JP2011029324A - Method and system for updating recipe - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and system for updating a recipe, capable of updating an optical condition stored in the recipe in order to detect a defect of a circuit pattern more easily without reducing an inspection time in a visual inspection device inspecting a semiconductor substrate including the circuit pattern formed thereon based on the recipe storing the optical condition. <P>SOLUTION: In a recipe updating method for updating the recipe in a visual inspection device inspecting a semiconductor substrate based on the recipe, a first image is obtained by imaging the semiconductor substrate according to an optical condition stored in a recipe storing means, and detection is performed whether or not a defect exists in the obtained first image. When an existence of the defect is detected, the optical condition is changed, a second image is obtained by imaging the semiconductor substrate according to the changed optical condition, and the optical condition stored in the recipe storing means is updated to the changed optical condition. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a recipe update method and a recipe update system for updating a recipe in an appearance inspection apparatus that inspects a semiconductor substrate on which a circuit pattern is formed based on a recipe in which optical conditions are stored. The present invention relates to a recipe update method and a recipe update system for updating the recipe to make it easier to detect defects.

  Conventionally, in a semiconductor manufacturing process for manufacturing a semiconductor device, in order to secure a high yield, it is important to accurately detect defects such as scratches, dust, unevenness, and dirt generated on a semiconductor substrate during the manufacturing process. is there.

  In order to detect these defects, the surface of the semiconductor substrate is irradiated with illumination light, the specularly reflected light, diffracted light, scattered light, etc. are imaged with an imaging device such as a line sensor camera, and the captured image data is captured. In general, defects are detected by image processing. For example, the stage on which the semiconductor substrate is placed is moved based on the defect position detected by the host inspection apparatus, and imaging is performed at a low magnification that allows the semiconductor substrate to be within the field of view. Further, the defect position is automatically detected by the image processing for the captured image and displayed on the display screen via the graphical user interface. And this imaging is performed based on a predetermined optical condition.

  In such a defect detection step, a technique for automatically detecting an optical condition for detecting specular reflection light or diffracted light using a non-defective semiconductor substrate is disclosed (for example, see Patent Document 1). .

In addition, a technique for automatically eliminating detection system noise by tuning (updating) recipe parameters is disclosed (for example, see Patent Document 2).
In addition, a defect detection inspection technique in which defects are not overlooked or duplicated by performing inspection under a plurality of optical conditions is disclosed (for example, see Patent Document 3).

WO2002 / 027305 JP 2006-17744 A JP 2005-17159 A

  However, in order to obtain good optical conditions in the conventional technology as described above, the circuit pattern of the semiconductor substrate to be inspected is simple, or design information such as a pattern pitch is necessary in advance. In the case of an actual semiconductor substrate, the circuit pattern is complicated, or multilayer wafers that have undergone multiple manufacturing processes may not find good optical conditions only with the surface layer design information. there were.

  In addition, because we focus only on noise removal of the detection sensor, it is not possible to select optical conditions with higher detection sensitivity, or to perform multiple inspections with more optical conditions to prevent missing defects, so inspection time There is a problem that sometimes becomes longer.

  The present invention has been made in view of the above-described actual situation, and in a visual inspection apparatus that inspects a semiconductor substrate on which a circuit pattern is formed based on a recipe in which optical conditions are stored, a circuit pattern defect is further detected. In order to facilitate, an object of the present invention is to provide a recipe updating method and a recipe updating system capable of updating the optical conditions stored in the recipe without reducing the inspection time.

The present invention employs the following configuration in order to solve the above problems.
That is, according to one aspect of the present invention, the recipe update method of the present invention is a recipe update method for updating the recipe in an appearance inspection apparatus that inspects a semiconductor substrate based on a recipe, and is stored in a recipe storage unit. According to the optical conditions, the semiconductor substrate is imaged to obtain a first image, whether or not a defect exists in the acquired first image, and the presence of the defect is detected, Changing the optical condition, imaging the semiconductor substrate according to the changed optical condition, obtaining a second image, and updating the optical condition stored in the recipe storing means to the changed optical condition; It is characterized by.

  Further, the recipe update method of the present invention determines whether or not a more prominent defect appears in the second image than the defect appearing in the first image, and appears in the first image. When it is determined that a defect more conspicuous than the existing defect appears in the second image, it is desirable to update the optical condition stored in the recipe storage unit to the changed optical condition.

In the recipe update method of the present invention, it is preferable that the second image is an image obtained by capturing an image of a semiconductor substrate on which the defect is detected.
In the recipe update method of the present invention, it is desirable that the second image is an image obtained by imaging another semiconductor substrate of the same type as the semiconductor substrate in which the defect is detected.

  In the recipe update method according to the present invention, it is preferable that when the appearance inspection apparatus sequentially inspects the semiconductor substrate, it is detected whether or not there is a defect in the acquired first image every predetermined number of times.

