JP2013148349A - Review method and review device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate observation of a sample by reducing the man-hour of a user, that is required for a coordinate conversion between a coordinate system of inspection data input from an external inspection device and a coordinate system of a defect review device concerning a defect review method and defect review device using an electron microscope.SOLUTION: A capturing of a dark field optical microscopic image is repeated multiple times in order to obtain a dark field optical microscopic image optimum to calculate the barycentric coordinate of a defect and fine alignment is performed concerning a defect to be a review object on the basis of a sorting method with previously registered defect sizes and optical filter selection method of a dark field optical microscope.

Description

  The present invention relates to a review method and a review apparatus for reviewing defects generated in a manufacturing process of a thin film device such as a semiconductor electronic circuit board and a liquid crystal display board using an enlarged imaging device such as a scanning electron microscope.

  The manufacturing process of thin film devices such as semiconductors, liquid crystal displays, and hard disk magnetic heads is composed of a number of processes.

  The number of pattern processing steps sometimes reaches several hundreds. If an appearance abnormality such as a foreign object or a broken wiring pattern occurs on the thin film device due to inadequate or abnormal manufacturing conditions of the processing apparatus, the probability that a product will be defective increases and the yield decreases. Therefore, it is important to identify the device in which the problem has occurred and take measures to maintain and improve yield. For this reason, inspections such as foreign matter inspection and appearance inspection are performed for each main process, and whether or not the processing is normally performed is monitored. At this time, since it is impossible to inspect all the substrates to be processed for each processing process due to time and labor constraints, it is usually a lot unit or a substrate unit for each series of processes. Alternatively, an inspection is performed on a substrate to be processed sampled by a combination thereof. Here, the substrate to be processed means a minimum unit for product processing, and in the case of a semiconductor, it refers to one wafer.

  In the inspection apparatus for performing the foreign substance inspection or the visual inspection, a scanning electron microscope or an optical microscope, for example, a wafer surface is scanned with a laser, and the presence or absence of scattered light is detected to obtain information on the position and number of foreign substances. Also, when performing defect inspection to detect both foreign matter and pattern abnormalities, for example, by capturing an image of the circuit pattern of the wafer with an optical enlargement imaging device, and comparing it with an image of the other similar pattern region in the vicinity, Get information about the position and number of singularities. Here, the “singular point” refers to a point that is output as a point where an abnormality has been found by the inspection of the inspection apparatus. The foreign object and the appearance abnormality are combined and hereinafter referred to as “defect”.

  The determination of the process abnormality is often performed using the number and density of defects detected by the foreign matter inspection or the appearance inspection as a management index. If the number or density of defects exceeds a preset reference value, it is determined that an abnormality has occurred in the apparatus. Based on the defect coordinate information detected by the inspection apparatus, the defect is detected using an optical microscope or scanning electron microscope (Scanning Electron Microscope). Microscope (hereinafter referred to as “SEM”) etc. is used to magnify images with a review device to obtain detailed information on the size, shape, texture, etc. of defects, detailed inspections such as elemental analysis, cross-sectional observation, etc. Identify the device and malfunction details. Then, based on the result, measures are taken for manufacturing apparatuses and processes to prevent a decrease in manufacturing yield. In recent years, the importance of SEM-type review devices has increased due to the need for finer inspection target patterns and the detection of fine foreign matter.

  The review device is a function that automatically acquires an enlarged image of a foreign object or defect based on inspection data acquired by a foreign object inspection device or an appearance inspection device in response to a request for automation and efficiency of defect review work (Automatic Defect Review). , Hereinafter referred to as ADR). However, in the case of an SEM-type review apparatus, if the defect position information acquired by the foreign substance inspection apparatus or the appearance inspection apparatus is used as it is as the image pickup position information of the review apparatus, the probability that the defect is out of the imaging field of view of the SEM increases greatly. . This is because the SEM type review device has a large enlargement magnification, so that a slight deviation between the coordinate system of the foreign substance inspection device or the appearance inspection device and the coordinate system of the review device is enlarged as it is.

  In order to prevent this, it is possible to generate a coordinate conversion formula for converting defect position information possessed by the foreign substance inspection apparatus and appearance inspection apparatus into defect position information for the defect review apparatus using position information of an appropriate reference point on the sample. Done. Patent Document 1 discloses such coordinate conversion.

  Furthermore, since this alone cannot ensure the accuracy of the defect position to be finally acquired, a defect image having a wide field of view including the defect position and a reference image having substantially the same size as the defect image are acquired, and the defect image and the reference image are image-processed. By doing so, imaging is performed with a two-step visual field size in which the center position of the defect is calculated and the visual field center of the narrow-field image to be finally acquired is calculated. Patent Document 2 discloses a method for acquiring a defect image with the above two-step visual field sizes.

  In Patent Document 3, a defect image is obtained for an appropriate standard sample using an appearance inspection device and a review device, and the direction and size of visual field deviation are measured from the difference in the defect center coordinates obtained by both. Then, the systematic error is removed from the obtained field deviation distribution, and the field size of the wide-field defect image when performing imaging with two stages of field sizes is determined from the direction and size of the remaining random field deviation. The invention is disclosed.

