle, the contrasts of the two images are equal. Thereby, the height information of the foreign matter / defect can be obtained.

On the other hand, in order to determine whether the foreign matter / defect is in the layer, it is necessary to detect the position of the outermost layer. If the position of the foreign matter / defect is lower than the position of the outermost layer, it means that the foreign matter / defect is in the layer. In FIG. 21, the outermost surface 210
Illumination at a shallow angle with respect to the surface
4, the position of the reflected light 215 is detected by the position sensor 216, and the position of the outermost surface 210 is detected.
Since the illumination is at a shallow angle, most of the light is reflected on the surface, and foreign matter / defects 213 in the surface layer 212 and the bottom layer 211 are not reflected.
Can be detected without being affected by In this case, it is more preferable that the illumination light has a vibration direction of the electric field parallel to the surface of the sample to be inspected. This is because the reflectance of light in this vibration direction is high.

As shown in FIG. 22, the position of the outermost surface 210 may be obtained using a stylus 214 of an atomic force microscope. Alternatively, the outermost surface 210 may be detected using an air micrometer 215 as shown in FIG. The detection of the outermost surface 210 by these non-optical means makes it possible to determine the position of the outermost layer 210 without any influence in the layer, and as a result, it is possible to clearly determine whether the foreign matter / defect is in the layer. There is.

The determination as to whether or not the foreign matter / defect exists in the layer may be made at the time of observation after the end of the inspection or during the inspection. When the inspection is performed during the inspection, the result of the determination as to whether the existing position is in the layer is added to the inspection result.

FIG. 24 shows an observation flow in the case where it becomes clear whether or not the existence position of the foreign matter / defect exists in the layer by means such as the above. The detection result 2401 is subjected to a determination 2402 as to whether it is in the layer or not, and separated into a foreign matter / defect 2403 in the layer and a foreign matter / defect 2404 in the outermost layer. And the outermost foreign matter /
Observation 2405 of only the defect with an electron microscope or, if necessary, with an optical microscope is performed.

FIG. 3 shows an example of a block diagram of the defect detection processing circuit. Defect determination signal 35 indicating that a defect has been detected
The defect detection coordinate information 36 such as the X, Y, and Z coordinates of the defect detection position is sent to the detected defect memory 33. At the same time, the defect detection signal 34 and the characteristic amount of the detected defect (the amount of light detected from the defect, the size of the detected defect, the fatality of the defect, etc.) obtained by processing the detected position in the design by the defect attribute processing system 32 are obtained. Similarly, it is stored in the detection defect memory 33.

The detected defect information control system 37 records the detected defect result information on the detected defect memory in a storage medium 38 such as a magnetic, optical, or magneto-optical recording medium, outputs the information to a display, or outputs the information to a network. The defect detection result information 31 is transmitted to another device (for example, a review device or a server or a workstation) via the interface.

In the present invention, since the defect detection coordinates are detected by X, Y, and Z, the defect detection result list may display Z coordinates in addition to the X and Y coordinates. The Z coordinate is also data indicating a position where a defect exists in the cross-sectional structure of the sample to be inspected, and the fatality of the defect can be determined by comparing the Z-coordinate with a design drawing of the cross-sectional structure. FIG. 4 a) shows a display screen 41 (with Z coordinates) of the detected defect list. However, the Z coordinate is considered to be internal data required by the device at the time of review, and may be selected so as not to be displayed on the display screen. 4b) shows a display screen 42 (without Z coordinates) of the detected defect list. FIG. 3C shows a display condition setting screen 4 for switching between the two display methods.
3, which shows a state where switching is performed by the pointer cursor 44 operated by the pointing device.

FIG. 5 shows the state of the sample stage on the Y-axis stage 52 having an inclination as an error. The sample stage 53 at the position where the defect is detected and the sample stage 25 at the observation position are higher by δY. Are different. In this case, there is a possibility that focusing will not be performed at the time of observation at the Z height at the time of detection. However, correction can be made by measuring δY in advance and using it as an offset amount. Alternatively, the mounting of the observation optical system 22 may be offset.