In the recipe update method of the present invention, it is desirable that the optical condition is changed every predetermined number of times when the appearance inspection apparatus sequentially inspects the semiconductor substrate.
In addition, the recipe update method of the present invention, when detecting the presence of the defect, changes a plurality of the optical conditions, and the defect that is more conspicuous than the defect appearing in the first image is the second When it is determined that the image appears in the image, it is desirable to update the optical condition stored in the recipe storing means to the optical condition in which the defect appears most prominently among the changed optical conditions.

In the recipe update method of the present invention, it is desirable to change the optical condition using a graphical user interface when the presence of the defect is detected.
Moreover, according to one aspect of the present invention, the recipe update system of the present invention is a recipe update system that updates the recipe in an appearance inspection apparatus that inspects a semiconductor substrate based on a recipe, the recipe storing optical conditions In accordance with the optical conditions stored in the recipe storage means, a storage means, a first image acquisition means for imaging the semiconductor substrate and acquiring a first image, and a first image acquired by the first image acquisition means A defect detecting means for detecting whether or not a defect exists in the image of the image, and an optical condition changing means for changing the optical condition stored in the recipe storing means when the defect detecting means detects that a defect exists. And capturing a second image by capturing an image of the semiconductor substrate on which the defect is detected according to the optical condition changed by the optical condition changing means. Compared with a defect appearing in the first image acquired by the image acquisition means and the first image acquisition means, a more prominent defect appears in the second image acquired by the second image acquisition means. Determining means for determining whether or not a defect that is more prominent than the defect appearing in the first image appears in the second image by the determining means; And optical condition updating means for updating the optical condition stored in the optical condition to the changed optical condition.

  The present invention changes the optical conditions when a defect in the circuit pattern on the semiconductor substrate is detected, and in the subsequent inspection of the semiconductor substrate, the changed optical conditions are used as a recipe. There is an effect that the defect of the circuit pattern can be found.

It is a figure which shows the outline of the external appearance inspection apparatus in 1st Embodiment to which this invention is applied. It is a figure for demonstrating the outline | summary of the imaging by the imaging means. It is a figure which shows the example of GUI provided by the optical condition update means 16 in 1st Embodiment. 4 is a flowchart showing a flow of inspection processing executed in the appearance inspection apparatus 1. It is a figure which shows the outline of the external appearance inspection apparatus in 2nd Embodiment to which this invention is applied. It is a figure which shows the example of GUI provided by the optical condition update means 16 in 2nd Embodiment. It is a figure which shows the data list required in order to automate an analysis data acquisition process. It is a figure which shows the data list required in order to automate an optical condition update process. It is a figure which shows the data list required in order to automate an optical condition update process in the modification of 3rd Embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that, in all the drawings, even if the embodiments are different, the same or corresponding members are denoted by the same reference numerals, and redundant description is omitted.
(First embodiment)
FIG. 1 is a diagram showing an outline of an appearance inspection apparatus according to a first embodiment to which the present invention is applied.

  In FIG. 1, an appearance inspection apparatus 1 includes a semiconductor substrate on which various circuit patterns are formed, such as a liquid crystal display (LCD), a plasma display panel (PDP), an organic electroluminescence (EL) display, and an SED (SED). This is an apparatus for inspecting a semiconductor substrate such as a flat panel display (FPD) such as a surface-conduction electron-emitter display, a semiconductor wafer, or a printed circuit board based on a recipe. As shown in FIG. 1, the appearance inspection apparatus 1 includes an inspection unit 2 that inspects a semiconductor wafer W (hereinafter referred to as a wafer W) that is a circular semiconductor substrate, and controls and various data processing of the inspection unit 2 And a control unit 3 that performs the above-described process, and the recipe can be updated as will be described later.

  The inspection unit 2 includes a scan stage 4 that is an operation means that can move in one axial direction. For the scan stage 4, for example, a linear guide or the like is used. On the scan stage 4, a rotation stage 5 for placing the wafer W is provided. The rotation stage 5 is a rotation means in which a mounting unit 6 concentric with the wafer W is attached to a rotation mechanism.

  An imaging unit 7 is disposed above the rotary stage 5. The imaging means 7 is composed of a line sensor camera 8 and a cylindrical lens 9 arranged so that the image plane passes through the center of the wafer W.

  The inspection unit 2 further includes a line illumination 11 as an illumination unit that illuminates the image plane on the opposite side of the imaging unit 7 with the placement unit 6 interposed therebetween. Although not shown, between the line illumination 11 and the mounting portion 6, there is a polarizing plate insertion / extraction mechanism and a multiple band pass filter selection mechanism for selecting the wavelength of light to be detected. .

FIG. 2 is a diagram for explaining an outline of imaging by the imaging unit 7.
When the imaging unit 7 captures an image of regular reflection light, the line sensor camera 8 is set to an imaging angle at which a regular reflection image of the surface of the wafer W can be captured. Further, as shown in FIG. 2, the line sensor camera 8 has a rotating means 10 that rotates around the image plane in order to capture an image other than regular reflection of the surface of the wafer W. It is possible to image not only the regular reflection light of the surface but also the diffracted light. Further, the inspection unit 2 has an illumination 12 for photographing scattered light in addition to the line illumination 11 for imaging regular reflection light and diffracted light with respect to the image plane.