Japanese Patent Laid-Open No. 2002-039959 JP 2000-30652 A JP 2006-145269 A

  According to Patent Document 1, an appropriate defect point on a sample that is a defect review target is sampled, and a coordinate conversion rule is generated using both the appearance / foreign particle inspection apparatus and the defect review apparatus. Since the obtained coordinate transformation rules vary depending on which defect point is selected, considerable experience and knowledge are required to perform reasonable sampling. When the defect review apparatus is of the SEM type, it is substantially meaningless from the problem of throughput to generate the coordinate transformation rule using all defect points on the sample.

  According to Patent Document 3, a coordinate conversion rule is generated using an appropriate standard sample, but it is uncertain whether or not the obtained coordinate conversion rule can be applied to any sample.

  An object of the present invention is to reduce the number of steps required by a user for defect coordinate conversion in an SEM type defect review apparatus, to enable easy observation of a sample regardless of the skill of an operator, and to capture defects. It is to provide a review method and a review apparatus that improve the rate.

  In the present invention, in the SEM type defect review apparatus, the defect position is acquired by an external inspection apparatus such as an appearance inspection apparatus or a foreign substance inspection apparatus using an image signal detected by the optical microscope. It is characterized in that a coordinate conversion rule for information is generated. Here, the coordinate conversion rule is a general term for calculation rules necessary for performing coordinate conversion, such as a coordinate conversion formula and a coordinate offset value.

  As the optical microscope, a dark field microscope capable of acquiring defect position information at high speed is suitable. However, in the case of an optical microscope, an image signal that is not appropriate for acquiring a defect position may be detected due to factors such as halation and bright spot generation. For this reason, a determination unit for determining the quality of the image signal detected by the optical microscope is provided to select inappropriate data for generating the coordinate conversion rule.

  Further, for data determined to be inappropriate by the determination unit, the image acquisition condition of the optical microscope is appropriately adjusted, and the image signal is acquired again.

  Since the position information used for generating the coordinate conversion rule is acquired using an optical microscope instead of the SEM, the number of defect points that can be used for generating the coordinate conversion rule can be significantly increased as compared with the conventional case. Therefore, the conversion accuracy of the coordinate conversion rule is improved and the defect supplement rate is improved. Further, since it becomes unnecessary to sample the defect point, it becomes possible to set the review condition of the defect review apparatus regardless of the skill of the operator.

  Further, since the image signal is re-acquired, the number of defect points used for generating the coordinate conversion rule can be further increased as compared with the conventional case.

The system block diagram which shows the connection of the apparatus in a part of semiconductor device manufacturing line. Schematic of the SEM type review apparatus of Example 1 or 2. FIG. 5 is a flowchart showing the operation of the SEM type review device according to the first or second embodiment. FIG. 4 is a contrast diagram of dark-field microscope images obtained when an optical filter is not applied and when an optical filter is applied. FIG. 4 is a contrast diagram of dark-field microscope images obtained when an optical filter is not applied and when an optical filter is applied. The figure which shows an example of the parameter setting screen for the selection by defect size. The figure which shows an example of the setting screens, such as a filter applied in optical condition change, and the number of luminescent spots, the number of trials.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Example 1
In the present embodiment, a configuration example of an SEM type defect review apparatus having a function of generating a coordinate conversion rule using an output signal of an optical microscope at the time of ADR execution will be described.

  First, the schematic diagram of the manufacturing line of the semiconductor device in which the defect review apparatus of a present Example is installed is shown in FIG. In the semiconductor device manufacturing line, as shown in FIG. 1, a semiconductor device manufacturing apparatus 2, an inspection apparatus 3, a review apparatus 4, an analysis apparatus 5, and a review analysis apparatus 6 are connected to each other by a data management server 1 and a network 7. The structure is made. Since semiconductor devices are manufactured by forming a large number of layers on a semiconductor substrate, the number of manufacturing lines shown in FIG. 1 is actually set according to the number of semiconductor device manufacturing processes.

  The manufacturing apparatus 2 is an apparatus used for manufacturing semiconductor devices such as an exposure apparatus and an etching apparatus. The inspection apparatus 3 is an apparatus that inspects whether or not a defect has occurred in the semiconductor device processed by the manufacturing apparatus 2 and whether or not there is a foreign substance, and detects the position of the defect or the foreign substance. A method of scanning a beam spot of light or an electron beam on a semiconductor wafer and specifying a defect position from the degree of diffuse reflection and potential contrast due to secondary electrons, and acquiring images of the formed pattern from two chips, respectively. A method is known in which these images are compared to determine a different portion as a defect, and the defect position is detected. Inspection information such as the coordinates of defects detected by the inspection apparatus 3 is sent to the review apparatus 4 via the network 7.

  The review device 4 is a device for observing defects based on the inspection information of the inspection device 3. The review device 4 moves the stage on which the semiconductor wafer is mounted, and on the semiconductor wafer based on the defect position information output from the inspection device 3. The defect to be targeted is positioned and the defect is observed. The observation includes a method of imaging with an optical microscope and a method of imaging with an electron microscope. In this embodiment, an example using a scanning electron microscope SEM is shown. The analysis device 5 performs elemental analysis using, for example, EDX that detects X-rays or Auger electron spectroscopy. Auger electron spectroscopy is a method of detecting and analyzing Auger electrons emitted from an object when the object is irradiated with an electron beam. The review analysis device 6 is a device that can perform defect observation and elemental analysis with a single device. The analysis device 5 or the review analysis device 6 can also position a defect based on the defect position information output from the inspection device 3 and perform a review or analysis.