FIG. 6 shows a situation where the Z-axis has an inclination. In FIG. 2B, X / X when the Z axis is not tilted is shown.
As shown in the Y-axis fluctuation confirmation visual field 63, the alignment mark 64 is located at the center of the visual field, and is located at the center of the visual field even when the Z height is changed. However, when there is a tilt of the Z-axis as an error, FIG. When the height of the Z-axis is changed from the state shown in FIG. 7 to the state shown in FIG. ΔZX in the X direction and δ in the Y direction
ZY transition.

Therefore, when the stored X, Y, and Z coordinates are called, the correction amount is placed on the Z-axis coordinates because the inclination of the Y-axis stage is described above or because the observation system is a separate device. In some cases, the defect does not exist at the called position. Therefore,
It is conceivable to measure δZX and δZY when the Z axis is moved by a fixed amount, and to correct the X and Y coordinates. FIG.
e) shows an apparatus condition setting screen 62 for inputting the amount of change in the X / Y coordinates due to the inclination of the Z axis as the error. This variation can be easily and automatically measured using an alignment mark as shown in b) of the same figure. However, depending on the condition of the sample, it may be more accurate to set manually, and the measurement can be selected. As shown on the screen.
FIG. 7 shows a state in which the change in the Z axis when the sample is scanned is caused by the inclination of the surface of the sample holder and the change in the thickness of the sample itself. In general, a change in the focal point position within the same inspection apparatus is exclusively due to these causes. Among them, a change in the thickness of the sample itself is considerably small in the case of a semiconductor-related sample. Therefore, if the inclination on the sample holder surface is measured in advance, it is possible to focus without knowing the Z height that is the focus position. FIG. 8a) shows an example 8 of the result of measuring the inclination of the sample holder surface.
1 is three-dimensionally illustrated.

FIG. 6B is a cross section of the curved surface 82 generated by complementing the Z height of the sample holder surface from the result of a), and the Z height is controlled from the X and Y coordinates according to this.
However, there is actually a change in the thickness of the sample, and depending on the amount of change in X and Y, an error in the Z height from the actual sample 83 may not be negligible. In this case, the observation system 24
Since it is noticed by the defocusing in the above, the correction surface 8 is manually adjusted and moved 84 from the correction surface to the actual sample surface, and the correction surface 8 after the movement in which the sample holder surface is offset by the adjustment amount δt.
The control is performed according to 5.

However, this function may malfunction if the sample thickness fluctuates greatly in a short cycle. As shown in FIG. 9, the function is stopped on the Z-axis offset correction selection screen 91. , Execution should be selectable.

When this function is used, when detection and review are performed by different devices, the accuracy of calling the Z coordinate can be improved.

The inclination of the holder surface is unique to the apparatus. Therefore, if the inclination is too large to be ignored, no matter how much the detected Z coordinate information is sent, the Z position cannot be reproduced by another apparatus. . So, received Z
The coordinates need to be corrected by the review device according to the inclination of the sample holder surface of the review device. In this case, the Z-axis coordinates from which the influence of the inclination of the sample holder surface on the inspection device side and the offset due to the difference of the origin of each device are removed before sending out may be sent out. And on the review device side,
Only the inclination of the sample holder of the device is corrected. Alternatively, the detected Z coordinate itself, the offset amount unique to the device, and the inclination amount of the sample holder surface may be sent. Alternatively, the unique amount of each device may be held in the review device in advance, or may be held by the network server.

FIG. 10 shows a situation in which the detection defect of the mat portion on the memory LSI is reviewed. Memory L
In a mat portion where cells in the SI are continuous, even a very small defect is often fatal, and the defect that must be detected by the inspection device is very fine. The task of checking is also difficult.

FIG. 10A) shows that one cell 101 exists continuously at a pitch Ps, and one of the cells 101 has a defect 102.
FIG. The called position is (Xm, Ym), and these coordinates often have a certain approximate value depending on the accuracy of coordinate management of the inspection device. And if the defect location is unclear,
Further, it is not possible to perform high-magnification observation (for this reason, the field of view is narrow), and it is not possible to perform detailed defect analysis, and the detection result merely indicates a monitor value indicating the number of defects on the inspection sample. Absent.