Returning to the description of FIG.
The image sensor of the line sensor camera 8 is arranged so that its longitudinal direction is perpendicular to the moving direction of the scan stage 4. Further, the imaging range of the surface of the wafer W of the line sensor camera 8 is linear, and the length thereof is longer than the diameter of the wafer W. The scan range of the scan stage 4 is a range in which the entire surface of the wafer W can be imaged by the line sensor camera 8.

  The control unit 3 is a control computer having a central processing unit, an internal memory, an external storage device, and the like. The control unit 3 includes analysis data acquisition means 13, inspection information storage means 14, defect extraction means 15, optical condition update means 16, and recipe information storage means 17.

  Such an appearance inspection apparatus 1 images the wafer W according to optical conditions stored in the recipe information storage means 17, for example, camera swing angle, stage rotation angle, illumination light quantity, polarizing plate, bandpass filter, etc. Whether or not a defect exists in the image is detected, and when the presence of a defect is detected, the optical condition is changed. Then, according to the changed optical condition, the wafer W is imaged again, and when it is determined that a more prominent defect appears in the image, the optical condition is updated to the changed optical condition.

  The analysis data acquisition means 13 is connected to the line sensor camera 8 and the rotary stage 5, and interlocks an output instruction for outputting image data line by line from the line sensor camera 8 and a drive instruction for the rotary stage 5. Generate an image. Here, an image obtained by this operation is called a stage rotation image. Similarly, the analysis data acquisition unit 13 can also generate an image in conjunction with the drive instruction of the rotation unit 10. Here, an image obtained by this operation is called a camera swing image. Further, the analysis data acquisition means 13 is also connected to the scan stage 4 so that an image can be generated in conjunction with the output of the line sensor camera 8 and the scan of the scan stage 4. Here, an image obtained by this operation is called an inspection image or an analysis image depending on the purpose.

  Furthermore, the analysis data acquisition means 13 can hold each image data of the stage rotation image, the camera swing image, and the inspection image or analysis image described above in the internal memory. As a specific configuration example of the analysis data acquisition unit 13, an imaging board that controls the line sensor camera 8 with a computer, a rotation drive of the scan stage 4 and the rotation stage 5, and a rotation drive of the rotation unit 10 are described. And a system in which driver software for operating each board individually and application software for linking the driver software are installed.

  The defect extraction unit 15 performs pixel comparison between a non-defective image obtained by imaging the non-defective wafer W in advance by the analysis data acquisition unit 13 and an inspection image captured by the analysis data acquisition unit 13 at the time of inspection by a known image processing technique. Implementation is performed, and pass / fail judgment information of the wafer W and defect position information are created, and those information are stored in the inspection information storage means 14.

  The recipe information storage unit 17 stores recipe information including parameters such as optical conditions and defect detection sensitivity when each image is captured by the imaging unit 7. Further, the recipe information storage unit 17 also stores information shown in FIGS. 7 and 8 to be described later.

  The optical condition update unit 16 updates the optical condition stored in the recipe information storage unit 17. At that time, a GUI (Graphical User Interface) can be provided, and a displayable coordinate system and various types of information can be selected and input to display the contents.

FIG. 3 is a diagram illustrating an example of a GUI provided by the optical condition update unit 16 according to the first embodiment.
In the GUI example shown in FIG. 3, the stage rotation imaging unit 18 in the upper third area of the screen, the camera swing imaging unit 19 in the middle third area of the screen, and the lower third of the screen. The recipe optical condition update unit 20 is provided in the area. Details of these will be described later.

Returning to the description of FIG.
The inspection information storage unit 14 stores image data such as a stage rotation image, a camera swing image, an inspection image, and an analysis image stored in the internal memory by the analysis data acquisition unit 13 in an external storage device or the like. Further, the inspection information storage unit 14 also stores the inspection image data and defect position data generated by the defect extraction unit 15 and the optical condition data obtained by the analysis data acquisition unit 13.

The appearance inspection apparatus 1 according to the first embodiment to which the present invention is applied has been described above.
Next, the flow of the inspection process executed in the above-described appearance inspection apparatus 1 will be described.
FIG. 4 is a flowchart showing a flow of inspection processing executed in the appearance inspection apparatus 1.

  First, in step S21, an image capturing process is executed. That is, recipe information corresponding to the wafer W to be inspected is read from the recipe information storage unit 17, and an inspection image of the wafer W is acquired by the analysis data acquisition unit 13 according to the optical conditions included in the read recipe information.

  Next, in step S22, defect extraction processing is executed. That is, the defect extraction means 15 extracts defects using the inspection image acquired in step S21. Note that any known algorithm is used as the defect extraction algorithm.

In step S23, it is determined whether a defect has been extracted by the defect extraction process executed in step S22.
If it is determined that a defect has been extracted (step S23: Yes), an analysis data acquisition process at step S24 and an optical condition update process at step S25 are then executed. At this time, the appearance inspection apparatus 1 displays a GUI as illustrated in FIG.