  In the example of FIG. 1, each device is described according to its function. However, one device may have various functions. For example, the function of the inspection device 3 may be added to the manufacturing device 2 or the inspection device 3 and the review device 4 may be integrated to have a respective function.

  The data management server 1 is a device that manages data obtained by the inspection device 3, the review device 4, the analysis device 5, and the review analysis device 6. The review device 4 and the analysis device 5 use the data management server 1. Thus, information such as defect position coordinates output from the inspection apparatus 3 can be acquired. Note that the review device 4 can be provided with all or some of the functions of the data management server 1.

  In FIG. 1, each device is connected via the network 7 and exchanges data. However, it is not always necessary to pass through the network 7, and data can be exchanged between the devices. It may be connected.

  FIG. 2 is a longitudinal sectional view showing a schematic configuration of the SEM review device of the present embodiment. The semiconductor wafer WF to be reviewed is mounted on the XY stage 15. The XY stage 15 is controlled to move in the X direction or the Y direction by the stage control unit 23 based on a control signal sent from an overall control unit 19 having a plurality or a single microprocessor. The XY stage 15 can also be controlled in the height direction. The imaging device 8 to which the SEM is applied captures an enlarged image of a predetermined area of the semiconductor wafer WF fixed to the XY stage 15.

  The electron source 9 that is the source of the electron beam EB of the imaging device 8 is applied with a constant controlled voltage from the high voltage stabilization power supply 24. The electron beam EB emitted from the electron source 9 is converged and narrowed by the first condenser lens 10, the second condenser lens 11, the first objective lens 13, and the second objective lens 14, and is deflected by the deflection scanning coil 12. The semiconductor wafer WF is scanned. By irradiation of the semiconductor wafer WF to be measured with the electron beam EB, secondary signals such as secondary electrons and reflected electrons are generated from the semiconductor wafer WF and detected by the signal detector 25. A set of various optical elements such as an electron source and an objective lens described above constitutes an electron optical system, and is held in a vacuum container schematically illustrated as an outline of the imaging device 8.

  The output signal of the signal detector 25 is processed by an A / D conversion unit 21 that converts an analog signal into a digital signal, and is input to an image calculation unit 20 having a microprocessor. The image calculation unit 20 performs image processing such as generation of an SEM image of the semiconductor wafer WF and defect detection processing. The image processing result by the image calculation unit 20 is displayed on the display 17 via the overall control unit 19. A storage device 16 is connected to the overall control unit 19 and stores an SEM image and accompanying data. On the display 17, a GUI (Graphical User's Interface) that functions as an operation screen of the review apparatus is displayed. A user who uses the review apparatus inputs information necessary for inspection, such as defect observation conditions, to the GUI on the display 17 using the input device 18 configured by a keyboard, a pointing device, or the like.

  The defect coordinate data of the semiconductor wafer WF is sent from the inspection apparatus 3 shown in FIG. 1 to the overall control unit 19 via the network 7 and stored in the storage device 16. Based on the defect coordinate data, the overall control unit 19 sends a stage movement command to the stage control unit 23 so that the defect enters the field of view of the imaging device 8, and moves the XY stage 15.

  Since the SEM is imaged at a high magnification, the field of view is narrow, and the defect coordinates based on the inspection data input from an external inspection apparatus are used as the defect coordinates on the SEM type defect review apparatus in order to capture the defect so that it is within the field of view. It is necessary to perform coordinate conversion (fine alignment) accurately. Therefore, the imaging apparatus 8 of the present embodiment is provided with a dark field optical microscope 27, and based on the defect coordinate data of the inspection data input from the external inspection apparatus, the coordinates at which the defect to be reviewed exists are coordinates. Irradiate a laser to take a dark field optical microscope image. The movement of the defect position to the laser irradiation position is performed by moving the stage.

  The dark field optical microscope 27 irradiates the observation target with the laser emitted from the laser oscillation unit 27-1 via the ND filter 27-2 and the wavelength filter 27-3, and generates laser scattered light generated in the observation target. To detect. When the dark field optical microscope 27 detects the laser scattered light, the laser scattered light is incident on the objective lens 27-4, and then the optical path is changed by the mirror unit 27-5. The scattered light bent by the mirror unit enters the TV camera unit 27-6 and is imaged by the detection element of the TV camera unit. The dark field optical microscope 27 is equipped with a polarizing filter 27-7, and the signal amount of the input laser scattered light can be adjusted by inserting and removing the polarizing filter 27-7. On the other hand, by selectively changing the settings of the ND filter 27-2 and the wavelength filter 27-3, the amount of laser light to be emitted can be adjusted, and the surface of the observation target is irradiated with the laser whose light amount has been adjusted. The laser scattered light generated from the defect on the surface to be observed is captured by the dark field optical microscope 27 to obtain a dark field microscope image. Note that only one or two of the three optical filters, the ND filter, the wavelength filter, and the deflection filter described above, may be provided. In some cases, four or more optical filters may be provided.