FIG. 6B is an image obtained by a) moving its own image by the cell pitch (here, the image is moved in the horizontal direction in the drawing, but may be in the vertical direction or in the oblique direction). FIG. 3C shows a state in which the images of a) and b) are added. The difference image between a) and b) is shown because of the addition of the inversion, but the difference 104 or 103 between the two images is shown.
Is appearing. Since all cells have the same shape except for the defective portion, the difference between the two indicates the defective portion. However, two differences occur for one defect because the position of the defect is moving between the two images. In c), the pointer cursor points to a place considered to be defective,
Defect or defect position candidate mark 105 at defect candidate position
This shows a state in which is generated. d) shows that the two candidate points are enlarged and one is confirmed to be defective. The coordinates at this time are (Xd, Yd), which means that more accurate coordinates have been obtained. As a result, it becomes possible to analyze a defect, furthermore, apply a high-powered apparatus such as an SEM, or perform analysis using an ion beam apparatus.

FIG. 11A is a screen for designating the amount of shift of the cell in FIG. Although there is a method of inputting design numerical values in this way, design data is required and is not suitable for the site. Therefore, in FIG. 2B, the cell cursor indicates a cell pitch indication pointer (reference) 1 with the pointer cursor on the confirmation image.
The cell pitch is indicated on the cell pitch display / input window 111 by specifying a position representing the cell pitch by using the reference numeral 13 and the cell pitch instruction pointer (reference) 112, and calculating the shift amount. In this way, it is possible to specify the cell pitch without design data. To improve the measurement accuracy at this time, a plurality of cell pitches may be measured and averaged, or a length corresponding to several cell pitches may be designated and divided by the number.

FIG. 11c) shows a state in which when a difference image is obtained by shifting the cell, a minute alignment error is generated, and false reports 114 and 115 due to the alignment error are generated. Therefore, fine adjustment of the shift amount is required, and the cell shift amount fine adjustment display / input window 11 is required.
6, the fine adjustment target handle 117 is dragged with a pointing cursor, and a fine adjustment amount is manually instructed. At this time, it is preferable that the correspondence between the drag amount of the target handle and the shift amount can be changed by the fine movement setting instruction button 118, the coarse movement setting instruction button 119, or the like.

As described above, as shown in FIG.
, A cell pitch (cell shift amount) 1603 is obtained, and a difference image 1605 between the created image 1604 obtained by shifting the input image 1602 by the cell shift amount and the input image 1602 is obtained and displayed 1606.

As shown in FIG. 17, an enlarged image 1703 centering on the defect candidate position 1702 is created from the displayed difference image 1701, and the detected foreign matter is extracted from the image 1704 called at the same position. A defect category such as the defect is determined 1705, and the information is stored 1707.

FIG. 12A is a sectional view of a foreign substance 122 attached to a circuit pattern 121 on a sample to be inspected. As described above, since the heights of the circuit pattern and the defect such as the foreign matter are different, the heights at which the focus is achieved are different. Further, the focusing position also differs depending on the difference in the material, that is, the difference in the refractive index of the material. In FIG. 13B, the circuit pattern 123 is focused, and in FIG. 13C, a defect such as a foreign substance is focused. In the focused state, the dark part becomes the darkest, the bright part becomes the brightest, and the contrast becomes maximum. As the focus deviates, both converge to an intermediate value, and the contrast decreases.

FIG. D) is a graph showing the situation.
The horizontal axis indicates the height, and the vertical axis indicates the brightness of the image. The brightness of the circuit pattern behaves like a detection signal 127 from a pattern that becomes bright at the in-focus position and a detection signal 128 from a circuit pattern that becomes dark at the in-focus position, and the contrast becomes maximum at the Z-axis height Zf. In addition, the brightness of a defect such as a foreign substance is detected by a detection signal 126 from a foreign substance that becomes dark at the in-focus position.
It behaves like 29, and the contrast is maximized at a height shifted from Zf by ΔF.

Therefore, a plurality of images having different heights are taken in, and the maximum or minimum of the change in contrast or brightness at each point is obtained from the result to display the in-focus position at each point. Thirteen. Although there is a three-dimensional graph display method for displaying the in-focus position, here, as an example, in order to improve the correspondence with the observation image, two-dimensional mapping is performed, and the difference in height is simulated. The display 131 is displayed in color or shade. Then, the contrast maximum height display menu screen 13 is displayed so that the height in the normal portion becomes zero.
In the pseudo color setting menu 134 in 3, the pseudo color setting condition pointer 135 is adjusted by operating with a pointing device.