First, in step S24, analysis data is acquired. Specifically, first, the operator sets conditions for acquiring analysis data using the GUI.
That is, any one of “regular reflected light 27”, “scattered light 28”, and “diffracted light 29” is selectively input as the measurement light condition 26 that is the type of light used for analysis. Only when the specularly reflected light 27 is selected and input, the selection input to the camera angle input 31, the camera swing imaging start button 32, and the set angle 33 among the items displayed in FIG. The angle of is displayed. The operations of items other than these can be selected and input in common for all measurement light conditions 26.

  Next, the stage rotation imaging unit 18 is operated. Specifically, after selecting the selection input of the band pass filter 30, the selection input of turning on or off the polarizing plate 34, and the selection input of the camera angle 31, the imaging start button 35 is pressed. By this operation, the analysis data acquisition unit 13 captures the stage rotation image according to the set optical condition and displays the captured stage rotation image on the display screen 36. On the display screen 36, line profile data 37 obtained by averaging the horizontal line data of the image in the vertical direction is displayed over the image data.

  Further, a scroll bar 38 is provided below the display screen 36, and the rotation angle of the rotary stage 5 can be set to an arbitrary angle. Then, the line 39 corresponding to the set angle is overlaid on the display screen 36, and at the same time, the set angle is displayed in the text box of the selected angle 40. In general, when a stage rotation image is captured on a pattern wafer, it can be seen that defect detection increases at the angle of the maximum peak of the aforementioned line profile such as the peak 41 if the light is specularly reflected or diffracted. ing. In the case of a dark field, defect detection becomes high at a minimum peak such as the peak 42.

  Then, by pressing the add button 44, the rotation angle of the rotary stage 5, the on / off state of the polarizing plate, etc. in the image data displayed on the display screen 36 can be added to the list 43. An unnecessary angle can be deleted by pressing the delete button 45 of the selection item in the list.

  With the above operation, the stage rotation angle condition determination button 46 is pressed while only the necessary angles are displayed in the list 43. Then, the analysis data acquisition means 13 stores the information displayed in the list 43 and the stage rotation image data in the inspection information storage means 14, and changes the contents of the menu list so that all the list IDs 47 displayed in the list 43 can be selected. Update.

  Next, the camera swing imaging unit 19 is operated. First, the stage rotation angle is selected and input from the menu list of the list ID 48, and the imaging start button 32 is pressed. By this operation, the analysis data acquisition unit 13 captures a camera swing image according to the set optical condition and displays it on the display screen 36a. The operation for registering the necessary camera angle in the list is omitted because it is the same operation as the stage rotation imaging. However, the camera angle setting angle 33 performs the same operation as the selection angle 40 described above. When the add to list button 44a is pressed, the target luminance value 49 of the maximum luminance value in the analysis captured image is also added to the list 43a.

  With the above operation, the “camera swing angle setting confirmation → automatic light quantity adjustment → save analysis data acquisition information” button 50 is pressed while only the necessary angles are displayed in the list 43a. Then, the analysis data acquisition unit 13 stores the information displayed in the list 43a and the camera swing image in the inspection information storage unit 14, and then performs the light amount automatic adjustment processing obtained by binary search or the like to determine the illumination light amount value 51 determined. The illumination light quantity value closest to the target luminance value 49 is added to the analysis data list 52 and displayed. That is, according to the changed optical conditions, the semiconductor substrate that has been detected to have the defect is imaged again, and compared to the defect that appears in the image acquired before and after the change, the image that has been re-imaged is more prominent. If it is determined whether or not is displayed, the updated optical condition is updated. Then, the contents of the optical condition list 53 currently registered in the recipe are updated by the add button 54, the delete button 55, etc. from the data list 52 for analysis. Thereby, in the subsequent inspection, the optical conditions of the updated optical condition list 53 can be used, so that the possibility of finding more defects more precisely becomes high.

  When the recipe update operation is completed by the above operation, the analysis data acquisition timing 56 is selected. Here, when the defective wafer detection time 57 is selected, the description is omitted because it has been described above. When the designated processing wafer interval 58 is designated, the analysis data acquisition process in step S24 and the step S25 are performed each time the number of wafers W having the value input in the text box 59 is inspected regardless of the presence or absence of the defective wafer W. The optical condition update process is performed.

  Finally, by pressing the recipe save & end button 60, the updated recipe information is saved in the recipe information storage means 17 as the optical condition update process in step S25, and the contents of the analysis data list 51 and the analysis image are stored. After the data is stored in the inspection result information storage unit 14, the GUI 61 is closed and the process is terminated.

After the above operation is performed, the determination of step S23 is performed at the latest timing 56 (defect / specified number of wafer processes) stored after the defect detection of the wafer W.
(Modification of the first embodiment)
In the first embodiment described above, the optical conditions of the stage rotation image capturing and the camera swing imaging are the same. However, these are not necessarily the same. For example, the setting of the polarizing plate 34 is Different settings may be used, such as turning off for stage rotation imaging and turning on for camera swing imaging.