  The output signal of the dark field type optical microscope is quantized by the A / D conversion unit 21 in the same manner as the output signal of the detector 25, and then input to the image calculation unit 20, where various image processes are performed. The coordinates of the center of gravity of the scattered light generated at the defect portion are calculated as defect coordinates on the SEM type defect review apparatus, and fine alignment is performed. By executing this operation for all defects to be reviewed, defects on the semiconductor device can be accurately captured and reviewed (imaged) by the SEM type defect review apparatus.

  The overall control unit 19 controls the electron optical system control unit 22-1, the dark field optical microscope control unit 22-2, and the stage control unit 23 in accordance with the input information and a control program stored in advance. The electron optical system control unit 22-1, dark field optical microscope control unit 22-2, and stage control unit 23 are in accordance with instructions from the overall control unit 19, the electron optical system, dark field optical microscope, or XY stage in the imaging device 8. Control 15 is performed. Furthermore, the overall control unit 19 includes functional blocks of a defect selection unit 26-1, a DFOM (Dark Field Optical Microscope) image determination unit 26-2, and a coordinate transformation calculation unit 26-3, which will be described later. .

  Next, the operation of the defect review apparatus of this embodiment will be described with reference to FIG. FIG. 3 is a flowchart for explaining a procedure for performing a defect review. The fine alignment, which is a feature of the defect review apparatus according to the present embodiment, is described in detail.

  When the semiconductor wafer to be observed is loaded into the imaging device 8 (STEP 100), the overall control unit 19 communicates with the inspection device 3 or the data management server 1 shown in FIG. 1 and reads the inspection data (STEP 101). The inspection data is configured such that defect feature amount information such as defect size and defect position information are associated with an appropriate defect ID assigned to a detected defect.

  After execution of STEP 101, the overall control unit 19 executes wafer alignment for roughly correcting an error between the coordinate system of the SEM type defect review apparatus and the coordinate system of the semiconductor wafer (STEP 102). Wafer alignment is an alignment process between the semiconductor wafer and the coordinate system of the SEM type defect review apparatus, and an image of an appropriate pattern whose position is known among the semiconductor patterns formed on the semiconductor wafer is imaged. Then, it is executed by determining the coordinates of the target pattern from the control information of the XY stage 15 and comparing the determined coordinates with the known position information. As the alignment pattern, a characteristic pattern having no similar pattern in the vicinity thereof is used. When executing wafer alignment, an apparatus user first designates an appropriate alignment pattern on the GUI. The overall control unit 19 instructs the electron optical system control unit 22-1 and the stage control unit 23 to image the designated alignment pattern, and obtains the position of the imaged alignment pattern and the known position information of each alignment pattern. The comparison is made, and the average value of the shift amounts is stored in the storage device 16 as the offset amount of the coordinate origin.

  By executing the wafer alignment, it is possible to calculate the correction amount for the rotational deviation of the coordinates that occurs when the semiconductor wafer is loaded into the imaging device 8. Further, when a defect review is performed on a bare wafer before forming a semiconductor pattern, an edge portion of the outer peripheral portion and the notch portion of the semiconductor wafer is designated and an error between the coordinates of the semiconductor wafer sent from the inspection apparatus 3 is obtained. Can also be corrected. In wafer alignment, an optical microscope or SEM may be used.

  After the wafer alignment is performed, fine alignment using a dark field optical microscope is performed. The fine alignment process of the present embodiment is characterized by including a redoing sequence of imaging in which the optical conditions of the dark field optical microscope are changed. This will be described in detail below with reference to FIG.

  First, the defect sorting unit 26-1 in the overall control unit 19 sorts the inspection data read in STEP 101 using a preset size threshold of the defect size (STEP 103). This is because if the defect size of the defect to be reviewed is large, the brightness of the scattered light due to laser irradiation increases, halation is likely to occur, and the center of gravity coordinates of the defect can be accurately calculated using a dark-field optical microscope image. This is because it becomes difficult. This process is executed by the overall control unit 19. When executing STEP 103, defect data may be sorted by sorting defect data in order of defect size.

  The filter threshold value for the defect size used in STEP 103 is set via a parameter setting screen as shown in FIG. 6, for example. The apparatus user calls the parameter setting screen of FIG. 6 to the GUI on the display 17 when setting the inspection recipe before executing ADR, and uses it in STEP 103 based on the defect size of each defect in the inspection data input from the external inspection apparatus. The defect size threshold value is input to the input field at the center of the screen, and the “OK” button is pressed to set the defect size filter threshold value. The set threshold information is stored in the storage device 16 and is referred to by the overall control unit 19.

  When the defect size selection is completed, the overall control unit 19 captures an image of the defect determined to be the defect size “large” in STEP 103 in accordance with a built-in control program (STEP 104). Specifically, the positional information of the selected defect is transmitted to the stage control unit 23 to execute visual field movement, and an imaging instruction is given to the dark field optical microscope control unit 22-2. The dark-field optical microscope control unit 22-2 controls the dark-field optical microscope 27 to acquire an image at the field movement destination, and returns the acquired image information to the overall control unit 19.