As described above, as shown in FIG. 18, data 1810 in which the height at which the contrast or the brightness is maximized at each pixel is created for the images 1801 to 1809 inputted with the height changed from Z1 to Zn. And display it 18
11

FIG. 14 shows a position where the coordinates are designated by the number of cells (for example, "36 from the upper left of the mat portion, 21
The method for calling “individual”) is described.
Such a designation method is often seen in the case of failure analysis, etc., but since the number to be counted tends to be large and the shape of the cells to be counted is exactly the same, counting errors are easy to occur, so Often, a great deal of time must be spent calling only the desired point.

In FIG. 7A, the upper left of the mat section is first called. Here, the cell pitch in the horizontal and vertical directions is shown in FIG.
Input in the same manner as in 1. The pitch is input by the X-direction cell pitch instruction pointer 141 and the Y-direction cell pitch instruction pointer 142. However, this time, unlike the case of FIG. 11, since the movement amount is large, the θ (rotation in the plane of the sample) alignment of the sample to be inspected is performed. The cell pattern 144 is also predicted by using the cell alignment inclination instruction pointer 143, and the cell pattern 144 is also predicted.
A cursor line aligned with one side of the column is drawn with a pointing device, and correction in the θ direction is performed based on the inclination of the cursor line. The result is confirmed in the cell pitch amount display / input window 145 or directly input numerically from here. Then, the moving cell number input / display window 146
Then, the visual field moves by the product of the cell pitch and the number. A target cell pattern display mark 148 is displayed on the called cell to improve visibility.

This type of work is often exchanged between various types of microscopes, and may involve a process such as a photomask and a wafer that is turned left and right and upside down. For this reason, the indication of the (relative) movement amount by the number often causes an error in the coordinate system. If the desired position is not called even after the number of movements, the operator has to consider the possibility of errors in various coordinate systems, thereby reducing the work efficiency. Therefore, in the present invention,
By clicking the coordinate system conversion button and the movement amount inversion button 147, various types of coordinate system changes can be easily performed instantaneously, and the possibility of studying the possibility is devised to improve the workability.

Next, a method of specifying a mat portion will be described. The mat portion is a much larger area than a single cell, and although it is not impossible to specify in the same way as a cell, it involves considerable difficulty. FIG. 15a) shows the magnification and the general field size of a microscope objective lens often used in an observation optical system. In contrast to the field size shown in the table, the chip size of the semiconductor LSI is several times larger than the field size of the objective lens as shown in FIG. , Memory LS
There is a peripheral circuit portion 152 of I. Although other optical devices, such as a photographic camera lens, have a sufficiently large field of view, it is not preferable to switch and share lenses of different optical devices with a single optical system by one optical system, and inconvenience occurs. .

Accordingly, in the present invention, as shown in FIGS. C) and d), the position of the field of view is moved in the field of view of the objective lens to capture a plurality of images together with the position of the field of view, and as shown in FIG. To create an image (macro image) with a wide field of view by synthesizing the image on the image memory of the above, and display it. On the macro image, a desired mat portion is designated with a pointing device or the like, and the visual field position is calculated back therefrom, and the desired mat portion is called to the visual field position of the objective lens.

In addition, if the capturing of FIGS. C) and d) is performed with a high-magnification image, the time required for capturing and the amount of memory for the image are increased, but the image on the memory is mainly used until detailed cell observation. Can be performed like an electronic zoom.

When a macro image is used, various settings such as an inspection area, a negative window, a chip interval, and the like at the time of inspection can be directly instructed not by a design drawing but by an image of a chip. This produces a greater effect in the case of a photomask or the like that is manufactured in a small quantity of other products.

FIG. 19 shows this state. A desired inspection area (or an undesired inspection area) 194 is designated by a pointing device or the like from a synthesized image 193 synthesized from the captured image 192 shown on the display screen (including the screen window) 191.