Further, the line illumination 11 and the illumination 12 as shown in FIG. 2 are not configured to have a plurality of illumination means at fixed positions, but can be configured to rotate as in the case of camera driving, and the camera angle and illumination. The rotation may be performed with the angle always maintaining the position of the specularly reflected light with respect to the image plane. As a result, the analysis data acquisition process in step S24 and the optical condition update process in step S25 are executed by operating the GUI 61 in the same manner as diffracted light and scattered light without fixing the camera angle even in the case of regular reflection light. Can do.
(Second Embodiment)
FIG. 5 is a diagram showing an outline of an appearance inspection apparatus according to the second embodiment to which the present invention is applied.

  As shown in FIG. 5, the appearance inspection apparatus 1a according to the second embodiment includes an inspection unit 2 included in the appearance inspection apparatus 1 according to the first embodiment and the appearance inspection apparatus 1 according to the first embodiment. A control unit 3a and a data processing unit 62 connected via a network are provided instead of the control unit 3 included in the network.

  The control unit 3a includes an analysis data acquisition unit 13 and a defect extraction unit 15. The data processing unit 62 includes an inspection information storage unit 14, an optical condition update unit 16, and a recipe information storage unit 17. . The data processing unit 62 is a control computer having a central processing unit, an internal memory, an external storage device, and the like, like the control unit 3.

FIG. 6 is a diagram illustrating an example of a GUI provided by the optical condition update unit 16 according to the second embodiment.
The GUI 63 of the optical condition update unit 16 in the second embodiment shown in FIG. 6 includes a stage rotation imaging unit 18 existing in the GUI 61 of the optical condition update unit 16 in the first embodiment shown in FIG. There is no camera swing imaging unit 19. Further, the GUI 63 can display the analysis data stored in the inspection information storage unit 14, so that the recipe can be updated by an easy operation. In order to automatically acquire analysis data, it is necessary to set the data shown in FIG.

Next, the operation of the second embodiment will be described.
FIG. 7 is a diagram showing a data list necessary for automating the analysis data acquisition process.
In the second embodiment, every time the defect extraction process in step S22 in the first embodiment described with reference to FIG. 4 is completed, it is confirmed whether the analysis data acquisition start condition 64 in FIG. .

  When the analysis data acquisition start condition 64 is the defective wafer detection 65, necessary inspection information is read from the inspection information storage unit 14, and the number of times of fail (failure) in the pass / fail determination is calculated. Then, when the calculated number is equal to or larger than the set value of the number of wafers 67, the analysis data acquisition process in step S24 is started. On the other hand, when the analysis data acquisition start condition 64 is the designated wafer processing 66, the number of inspection images stored in the inspection information storage unit 14 is read, and the number of inspection images read out is equal to or larger than the set value of the number of wafers 67. In this case, the analysis data acquisition process in step S24 is started.

  In the analysis data acquisition process in step S24, first, each piece of information is collected according to preset contents. This is determined by the optical condition selection number 68 of FIG. For example, when collecting regular reflection light and scattered light, the optical condition selection number 68 is set to 2, regular reflection light is set to each condition setting 69 [1], and scattered light is set to each condition setting 69 [2]. Keep it.

  Next, the camera angle when performing stage rotation imaging is set according to the camera angle selection number 70 and the camera angle setting 71 of the stage rotation imaging number in FIG. For example, when collecting regular reflection light and scattered light, 2 is set in the camera angle setting 71 (i = 2). For regular reflection light, the camera angle is fixed, so the camera angle selection number 70 is always set to 1, and if you want to capture scattered light at camera angles of 40 degrees and 70 degrees, the value of the camera angle selection number 70 Is set to 2, camera angle setting 71 [1] is set to 40.0, and camera angle setting 72 [2] is set to 70.0.

  Furthermore, a stage rotation image can be acquired by setting the illumination light quantity setting 72, the band pass filter setting 73, and the polarizing plate setting 74 shown in FIG. 7 in advance. The captured stage rotation image is stored in the stage rotation image data 75.

  Further, line profile data 37 similar to that in the first embodiment described above is generated based on this stage rotation image, and the peak of the luminance value is detected. The peak detection method is preset to the peak detection rule 76 and the maximum number 77 of peak detections. When the detection rule 76 is the minimum peak 79 and the maximum number 77 of peak detections is 10, the rotational stage angle amounts of the 10 peak positions are stored in the stage rotation angle 80 in order of increasing luminance value from the detected minimum peak. To do. Similarly, when the detection rule 76 is the maximum peak 78 and the maximum number 77 is set to 7, the rotation stage angle amount of the 7 peak positions from the found maximum peak in descending order of the luminance value is rotated by the stage. Store at angle 80.