  Next, the DFOM image determination unit 26-2 in the overall control unit 19 determines whether or not the image acquired in STEP 104 is an appropriate image for calculating the defect centroid (STEP 105). To determine whether the image is appropriate, for example, the image data is binarized, the number of pixels in the area corresponding to the defect is calculated, and an algorithm such as whether the number of pixels is larger than a certain threshold is used. In addition, when there are two or more regions that can correspond to defects in the binarized image due to halation or the like, an algorithm that determines that the image is not appropriate for calculating the defect centroid may be used. it can. Furthermore, even when there are two or more regions that can correspond to defects in the binarized image, a defect candidate that is closest to the center of the field of view and has a certain number of pixels in the defect candidate region. It is possible to use an algorithm that determines whether the value is also large.

  If the image is appropriate as a result of the determination, the coordinate transformation calculation unit 26-3 calculates the center of gravity coordinates of the defect by pixel calculation, and calculates the difference between the defect position information included in the inspection data and the calculated center of gravity coordinates of the defect. Fine alignment is performed by calculating (STEP 107). The fine alignment correction value calculated in STEP 107 is stored in a defect file stored in the storage device 16. The defect file stores the defect ID of the defect acquired by the external inspection apparatus and attribute information such as the position information or size of the defect corresponding to the ID, and the fine alignment correction value is also the above-described defect ID. Are recorded in correspondence.

  When it is determined in STEP 105 that the image is not appropriate, the DFOM image determination unit 26-2 instructs the dark field optical microscope control unit 22-2 to perform re-imaging (retry) while changing the optical condition. As described above, the dark field optical microscope of the present embodiment includes three optical filters, a neutral density filter, a polarization filter, and a wavelength filter, and can be adjusted by a combination of filters that use the image quality of an image to be acquired. it can.

For example, assuming that the optical characteristics of the neutral density filter, the deflection filter, and the wavelength filter are F ₁ , F ₂ , and F ₃ , the optical characteristics of the composite filter that combines them can be expressed as the following Expression 1. .

F (k) = αn ₁ F ₁ + n ₂ F ₂ + n ₃ F ₃ (1)
Here, alpha, or the coefficient of n _1, n _2, n _3, the coefficient indicating the dimming degree of alpha is the neutral density filter, is n _1, n _2, n _3, it is used or not the filter It takes a value of 0 or 1. k is an argument for designating a composite filter.

The storage device 16 stores optical conditions of the composite filter expressed by the permutation combination conditions of F ₁ , F ₂ , and F ₃ , and information on arguments for specifying the filter conditions. A specific filter condition can be called by specifying. Even if the dark field microscope has other than three optical filters, the optical characteristics of the composite filter are basically expressed by a linear combination of the optical characteristics of the individual filters as shown in Equation (1). Therefore, if the permutation combination condition of the optical characteristics of the individual filters is determined, the optical characteristics of the composite filter composed of an arbitrary number of filters can be designated.

  The DFOM image determination unit 26-2 calls the filter condition by specifying the above argument, and changes the optical condition at the time of retrying by instructing the dark field optical microscope control unit 22-2 (STEP 106). The dark field optical microscope control unit 22-2 adjusts the imaging conditions of the dark field optical microscope 27 according to the instructed conditions, and performs reimaging (STEP 104). At this time, the composite filter condition for the defect with the defect ID that was retried last time is stored in the storage device 16, and the optical condition of STEP 106 may be changed with reference to the stored filter condition at the time of this retry. Good. This process is particularly effective when the defects are sorted in size order and the dark field microscope is imaged in size order. This is because if the defect size is close, the optimum imaging condition of the dark field optical microscope is considered close.

  The DFOM image determination unit 26-2 determines whether or not the retried image can be used (STEP 105). If the image can be used, the coordinate transformation calculation unit 26-3 calculates the position of the center of gravity of the defect to obtain a fine image. Alignment is executed (STEP 107), and the calculated fine alignment correction value is stored in the storage device 16 in association with the defect ID. If it cannot be adopted, the flow of retrying by changing the optical conditions is repeated.

  After execution of STEP 107, it is determined whether or not imaging has been completed for all defects determined to have a defect size of “large” (STEP 111). If not completed, the process returns to STEP 104 to move the visual field. The next defect is imaged. If completed, the process proceeds to STEP 103, and a fine alignment sequence is executed for a defect determined to have a defect size of “medium to small”.

  Since the contents of the processing executed in STEP 108 to STEP 111 are substantially the same as the processing in STEP 104 to STEP 105 to STEP 107, the description will not be repeated. However, the initial setting value of the filter condition of the dark field optical microscope for the defect selected as the defect size “medium to small” is set to a value different from the initial setting value for the defect size “large”. It is stored in the storage device 16 together with information on the size class of the defect size “large” or “medium to small” and is referred to by the DFOM image determination unit 26-2.

  Even when the defect size is medium or smaller, an image inappropriate for calculating the defect centroid may be acquired. For example, there is a case where halation occurs due to a defect or a surface condition around the defect, or a case where the surface of the semiconductor wafer around the defect is rough and a plurality of bright spots due to scattered light exist. Even in that case, the optical conditions are automatically changed based on the determination result of STEP 109 (STEP 106), and the operations of STEP 104 to STEP 106 are executed until a dark-field optical microscope image suitable for calculating the barycentric coordinates of the defect can be captured. . In the determination of STEP 109, for example, it is possible to set so that the barycentric coordinates of the defect are calculated under an optical condition in which the number of bright spots on the dark-field optical microscope image or the size thereof is a certain value or less.