From the designated position 195 on the screen to the screen coordinates-
The actual inspection coordinates 196 are obtained from the conversion parameters 198 of the actual inspection (observation) coordinates.

Next, a repetition parameter 199 for expressing a multichip or the like is used to generate a coordinate group 197 taking into account the repeatability of the multichip or the like. Using this data, a detection signal (or detection result) 1910 is masked by a masking means (or enable means) 1911.
And a detection result 1912 is obtained.

[0080]

According to the present invention, when a defect is detected, in addition to recording the coordinates of the detected defect in the plane of the substrate, the position of the detected defect in the height direction is recorded at the same time. By reproducing not only in the plane but also in the height direction, there is an effect of improving the efficiency of the work of reviewing the detected defects.

[Brief description of the drawings]

FIG. 1 is a schematic front view of an apparatus including a defect detection optical system and an observation optical system according to an embodiment of the present invention.

FIG. 2 is a cross-sectional view of a sample illustrating surface foreign matter and interlayer foreign matter to be detected.

FIG. 3 is a block diagram illustrating a detection system that processes X, Y, and Z coordinates as foreign object data according to an embodiment of the present invention.

FIG. 4 shows an example of a display screen of a detected defect list according to the present invention.

FIG. 5 is a schematic front view of an apparatus including a defect detection optical system and an observation optical system according to an embodiment of the present invention, for explaining the effect of an inclination of a stage as an error on observation. It is.

FIG. 6 is a view showing the relationship between X and X by a Z-axis stage having a tilt;
FIG. 7 is a schematic front view of a Z-axis stage for explaining that Y-axis coordinates change.

FIG. 7 is a schematic front view of a Z-axis stage illustrating an influence of a tilt of a sample holder surface and a change in thickness of a sample on a Z height.

FIG. 8 is a diagram for explaining the correction of the inclination of the sample holder.

FIG. 9 is a front view of a display for explaining ON / OFF of a correction function.

FIG. 10 is a cell shift image for explaining how to improve the confirmability of a defect based on the difference between the cell shift images.

FIG. 11 is a front view of a display in which a function for performing fine adjustment of cell shift alignment is displayed on a screen.

FIG. 12 is a diagram for explaining how to improve the confirmation of foreign matter.

FIG. 13 is a front view of the display on which a screen for setting a display method of the maximum contrast height is displayed.

FIG. 14 is a front view of the display on which a screen for supporting the movement of the number of cells is displayed.

FIG. 15 is a plan view of the field of view of the objective lens, showing a method of creating a macro image with an objective lens having a narrow field of view.

FIG. 16 is a block diagram showing a flow for obtaining a difference image.

FIG. 17 is a block diagram illustrating a flow for obtaining high-precision coordinates from a difference image.

FIG. 18 is a block diagram showing a flow for obtaining a maximum brightness image from an image input at different heights.

FIG. 19 is a block diagram illustrating a procedure for masking an inspection result from a composite image.

FIG. 20 is a schematic front view of an optical system illustrating a configuration for measuring a Z coordinate of a foreign matter / defect position.

FIG. 21 is a schematic cross-sectional view of a sample and a position sensor for explaining a method of optically measuring the position of the outermost layer of a multilayer film.

FIG. 22 is a schematic cross-sectional view of a sample and a stylus of an atomic force microscope for explaining a method of mechanically measuring the outermost layer position of a multilayer film.

FIG. 23 is a schematic cross-sectional view of a sample and an air micrometer for explaining a method of mechanically measuring the outermost layer position of a multilayer film.

FIG. 24 is a block diagram illustrating a procedure for selecting a foreign substance / defect to be observed with an electron microscope.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Inspection sample, 2 ... Sample holder, 3 ... Z-axis stage, 4 ... X-axis stage, 5 ... Y-axis stage, 9
... Autofocus control system, 10 ... Autofocus sensor, 11
... Stage control system, 12 ... Defect detection sensor, 13
... Defect detection processing system, 16 ... Display for detecting detected defects, 17 ... Display for operating condition display / input, 1
8 ... Observation focus mechanism control system, 19 ... Observation image sensor, 20 ... Observation illumination system, 21 ... Observation focus system drive mechanism, 22 ... Observation optical system, 23 .. Illumination system for defect detection, 24 ... Detection optical system, 26 ... Sample substrate to be inspected, 2
7 ... lower thin film layer of the sample to be inspected, 28 ... upper thin film layer of the sample to be inspected, 29 ... surface foreign matter, 30 ... interlayer foreign matter, 3
7 ... detection defect information control system, 83 ... actual sample surface, 1
02: pattern defect, 103: defect candidate, 10
4 ... Defect candidate, 105 ... Defect position or defect candidate position indicating mark, 114 ... False report due to alignment error, 115 ... False report due to alignment error, 122
... Foreign matter, 126 ... Detection signal from the foreign matter darkened at the in-focus position, 127 ... Detection signal from the pattern brightened at the in-focus position, 128 ... From the circuit pattern darkened at the in-focus position 129 ... Detection signal from a foreign substance that becomes brighter at the in-focus position, 131 ... Point where the Z height at which the contrast becomes maximum differs from the circuit pattern, 133 ... Contrast maximum height display menu screen 136 ... Pointing device, 145 ... Cell pitch amount display / input window, 146 ... Moving cell number input / display window, 147 ... Moving amount reverse button, 148 ... Target cell pattern display mark .

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Atsushi Shimoda 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Inside of Hitachi, Ltd. Inside Hitachi, Ltd. Production Technology Laboratory

Claims (20)

[Claims] 1. A method for inspecting a defect of a sample, the method comprising: imaging a sample to obtain an image of the sample; detecting a defect of the sample from the image of the sample; and detecting three-dimensional coordinates of the detected defect. A defect inspection method which stores information and locates the detected defect in an observation field of view based on the stored information on the three-dimensional coordinates of the defect and observes the defect. 2. The defect inspection method according to claim 1, wherein the detection of the defect is performed over an entire inspection target area of the sample. 3. Observing the detected defect based on the stored information on the three-dimensional coordinates of the defect. A) Positioning the defect detection position in an observation field of view; and b) Observing with a low magnification objective lens. C) Switching to a high-magnification objective lens 2) Using the above two objective lenses in a preset amount Z
3. The defect inspection method according to claim 1, wherein the defect inspection method is performed by relatively moving in an axial direction, and e) observing with a high-magnification objective lens. 4. A defect inspection apparatus for inspecting a defect of a sample, comprising: a defect detection unit configured to image a sample and detect a defect of the sample from an image of the sample obtained by the imaging; Storage means for storing information relating to the three-dimensional coordinates of the detected defect of the sample; output means for outputting information relating to the three-dimensional coordinates of the defect stored in the storage means; A defect inspection apparatus comprising: an observation unit configured to move a defect of the sample into an observation visual field based on information on dimensional coordinates to obtain an enlarged image of the defect. 5. A positioning section for positioning the defect detection position by the observation means, a low magnification observation section for observing with a low magnification objective lens, and a high magnification observation section for switching to a high magnification objective lens for examination. The defect inspection apparatus according to claim 4, further comprising: a correction unit that relatively moves the two objective lenses in the Z-axis direction. 6. A defect inspection method for inspecting a defect at a desired portion on a sample, wherein the plurality of images of the sample are obtained by sequentially imaging the sample using an optical system having a first magnification. A plurality of images of the sample are recorded, and the recorded plurality of images are combined to obtain an image of an area larger than the field of view on the sample by the first magnification optical system. The above-mentioned desired portion is obtained, the desired portion is imaged using an optical system of a second magnification, and information on the image of the desired portion of the sample imaged using the optical system of the second magnification is obtained. A defect inspection method characterized by outputting. 7. The optical system according to claim 6, wherein the second magnification of the optical system is equal to the first magnification.
7. The defect inspection method according to claim 6, wherein the magnification is larger than the magnification of the defect. 8. A defect inspection apparatus for inspecting a defect at a desired position on a sample, a first imaging means for sequentially imaging the sample to obtain a plurality of images of the sample, and the first imaging Storage means for storing a plurality of images of the sample taken by the means; and combining the plurality of images stored in the storage means to create an image of an area larger than the field of view on the sample of the first imaging means. Means for obtaining a desired location on the sample from the image of the large area created by the synthetic image creating means; and capturing the desired location determined by the setting means to obtain the desired location A defect inspection apparatus, comprising: a second imaging unit that obtains an image of the sample; and an output unit that outputs information relating to an image of a desired portion of the sample captured by the second imaging unit. 9. The defect inspection apparatus according to claim 8, wherein said second imaging means obtains an image of said sample having a larger magnification than said first imaging means. 10. A defect inspection method for detecting a defect of the repetitive pattern on a sample on which a repetitive pattern is formed and observing the detected defect, comprising the steps of: a) imaging a sample having a repetitive pattern; B) detecting a defect candidate of the repetitive pattern from the obtained image; c) displaying an image of the detected defect candidate on a screen; d) selecting an image of the defect candidate displayed on the screen. E) A defect inspection method, wherein a part to be enlarged and displayed is designated, and e) the designated part is enlarged and displayed on the screen. 11. The apparatus according to claim 1, wherein coordinates of the portion enlarged and displayed on said screen are stored as defect coordinates.
0. 12. The defect inspection method according to claim 11, wherein said defect coordinates are three-dimensional coordinates. 13. A defect inspection method for detecting a defect on a substrate sample and observing the result of the detection, the method comprising: a) positioning a corresponding position in a visual field of an observation system from defect detection coordinates of the detection result; An image at a plurality of heights is taken into an image processing system to obtain a taken image. C) A height at which the brightness information is maximized is obtained from the taken image. F) The obtained height information is displayed on a screen together with the taken image. G) pointing a defect candidate point of the captured image displayed on the screen with a pointing device to mark the defect; h) displaying the marked defect candidate point further enlarged and displayed. Method. 14. The defect inspection method according to claim 13, wherein coordinates of a position designated on the enlarged image displayed on the screen are stored as defect coordinates. 15. The defect inspection method according to claim 14, wherein said defect coordinates are three-dimensional coordinates. 16. A defect inspection apparatus for detecting a defect of the repeated pattern on a sample on which a repeated pattern is formed and observing the detected defect, comprising: a) imaging a sample having a repeated pattern; B) a defect candidate detecting means for detecting a defect candidate of the repetitive pattern from an image taken by the imaging means; and c) an image of a defect candidate detected by the defect candidate detecting means. Display means for displaying on the screen, d) enlarged display designating means for designating a portion to be enlargedly displayed from the image of the defect candidate displayed on the screen of the display means, and e) the defect candidate is designated by the enlarged display designating means. A defect inspection apparatus comprising: an enlarged display unit for enlarging and displaying a designated portion in an image on a screen of the display unit. 17. The defect inspection apparatus according to claim 16, wherein said enlarged display designating means is a pointing device for designating a position on said screen of said display means. 18. A method for inspecting a defect of a sample, the method comprising: picking up an image of the sample by using a first imaging means to obtain an image of the sample; detecting a defect candidate of the sample from the image of the sample; Obtaining position information of the defect candidate in the sample surface; obtaining position information of the sample surface in a normal direction of the sample surface; and obtaining position information of the defect candidate in the sample surface and a method of the sample surface. The detected defect candidate is located in the observation field of view of the second imaging means based on the position information in the line direction, and the second imaging means images the defect candidate in the observation field of view and A defect inspection method characterized by displaying an image on a screen. 19. The defect inspection method according to claim 18, wherein said second imaging means images said sample at a higher magnification than an image of said sample imaged by said first imaging means. 20. A defect inspection apparatus for detecting a defect candidate of a sample and observing the detected defect candidate in detail, comprising: first imaging means for imaging a sample to obtain an image of the sample; A defect candidate detecting unit that detects a defect candidate of the sample from an image of the sample obtained by the imaging unit and obtains positional information of the defect candidate in the sample surface; and a normal direction of the sample surface. Height detection means for obtaining position information of the sample surface in the above, position information in the sample plane of the defect candidate obtained by the defect candidate detection means and the normal direction of the sample surface obtained by the height detection means A second imaging unit for positioning the detected defect candidate in the observation field of view based on the position information of the second imaging unit; and an image of the defect candidate in the observation field of view by the second imaging unit. Display means for displaying on a screen. Apparatus.