  Further, the camera swing imaging can be performed by previously setting the light amount setting 81, the filter setting 82, and the polarizing plate setting 83 shown in FIG. However, only when each condition setting 69 is specular reflection light, camera swing imaging is not performed, and the camera swing angle of the analysis image is fixed. On the other hand, when the condition setting 69 is diffracted light or scattered light other than the regular reflection light, that is, the camera angle image is stored in the camera swing image data 84.

  Furthermore, line profile data 37 similar to that in the first embodiment described above is generated based on the camera swing image, and the peak of the luminance value is detected. Since the peak detection method is the same as in the case of stage rotation imaging, description thereof is omitted.

  By previously setting the stage rotation angle 85, the camera swing angle 86, the illumination light amount setting 87, the band pass filter setting 88, the polarizing plate setting 89, and the target image luminance value 90 obtained as described above, The image after the light amount automatic adjustment processing similar to that of the first embodiment can be stored in the captured image data 91 for analysis.

The analysis data can be automatically acquired as described above.
Next, in the recipe optical condition update unit 20a included in the GUI 63 of the optical condition update unit 16 shown in FIG. The setting button 94 and the setting button 95 are used for reading. Then, the wafer map display screen 92 on which the respective captured images for analysis are displayed and the thumbnails 120 of the wafer W designated on the wafer map display screen 93 are displayed by the number of captured images.

  Further, the stage rotation image 121 and the camera swing image 96 stored when the analysis captured image is captured are respectively displayed. When the thumbnail 120 is clicked (instructed), the stage rotation angle position of the selected image is displayed as a line 97 on the stage rotation image 121 and the camera angle is displayed as a line 98 overlaid on the camera swing image 96. Furthermore, the display can be switched from the thumbnail 120 by designating the histogram display 99.

  Using such a GUI 63, for example, the thumbnails 120 of defective chips and non-defective chips are compared, and a condition with a large difference in luminance value is set as a condition with higher defect detection sensitivity.

As a result, it is possible to more easily determine a good optical condition for defect detection than the GUI 61 described in the first embodiment.
Finally, the best conditions can be registered with the add button 54a and the delete button 55a for the optical conditions of the current recipe, and the work can be ended with the recipe save & end button 60a.
(Modification of the second embodiment)
In the second embodiment described above, the analysis image included in the analysis captured image data 91 is compared with two images. However, the present invention is not limited to this, and three or more analysis capture images are used. Images may be compared simultaneously.

  Furthermore, a function may be added in which the defect detection result (information such as the found defect position, defect name, and individual size) of the analysis image to be compared is displayed on the wafer map. In this case, the position for displaying the wafer W is limited to display only the wafer W in which a defect is found, or the difference between the comparison images is calculated for all the pixels of each wafer W included in the analysis imaged image. A function for restricting display selection to only the wafers W having a predetermined difference or more may be added.

  Further, before obtaining the maximum peak from the line profile generated based on the stage rotation captured image or the camera swing image, the minute fluctuation noise may be cut from the line profile by, for example, a noise cut filter. Further, a process of rounding the obtained peak to a specified angle interval (30 degrees if the peak angle is 33.4 degrees and the rounding angle interval is 10 degrees) may be added.

Further, a threshold value for the luminance value of the detected peak may be provided, and processing such as counting only the peak having the luminance value equal to or higher than the threshold value in the case of the maximum peak and the luminance value not exceeding the threshold value in the case of the minimum peak may be added. .
(Third embodiment)
Since the appearance inspection apparatus according to the third embodiment to which the present invention is applied has the same configuration as the appearance inspection apparatus 1a according to the second embodiment, the description thereof is omitted.

The operation of the third embodiment will be described.
FIG. 8 is a diagram showing a data list necessary for automating the optical condition update process.
In the third embodiment to which the present invention is applied, it is confirmed whether or not the tuning start condition 100 in FIG. 8 is met every time the defect detection process in step S22 is completed in the second embodiment described above. To do. Therefore, in the third embodiment, it is necessary to store data as shown in FIG. 8 in advance in the recipe information storage means 17 in the second embodiment described above.

  First, when the tuning start condition 100 in FIG. 8 is the defective wafer detection 101, necessary inspection information is read from the inspection information storage unit 14, and the number of times of fail in pass / fail determination is calculated. Then, when the calculated number of times is equal to or greater than the set number of wafers 103, the optical condition update process in step S25 is started. On the other hand, when the tuning start condition 100 is 102 for each designated wafer process, the number of inspection images stored in the inspection information storage unit 14 is read, and when the number of inspection images read is equal to or greater than the set number of wafers 103, The optical condition update process in step S25 is started.

  Next, the analysis image corresponding to the setting of the optical condition selection 104 and the optical condition data obtained by capturing the analysis image are read from the inspection information storage unit 14, and the analysis image is read according to the optical condition selection rule 105 described later. Perform prioritization.

  If the SN descending order 106 is set in the optical condition selection rule 105, the condition candidates are ranked in descending order of the sum of deviations from the average value of the analysis images stored in the examination information storage means 14 in the previous examinations. To do. This descending ordering setting is effective when the optical condition selection 104 is regular reflection light and diffracted light with respect to the wafer W with a pattern.