  It is also possible to set an upper limit for the number of retry attempts, and to exclude the defect from fine alignment if an appropriate dark-field optical microscope image cannot be obtained even after a specified number of retries. It is. For the defect with the defect ID excluded from the fine alignment target, there is no fine alignment correction value for the defect with the defect ID. For such a defect, the fine defect ID with the closest position is fine. Processing such as substituting the alignment correction value or substituting the average value of the fine alignment correction values of all other defects is possible. Such registration processing of the substitute correction value or calculation processing of the average value is executed by the coordinate transformation calculation unit 26-3.

  By determining the upper limit value of the number of retries, erroneous recognition of the center of gravity coordinates of the defect can be avoided, the coordinate conversion accuracy can be improved, and the defect capture rate in the SEM type review apparatus can also be improved.

  If it is determined in STEP 111 that fine alignment has been completed for all defects of both defect size “large” and “medium to small”, the control flow proceeds to STEP 112, and the SEM image capturing sequence is started.

  In the flow of FIG. 3 described above, an appropriate optical filter is automatically set for each defect by repeating the retry flow for the defect whose defect size is determined to be “large”, and on the dark field optical microscope image. It is possible to obtain a defect image in which the size of the defect appearing in (2) becomes an appropriate size for calculating the defect centroid. With this sequence, the defect centroid can be accurately calculated even for a defect whose defect size is determined to be “large”.

  On the other hand, for defects whose defect size is determined to be “medium to small”, the probability of halation is low, and it is possible to capture an appropriate dark field microscope image even if the initial setting values of the filter conditions of the composite filter are used as they are Therefore, it is possible to efficiently perform the imaging process to the defect centroid calculation process in STEPs 108 to 109.

  Further, in the flow of FIG. 3, for a defect whose defect size is determined to be “large”, the retry flow is scheduled to be repeated separately from the defect whose defect size is determined to be “medium to small”. . If the defect size is close, the optical filter conditions at the time of retry are likely to be close, so the optical filter conditions optimized for defects of the same size class of “large” or “medium to small” are the same size class. There is a high possibility that it can be applied as it is to a defect of the above or with a little adjustment. Therefore, by sorting the defects in order of size, it is possible to automatically set the optimum optical filter condition in a short cycle for the next defect having a similar defect size.

  In this way, by dividing the defect size into “large” and “medium to small” and separately capturing dark field optical microscope images, even if defects of various defect sizes coexist, In addition, accurate defect centroid calculation and defect coordinate correction can be performed in a short time. Not only the two large size classes “Large” and “Medium to Small” but also “Large”, “Medium”, “Small”, or “Class 1” “Class 2” “Class 3” “Class 4”. It is also possible to take an optical microscope image for fine alignment by performing size classification.

  FIG. 4 is a diagram illustrating an example in which a dark field optical microscope image optimal for calculating the barycentric coordinates of a defect is obtained by applying an optical filter to halation that occurs when the defect size is large. In this case, since the neutral density filter is particularly effective in many cases, when changing the optical condition (STEP 106), it may be configured to preferentially apply the neutral density filter.

  FIG. 5 is a diagram illustrating another example in which an appropriate image can be acquired by retrying. The left figure of FIG. 5 is an example of an image having a medium size or less but having many bright spots due to scattered light. Such imaging results often occur when the surface of the semiconductor wafer is rough. The right figure of FIG. 5 is an image obtained as a result of correction by the composite filter, and it is obtained that a dark-field microscope image including only a bright spot due to scattered light of a defect to be truly reviewed is obtained. I understand. In the case as shown on the left in FIG. 5, a combination of a neutral density filter and a polarizing filter or a wavelength filter is often effective, and therefore filter conditions to be preferentially applied may be set and registered in advance. Further, as described above, an upper limit value of the number of bright spots may be set.

  Returning to FIG. 3, the imaging sequence of the SEM image will be described. When the fine alignment is completed for all the defects and the coordinate conversion is completed, the overall control unit 19 captures a low-magnification SEM image at a preset magnification based on the coordinates after the fine alignment. The SEM control unit 22-1 and the stage control unit 23 perform imaging based on the instructions (STEP 112). The image calculation unit 20 calculates a more accurate position of the defect center using the low-magnification SEM image and transmits it to the overall control unit 19 (STEP 113). The overall control unit 19 instructs the SEM control unit 22-1 to move the defect portion to the center of the field of view by image shift and capture a high-magnification SEM image (STEP 114). The imaging magnification of the high-magnification SEM image is set in advance. Although illustration is omitted, the processing of STEPs 112 to 114 is repeated for all the defects, and the review image acquisition sequence is sequentially executed. When imaging is completed for all defects, the semiconductor wafer to be reviewed is unloaded from the defect review apparatus (STEP 115), and the captured dark-field optical microscope image or low-magnification or high-magnification SEM image is uploaded to the data management server 1. (STEP 116).