  Further, when the SN ascending order 107 is set in the optical condition selection rule 105, ranking is executed for each analysis image in ascending order of deviation sum from the average value. This ascending ordering setting is effective when the bare wafer W has no pattern and the optical condition selection 104 is regular reflection light and diffracted light, or scattered light.

  Further, when the optical condition selection rule 105 is set in the descending order 108 with respect to the recipe image, the non-defective product image included in the recipe and the analysis image stored in the inspection information storage unit 14 in the previous inspection are displayed. In the meantime, the luminance difference between the pixels at the same position is calculated for all the pixels, and the difference is ranked in descending order of the average value. The ranking setting of the luminance difference descending order does not depend on the characteristics of the wafer W and the setting of the optical conditions.

  Further, when the optical condition selection rule 105 is set to 109 in descending order of the number of fail wafers, the defect extraction process in step S22 is executed on the analysis image, and the number of wafers that have failed for each condition is counted. And rank them in descending order. The ranking setting in the descending order of the number of fail wafers is also a setting that does not depend on the characteristics of the wafer W and the optical condition setting, as in the case of the recipe image and the luminance difference descending order.

Then, the optical conditions of the recipe are updated in descending order of the condition candidates determined as described above.
(Modification of the third embodiment)
FIG. 9 is a diagram showing a data list necessary for automating the optical condition update processing in the modification of the third embodiment.

  In the third embodiment described above, since the selection rule of the optical condition is a predetermined setting, the selection rule is the same when the optical condition update process in step S25 is executed. As shown in FIG. 9, the first optical condition selection rule 105a, the number of times 110 the selection rule is changed by the total number of processed wafers, and the total number of wafer processing 111 for starting the application of the rules are newly set and inspected. The first optical condition selection rule 105a can be changed to the optical condition selection rule 105b depending on the number of wafers.

  Thus, for example, in the first inspection, the optical conditions are selected in the SN ascending order 107a or SN descending order 106a, and in the inspection when the number of inspection wafers is 10, the recipe image and the luminance difference descending order 108b are optically selected. When the conditions are selected and inspection is performed with the number of inspection wafers being 100, the optical conditions can be selected in descending order 109b of the number of fail wafers. Thereby, ranking with higher accuracy can be performed.

  Further, in the above-described third embodiment, when the optical wafer selection rule 105 sets the fail wafer number descending order 109, the target analysis image is executed when the optical condition update process in step S25 is executed. However, the optical condition selection rule 105 is determined in advance. Therefore, for example, if the defect extraction process in step S22 is performed every time the inspection process is performed, the time for the optical condition update process in step S25 can be shortened.

  Further, the optical condition selection rules 105, 105a, and 105b are not limited to the method of the example described above. For example, instead of the fail wafer number descending order 109, 109a, and 109b, the total of the detected defect areas is calculated. Any method may be used as long as it can rank the candidate candidates for the condition such as descending order so that the appearance inspection apparatus 1a can detect defects more.

As mentioned above, although each embodiment to which this invention is applied has been described, it is not limited to these.
For example, the optical conditions for stage rotation imaging and camera swing imaging are the same, but they are not necessarily the same. For example, in stage rotation imaging, the polarizing plate setting is off, and camera swing imaging is performed. These may be different settings, such as the polarizing plate setting being on.

  Further, like the line illumination 11 and the illumination 12 shown in FIG. 2, these are not a configuration in which a plurality of these are fixed at a specific position, but a configuration that can be rotated like a camera drive, and a camera angle and illumination. The angle may be rotated in a state where the regular reflection position is always maintained with respect to the image plane. By doing so, the analysis data acquisition process in step S24 and the optical condition update process in step S25 may be performed without fixing the camera angle even in the case of regular reflection light.

  That is, the present invention is not limited to the above-described embodiments and the like, and can take various configurations or shapes without departing from the gist of the present invention.