FIG. 7 is a diagram showing an example of a setting screen for filters, the number of bright spots, the number of trials, and the like applied in changing the optical conditions of the dark-field optical microscope. This setting screen is displayed on the display 17 and is applied when the apparatus user changes the optical condition in order to obtain an optimum dark field optical microscope image for calculation of the center of gravity coordinates of the defect via this setting screen. As the type of optical filter, a neutral density (ND) filter, a polarization (PL) filter, a wavelength (WL) filter, or the like can be selected, and the intensity can be selected for the neutral density filter. When selecting an optical filter to be used, the selection is performed by turning on / off a selection button shown in a dotted line frame 71 in FIG. When the intensity of the neutral density filter is selected, a numerical value is entered in the neutral density filter coefficient setting field shown in the dotted line frame 73. If there is an optical filter to be preferentially applied, it is selected by turning on / off the “priority” button shown in the dotted frame 73. The result of the selection button on / off and the numerical value input result to the dark filter coefficient setting field are reflected in the coefficients n ₁ , n ₂ , n ₃ and α included in Equation 1, and are used when executing the fine alignment sequence. Complex filter candidates are set. The composite filter combination condition is calculated by the overall control unit 19.

  Furthermore, when setting the parameters used when determining whether or not to adopt an image with a bright spot as shown in FIG. 5, for example, the upper limit number of bright spots, the size of the bright spot, or the upper limit value of the number of retries In this case, an arbitrary numerical value is input to each input field shown in the dotted line frame 74. By pressing “OK” after selecting and inputting the parameters, the setting conditions are applied to the sequence in the automatic review. Note that the size of the bright spot set in advance may be based on the number of pixels calculated based on the dark field optical microscope image or on the actual measurement value in the dark field optical microscope image. good.

  As described above, in the defect review apparatus according to the present embodiment, when capturing an optical microscope image for fine alignment, a plurality of target defects to be imaged are selected according to size, and an image and image are used for each selected size class. A sequence for determining availability / re-imaging is executed. Thereby, even when defects of various defect sizes coexist, accurate defect centroid calculation and defect coordinate correction can be performed for all defects in a short time.

  In addition, for an image that is determined to be unusable, re-imaging is performed by changing the filter condition of the optical filter. Thereby, when performing fine alignment, it becomes possible to employ even a defect that has not been conventionally employed, and the number of defects used for fine alignment increases. As a result, defect review points having coordinate correction values are increased as compared with the prior art, so that the probability of searching the field of view of the SEM image during ADR execution can be reduced.

  Furthermore, in the defect review apparatus of the present embodiment, in principle, coordinate correction values can be given to defects of all defect IDs. As a result, the defect can be moved and captured near the center of the field of view of the SEM image when the SEM review is executed. Specifically, when taking a low-magnification SEM image of STEP 112, it is possible to greatly reduce the probability that a defect is out of the field of view. Re-imaging a low-magnification SEM image) can reduce the occurrence of processing.

  Further, since the defect can be positioned substantially at the center of the field of view of the SEM image when the SEM review is executed, when the fine alignment is completed (step 111), the SEM image acquisition / defect recognition sequence at a low magnification (step 112). And 113) can be omitted, and only SEM image acquisition at a high magnification (step 114) can be executed. This flow is particularly effective when the number of defect points to be reviewed is very large or when the review time allowed for one wafer is short.

  Further, since the distribution state of the coordinate correction value on the sample is known, it can be used as data for calculating the coordinate correction value more accurately.

  In the above description, the various arithmetic processes executed by the overall control unit 19 or the functional blocks formed inside are described as being realized by software. However, the dedicated processing circuit or chip of each functional block is used. It may be realized as hardware.

(Example 2)
In the first embodiment, a configuration example of a defect review apparatus that executes a flow for performing fine alignment on all defects on a semiconductor wafer at the time of ADR execution has been described, but in this embodiment, at the recipe setting stage before ADR execution. A configuration example of a defect review apparatus having a function of performing fine alignment and reflecting the result on the imaging magnification setting of the low-magnification SEM image will be described.

  The hardware configuration and operation flow of the defect review apparatus of this embodiment are almost the same as those in FIGS.

  In the case of the present embodiment, the flow of FIG. 3 is executed at the recipe creation stage, and fine alignment correction values for all defect IDs obtained by executing the flow of FIG. 3 are stored in the storage device 16 as inspection recipes. The set inspection recipe is used for defect review of semiconductor wafers manufactured through the same manufacturing process.Of course, the defect position also varies with different wafers, so the created fine alignment correction value itself is used. I can't turn it.

  Therefore, the defect review apparatus of the present embodiment calculates the average value using the fine alignment correction values for all defects (or defects with a number of points that can be said to be substantially all defects) created at the recipe stage, Based on the average value, the visual field size of the low-magnification SEM image acquired in STEP 112 is determined. This function is realized when the overall control unit 19 calculates an average value using the fine alignment correction value stored in the defect file.

  As a result, the field-of-view size of the low-magnification SEM image can be optimized, so that the imaging time is shortened, and as a result, the ADR execution time can be shortened.

  As described above, according to the present invention, the SEM type defect review method and the SEM type defect review apparatus capable of reducing the man-hours required for coordinate conversion at the time of defect review and easily reviewing the defect are realized.

  In the first and second embodiments described above, the SEM is used as the enlarged imaging device for review. However, as the enlarged imaging device, an optical microscope using visible light or a microscope using ultraviolet light is used. But you can. The same function and effect can be obtained as long as the apparatus has a function capable of enlarging and imaging regardless of the type of energy used, the intensity of energy, and the visualization method.

DESCRIPTION OF SYMBOLS 1 Data management server 2 Manufacturing apparatus 3 Inspection apparatus 4 Review apparatus 5 Analysis apparatus 6 Review analysis apparatus 7 Network 8 Imaging apparatus 16 Storage apparatus 17 Display 18 Input apparatus 19 Overall control part 20 Image calculation part 26-1 Defect selection part 26-2 DFOM image determination unit 26-3 Coordinate transformation calculation unit 27 Dark field optical microscope

Claims (13)

In a defect review apparatus that acquires an image of the defect position on the sample using known defect position information on a predetermined sample,
A scanning electron microscope that scans a primary electron beam with respect to the defect position and outputs a detected signal as image data;
An optical microscope that irradiates light to the known defect position and outputs information of reflected light or scattered light detected as a detection signal;
An overall control unit for calculating a conversion rule for converting the known defect position information into imaging position information of the scanning electron microscope based on a defect detection signal imaged by the optical microscope;
The overall control unit selects a defect used for calculating the conversion rule on the basis of defect size information obtained from the reflected light or scattered light information. The defect review apparatus according to claim 1,
An optical microscope control unit for adjusting the operating conditions of the optical microscope;
The optical microscope control unit controls the optical microscope to reacquire information on reflected light or scattered light by changing an acquisition condition for a defect that has not been adopted for calculating the conversion rule. Defect review device to do. The defect review apparatus according to claim 2,
The optical microscope includes at least one of a neutral density filter, a polarizing filter, and a wavelength filter,
A defect review apparatus, wherein the acquisition condition is changed by detecting the reflected light or scattered light using any one of the neutral density filter, polarizing filter, and wavelength filter. The defect review apparatus according to claim 2,
The defect used for calculating the conversion rule is selected based on the defect size information obtained from the re-acquired reflected light or scattered light information for the defect that has not been adopted for calculating the conversion rule. A defect review apparatus characterized by that. The defect review apparatus according to claim 4,
The overall control unit calculates the conversion rule using information of the reflected light or scattered light with respect to the selected defect, and the defect review apparatus. The defect review apparatus according to claim 4,
The defect control apparatus, wherein the overall control unit does not use the information on the reflected light or scattered light with respect to the defect leaked from the sorting to calculate the conversion rule. The defect review apparatus according to claim 4,
The defect control apparatus according to claim 1, wherein the overall control unit includes a storage unit that stores a selection rule for executing the selection. The defect review apparatus according to claim 2,
A defect review apparatus comprising: display means for displaying a GUI for setting the acquisition condition of the optical microscope. The defect review apparatus according to claim 8,
The optical microscope includes at least one of a neutral density filter, a polarizing filter, and a wavelength filter,
The defect review apparatus, wherein a filter condition setting column for the neutral density filter, polarization filter, or wavelength filter is displayed on the GUI. In a defect review method for acquiring an image of the defect position on the sample by a scanning electron microscope using known defect position information on a predetermined sample,
Illuminating the known defect location with light and performing a first determination on the defect based on information of reflected or scattered light detected;
For the defect that has fallen in the first determination, change the acquisition condition to re-acquire information on the reflected light or scattered light,
A second determination is made under the same conditions as in the first selection for the defect that is the target of reacquisition of the reflected light or scattered light,
A conversion rule is generated for converting the known defect position information into imaging position information of the scanning electron microscope based on information of reflected light or scattered light with respect to the defect that has passed the first determination and the second determination. ,
A defect review method characterized by acquiring a scanning electron microscope image of a defect position converted based on the conversion rule. The defect review method according to claim 10,
The defect review method, wherein the first determination and the second determination are performed based on the presence or absence of halation with respect to the reflected light or scattered light. In a defect review apparatus that acquires an image of the defect position on the sample using known defect position information on a predetermined sample,
A scanning electron microscope,
A dark field microscope,
A dark field microscope control unit for controlling the operation of the dark field microscope;
An overall control unit that performs overall control of the control unit of the scanning electron microscope and the control unit of the optical dark field microscope,
The overall controller is
Using the defect image data acquired by the dark field microscope, a coordinate conversion calculation unit that generates a conversion rule for converting the known defect position information into imaging position information of the scanning electron microscope,
A determination unit that determines whether image data acquired by the dark field microscope is appropriate for generation of the conversion rule,
The defect review apparatus, wherein the dark field microscope control unit adjusts an image acquisition condition of the dark field microscope based on the determination of the determination unit. The defect review apparatus according to claim 12,
The dark field microscope includes at least one of a neutral density filter, a polarizing filter, and a wavelength filter,
The defect review apparatus, wherein the dark field microscope control unit adjusts filter conditions of the neutral density filter, polarization filter, or wavelength filter based on a determination result of the determination unit.

Images