W Semiconductor wafer (wafer)
DESCRIPTION OF SYMBOLS 1, 1a Appearance inspection apparatus 2 Inspection | inspection part 3, 3a Control part 4 Scan stage 5 Rotation stage 6 Mounting part 7 Imaging means 8 Line sensor camera 9 Cylindrical lens 10 Rotating means 11 Line illumination 12 Illumination 13 Analysis data acquisition means 14 Inspection Information storage means 15 Defect extraction means 16 Optical condition update means 17 Recipe information storage means 18 Stage rotation imaging part 19 Camera swing imaging part 20, 20a Recipe optical condition update part 26 Measurement light condition 27 Regular reflection light 28 Scattered light 29 Diffracted light 30 Band pass filter 31 Camera angle 32 Imaging start button 33 Setting angle 34 Polarizing plate 35 Imaging start button 36, 36a Display screen 37, 37a Line profile data 38, 38a Scroll bar 39, 39a Line 40 Selection angle 41, 41a Peak 42, 2a peak 43,43a list 44,44a add button 45,45a delete button of the list in the selected item 46 stage rotation angle condition determination button 47 list ID
48 List ID
49 Target brightness value 50 Camera swing angle setting confirmation → Automatic light quantity adjustment → Analysis data acquisition information save button 51 Illumination light quantity value 52 Data list 53 Optical condition list 54, 54a Add button 55, 55a Delete button 56 Timing 57 Detect defective wafer Hours 58 Specified processing wafer interval 59 Text box 60, 60a Recipe save & end button 61 GUI
62 Data processor 63 GUI
64 Analysis data acquisition start conditions 65 Defective wafer detection 66 Designated wafer processing 67 Number of wafers 68 Optical condition selection number 69 Each condition setting 70 Camera angle selection number 71 Camera angle setting 72 Illumination light quantity setting 73 Band pass filter setting 74 Polarizing plate setting 75 Stage Rotation image data 76, 76a Peak detection rule 77, 77a Maximum number of peak detections 78 Maximum peak 79 Minimum peak 80 Stage rotation angle 81 Light amount setting 82 Filter setting 83 Polarizing plate setting 84 Camera swing image data 85 Stage rotation angle 86 Camera swing Moving angle 87 Illumination light quantity setting 88 Band pass filter setting 89 Polarizing plate setting 90 Target image luminance value 91 Image data for analysis 92, 93 Wafer map display screen 94, 95 Setting button 96 Camera swing image 97, 98 lines 99 Histogram display 100 Tuning start condition 101, 101a Detect defective wafer 102, 102a Each specified wafer processing 103 Number of wafers 104, 104a Optical condition selection 105, 105b Optical condition selection rule 105a First optical condition selection rule 106, 106a, 106b SN descending order 107, 107a, 107b SN ascending order 108, 108a, 108b Descending order of brightness difference from recipe image 109, 109a, 109b Fail Number of wafers descending order 110 Number of times the selection rule is changed based on the total number of processed wafers 111 Total number of wafer processes starting to apply rules 120 Thumbnail 121 Stage rotation image

Claims (9)

A recipe update method for updating the recipe in an appearance inspection apparatus for inspecting a semiconductor substrate based on a recipe,
According to the optical conditions stored in the recipe storage means, the semiconductor substrate is imaged to obtain a first image,
Detecting whether a defect exists in the acquired first image;
If it detects that the defect exists, change the optical condition,
According to the changed optical conditions, the semiconductor substrate is imaged to obtain a second image,
Updating the optical conditions stored in the recipe storage means to the changed optical conditions;
A recipe update method characterized by that. Determining whether a more prominent defect appears in the second image compared to the defect appearing in the first image;
When it is determined that a defect more prominent than the defect appearing in the first image appears in the second image, the optical condition stored in the recipe storage means is updated to the changed optical condition. To
The recipe update method according to claim 1, wherein: The second image is an image obtained by capturing an image of a semiconductor substrate in which the defect is detected.
The recipe updating method according to claim 1, wherein the recipe is updated. The second image is an image obtained by imaging another semiconductor substrate of the same type as the semiconductor substrate that has detected the presence of the defect.
The recipe updating method according to claim 1, wherein the recipe is updated. When the appearance inspection apparatus sequentially inspects the semiconductor substrate, it detects whether or not there is a defect in the acquired first image every predetermined number of times.
The recipe update method according to any one of claims 1 to 4, wherein: When the appearance inspection apparatus sequentially inspects the semiconductor substrate, the optical condition is changed every predetermined number of times.
The recipe update method according to claim 1, wherein the recipe is updated. If it is detected that the defect exists, a plurality of the optical conditions are changed,
When it is determined that a defect more prominent than the defect appearing in the first image appears in the second image, the optical condition stored in the recipe storage means is changed from the changed optical condition Update to optical conditions where defects are most prominent
The recipe update method according to claim 1, wherein the recipe is updated. If the presence of the defect is detected, change the optical condition using a graphical user interface;
The recipe update method according to claim 1, wherein the recipe is updated. A recipe update system that updates the recipe in an appearance inspection apparatus that inspects a semiconductor substrate based on a recipe,
Recipe storage means storing optical conditions;
First image acquisition means for imaging the semiconductor substrate and acquiring a first image in accordance with the optical conditions stored in the recipe storage means;
Defect detection means for detecting whether or not a defect exists in the first image acquired by the first image acquisition means;
An optical condition changing means for changing the optical condition stored in the recipe storing means when the defect detecting means detects the presence of a defect;
In accordance with the optical condition changed by the optical condition changing means, second image acquisition means for acquiring a second image by imaging a semiconductor substrate that has been detected to have the defect;
It is determined whether or not a more prominent defect appears in the second image acquired by the second image acquisition means than the defect that appears in the first image acquired by the first image acquisition means. A judgment means to
If it is determined by the determining means that a more significant defect appears in the second image than the defect appearing in the first image, the optical condition stored in the recipe storing means is changed. Optical condition updating means for updating to the optical condition,
A recipe update system comprising: