JP2000306964A - Inspection data processing method, and inspection data processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To review a defect at high speed, by judging the fatality of a defect from the fatality judgment data corresponding to the in-chip existence region of the defect and the data of the size of the defect, and writing the information to show the degree of the fatality of each defect based on this judgment result into an inspection data base. SOLUTION: The inspection data such as the inspection process, the defect coordinate, the size of the defect, etc., stored in an inspection data storage 21, and are read in a memory 6, and also chip region coordinate data stored in a chip region coordinate data storage 23 are retrieved with the retrieval part 5, and further are read in the memory 6. Furthermore, the fatality judgment data corresponding to each chip region of an inspection process are retrieved with the retrieval part 5 from among a judgment data storage means 24, and are read in the memory 6. Then, an operating part 9 judges the fatality of the defect from the fatality judgment data and the defect size data from the memory 6, and writes the information to show the degree of fatality of each defect based on the judgment result into the memory 6.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a defect generated in a circuit pattern of a semiconductor wafer, a mask, a liquid crystal or the like (a defect referred to in the present specification is a defect such as a short-circuit, a disconnection, a film residue or the like generated in a wafer process or the like). And inspection data processing technology that can determine the criticality of defects due to attached foreign matter). Techniques suitable for determining the criticality of element defects and selecting defects for review and analysis, and techniques suitable for determining and correcting criticality using defect categories added after review The present invention relates to a technique suitable for editing, searching, displaying, and externally outputting used inspection data.

[0002]

2. Description of the Related Art In a semiconductor manufacturing process, it is essential to find a cause of a defect early and feed it back to a process and a manufacturing apparatus in order to maintain and improve the yield. This includes
It is important to detect defects using an inspection device and analyze the inspection data.

According to the study by the present inventors,
For example, short-circuits, disconnections, and the like, and foreign substances generated in the wafer process are automatically inspected by image processing or an inspection apparatus using dark-field / bright-field illumination by laser light.

[0004] The inspection apparatus described above calculates coordinate data and a defect size of a defect (including a foreign substance) in a semiconductor wafer.
The data is output to an analysis system that stores the data. Then, the inspected wafer is transferred to a stage of a review device such as a metal microscope or a scanning electron microscope, the stage is moved to the position of the coordinate data of the detected defect, and the defect is classified based on an enlarged image of the defect. This classification work is called a review, and there are the following two types of classification for this review.

[0005] One is a classification of a defect itself, which is a classification of a detected defect such as a disconnection, a short circuit, a film residue, or a foreign matter. The other is a classification of a defect to a function of a semiconductor device. This is a fatal / non-fatal classification that determines whether a defect is fatal.

[0006] After the review is completed, an inspection personnel who inspects using a metallographic microscope or a scanning electron microscope outputs a predetermined multi-valued classification number to the analysis system in accordance with the classification of the defect itself and fatal / non-fatal. I do.

[0007]

However, among the above review work, the inventors of the present application have found that there are the following problems, particularly regarding the classification of fatal / non-fatal. The term “fatality” as used herein indicates the degree of influence of a defect on a circuit pattern, and means the possibility that the defect will cause the defect. Means that the electrical operation of the circuit pattern is likely to be defective (causing a defect)).

That is, since the above-mentioned review operation is a manual operation, it takes a long time and is a constraint on improving the throughput of the inspection process. In addition, regarding the determination of fatality, not only the appearance of the defect itself but also knowledge of the function and structure of the circuit pattern in the area where the defect exists is required, and it is left to a specific expert with specialized knowledge. Was.
Furthermore, since the criteria for judging the lethality differ among these specialists, there is a problem that the judgment results vary from person to person.

If the criticality determination by the review is not performed at all or only a part of the defect is determined because of the throughput, the following inconvenience occurs. That is, as shown in FIG. 15A, there is no correlation between the total number of defects including non-fatal defects and the yield of semiconductor chips (elements), and as a result, a meaningful management value cannot be set. For this reason, as shown in FIG. 15B, the yield abnormality is predicted based on the time series transition of the total number of defects for each inspection wafer, and even if an early measure is taken, the total number of defects functions as a monitor value of the yield. Therefore, even if the control value is set, the abnormality cannot be detected skillfully. Therefore, it is important to enable the criticality determination of all the detected defects by increasing the throughput of the criticality determination.

At the time of review, it is necessary to view an enlarged image with a metal microscope or a scanning electron microscope, as described above. Reviewing all detected defects is difficult from a throughput perspective. Therefore, it is necessary to narrow down the defects to be reviewed. This narrowing down operation is also left to the reviewer, and the selection result of the review target may vary depending on the person.

Further, in the analysis / analysis work after the review, it is necessary to narrow down the defects. However, since it depends on the reviewer, there is a high possibility that the target will be out of target, and it is inefficient.

Accordingly, a technical problem to be solved by the present invention is to solve the above-mentioned problems of the prior art, and a first object of the present invention is to perform a defect review operation at high speed, that is, It is an object of the present invention to provide a method and an apparatus for performing defect fatality determination at high speed without retaking an image. A second object of the present invention is to provide a method and apparatus for automatically selecting a review target. It is a third object of the present invention to provide a method and an apparatus for improving the accuracy of defect fatality determination by a case-study learning effect using a defect category added after a review. A fourth object of the present invention is to
Using these inspection / analysis result information, to provide a method and apparatus for editing, searching, displaying, and externally outputting inspection data, and a method for efficiently performing analysis and analysis such as cross-sectional observation of a defective portion, and It is to provide a device.

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

[0014]

SUMMARY OF THE INVENTION Among the inventions disclosed in the present application, typical ones will be outlined as follows.

That is, according to the inspection data processing method of the present invention, chip area coordinate data on a semiconductor wafer is
A step of obtaining criticality determination data of each chip area, coordinate data and size data (defect size data) of a detected defect, and criticality determination data and defect size data corresponding to a defect existing area in a chip. And a step of determining (calculating) the criticality of the defect from the above, and a step of writing a multi-value classification number indicating the criticality / non-fatality of each defect based on the determination result in a defect database of the inspection data processing system.

Thus, the criticality of the defect detected by the inspection device can be automatically determined without re-enlarging an enlarged image with a metal microscope or a scanning electron microscope. At this time, the criticality determination of the defect can be performed at high speed, and the criticality determination of all the detected defects can be performed. The fatality determination is to perform filtering from total defects to fatal defects. As shown in FIG. 16A, the number of fatal defects and the chip yield can be correlated. For this reason, the management value can be set from the yield value, and if the abnormality for each inspection wafer is managed by the number of fatal defects as shown in FIG. 16B, the abnormality can be detected with high accuracy. .

The method for identifying the defective area in the chip includes, for example, a step of generating chip area image data by dividing the chip into pixels and assigning a class value corresponding to the chip area to each pixel. And converting the coordinate data of the detected defect into an address of a pixel in the chip area image; and identifying the defective area in the chip from the class value of the pixel at the pixel address corresponding to the detected defect. To do.

This makes it possible to identify a defective chip existing area at high speed.

The method of selecting a defect to be reviewed / analyzed in the inspection data processing method of the present invention includes a step of reading chip area coordinate data on a semiconductor wafer, a step of reading fatality determination data of each chip area, Reading the coordinate data and size data of the defect, determining the existence area of the defect in the chip, calculating the fatality of the defect from the acquired data, And a step of sequentially selecting until the total number of defects to be reviewed / analyzed reaches a preset total number of defects to be reviewed / analyzed;

As a result, it becomes possible to automatically select the defects to be reviewed / analyzed in descending order of fatality.

In the inspection data processing method according to the present invention, the method of selecting a defect to be reviewed / analyzed may be such that after reading the coordinate data and size data of the detected defect, the detected defect in the step prior to the semiconductor manufacturing process is read. A step of comparing the coordinate data with the coordinate data of the detected defect in the process, and selecting only the detected defect in the process that does not match.

This eliminates the waste of reselecting the defects already reviewed in the previous step.

The method of selecting a defect to be reviewed / analyzed in the inspection data processing method according to the present invention employs a method of selecting coordinate data of a detected defect that does not match the detected defect in a step before the semiconductor manufacturing process. And classifying them into dense defects and random defects.

Thus, there is no waste in selecting a plurality of defects belonging to the same density and having the same classification as objects to be reviewed and analyzed.

Further, in the inspection data processing method of the present invention, the review / analysis target defect selection method is the same as the above-described method of classifying dense defects and random defects in the relevant step, but also detecting defects included in the dense defects in the previous step. Is compared with the coordinate data of the detected defect included in the dense defect in the process, and when one or more detected defects match, the dense defect in the previous process and the dense defect in the process are compared. All the defects included in the same classification.

As a result, it is possible to omit a defect in the process caused by the dense defect generated in the previous process from a review / analysis selection target.

In the inspection data processing method according to the present invention, the method for improving the accuracy of the criticality determination after review and the method for editing and retrieving inspection data include the steps of reading chip area coordinate data on a semiconductor wafer, Reading the criticality determination data, reading the coordinate data and size data of the detected defect, determining the existence area of the defect in the chip, and calculating the first criticality of the defect from the obtained data. Calculating the second criticality in consideration of the defect category using the defect category added by the review, writing the criticality into an inspection database of the analysis system, editing the inspection data, Searching.

Thus, defects are automatically selected in descending order of criticality, defects are automatically selected in descending order of first criticality for each defect category, and a defect category and a first criticality are determined. Edit using parameters such as dimensions (defect size) and view wafer maps (distribution) of defects,
It is possible to browse, edit, and rearrange desired images from the images detected during the review. In particular, when defect category classification by automatic review is used, a highly fatal defect image is provided to an operator, and efficient review and analysis can be performed. Further, the result of the electrical inspection is fed back to the criticality determination procedure, and the precision of the criticality determination can be improved.

Further, the review / analysis order and the review / analysis target can be edited using the inspection data having the input chip coordinates, defect coordinates, criticality, defect size, defect category, and the like. Therefore, an inspection data processing system (review / analysis system) capable of controlling the review / analysis order and the review / analysis target can be constructed. In addition, by browsing the images detected by the review / analysis device and providing an editable viewer such as numbering, the user can freely edit the review results,
Can be used to create reports. Further, even when a cross-sectional photograph of a defective portion is taken by an SEM, it is possible to narrow down data more efficiently by using the above-described inspection data processing method.

Further, the inspection data processing apparatus of the present invention has a first storage means for storing chip area coordinate data on a semiconductor wafer, a first storage means for dividing a chip into pixels, and assigning a class value corresponding to the chip area to each pixel. A first calculating means, a second storing means for storing the chip area image data calculated by the first calculating means, a third storing means for storing the criticality judgment data of each chip area, A second calculating means for converting the coordinate data of the pixel into an address of a pixel in the chip area image; and a first control means for identifying a defective area in the chip from a class value of a pixel at a pixel address corresponding to the detected defect. A second control means for judging the fatality of the defect from the existence area of the defect in the chip and the fatality judgment data corresponding thereto, and a multi-level classification number indicating the fatality / non-fatal of each defect is inspected. It includes a data input means for writing the database, the first, the second, the display unit for displaying the data stored in the third storage means. As a typical thing displayed on the display unit, as shown in FIG. 16B, the inspection wafer is shown on the horizontal axis, and the number of fatal defects of each inspection wafer obtained by the fatality determination is shown on the vertical axis. There is a fatal defect count control chart to be displayed for each.

Further, the inspection data processing apparatus of the present invention comprises a first storage means for storing chip area coordinate data on a semiconductor wafer, a first storage means for dividing a chip into pixels, and a class value corresponding to the chip area allocated to each pixel. A first calculating means, a second storing means for storing the chip area image data calculated by the first calculating means, a third storing means for storing the criticality judgment data of each chip area, A second calculating means for converting the coordinate data of the pixel into an address of a pixel in the chip area image; and a first control means for identifying a defective area in the chip from a class value of a pixel at a pixel address corresponding to the detected defect. A second control means for judging the fatality of a defect from the existence area of the defect in the chip and the fatality judgment data corresponding thereto, and a defect size obtained by the second control means. The size threshold ratio (value indicating the degree of fatality), which is the ratio of the criticality judgment threshold value corresponding to the area where the defect exists in the chip, is used to determine the defect to be reviewed and analyzed. Comparing the coordinate data of the detected defect in the previous step with the coordinate data of the detected defect in the previous step, and the third control means for selecting until the preset total number of defects to be reviewed / analyzed in order from the largest one, A fourth control unit that selects only the detected defect in the process that does not match, and classifies the detected defect that does not match the detected defect in the previous process into a dense defect and a random defect based on the coordinate data of the detected defect. The fifth control means compares the coordinate data of the dense defect in the previous step with the coordinate data of the detected defect included in the dense defect in the previous step. Congestion defects and the process Sixth control means for classifying all the defects included in the dense defect into the same classification, data stored in the first, second, and third storage means and review selected by the third control means -It has a display unit and an output unit for displaying the defect to be analyzed.

Further, the inspection data processing apparatus of the present invention comprises a first storage means for storing chip area coordinate data on a semiconductor wafer, and a chip for dividing a chip into pixels and assigning a class value corresponding to the chip area to each pixel. A first calculating means, a second storing means for storing the chip area image data calculated by the first calculating means, a third storing means for storing the criticality judgment data of each chip area, A second calculating means for converting the coordinate data of the pixel into an address of a pixel in the chip area image; and a first control means for identifying a defective area in the chip from a class value of a pixel at a pixel address corresponding to the detected defect. A second control means for determining a first fatality of a defect from a defect existing area in the chip and the fatality determination data corresponding thereto, and a defect obtained by the second control means. A second threshold using a size threshold ratio (first fatality) which is a ratio of a criticality determination threshold value corresponding to the size of the defect and the defect existence area in the chip, and a defect category added by the review. A third calculating means for judging the sex, editing data stored in the first, second, and third storing means and data obtained by the second controlling means or the third calculating means; A display unit and an output unit for searching and displaying these maps and images are provided.

[0033]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing a configuration of a system using a test data processing apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the inspection data processing apparatus 1 of the present embodiment is connected via a network Nt to an inspection apparatus WI for inspecting a defect of a circuit pattern formed on a semiconductor wafer. In the process in which the wafer W is processed with the film forming apparatus, the exposure apparatus, and the etching apparatus, for example, the inspection apparatus WI inspects the wafer W at the time when the etching is completed. The process is returned to the beginning. The data of the coordinates and the size of the defect output from the inspection device WI are stored in the inspection data storage means 21 provided in the inspection data processing device 1.

The inspection data processing device 1 includes an inspection device WI
Communication control unit 3 for controlling communication of data input / output from / to
Is provided.

The test data processor 1 is provided with a data input / output unit 4 for inputting / outputting data to / from a communication control unit 3 for controlling data communication. 4 is connected to the communication control unit 3.

Further, the test data processing device 1 includes a search unit 5 for searching for data stored in the test data storage unit 21, the chip area coordinate data storage unit 23, and the determination data storage unit 24, and the search unit 5 A memory 6 for storing the data searched by the search unit 5 is provided. The search unit 5 is connected to the memory 6, and the memory 6 is connected to the data input / output unit 4.

The test data processor 1 is provided with an input / output interface 7 for adjusting the timing of data exchange with the test data storage unit 21, the chip area coordinate data storage unit 23, and the determination data storage unit 24. The input / output interface 7 is connected to a search unit 5, a data input / output unit 4, an inspection data storage unit 21, a chip area coordinate data storage unit 23, and a determination data storage unit 24.

Further, the test data processing apparatus 1 has a program storage unit 8 in which a program which is software for performing necessary processing is stored, and an operation for performing an operation based on the program stored in the program storage unit 8. A unit (which serves as first and second calculation means) 9 is provided, and the calculation unit 9 is connected to the memory 6.

The test data processing apparatus 1 is provided with a main control section (first and second control means, etc.) 10 which controls all controls in the test data processing apparatus 1. The control unit 10 is connected to a data input / output unit 4, a search unit 5, a memory 6, a calculation unit 9, and a program storage unit 8.

Further, the display section 11 is connected to the data input / output section 4.

Note that the display unit 11 displays the inspection data storage unit 2
1. Display data of the chip area coordinate data storage unit 23 and the determination data storage unit 24, a map of defect information, and the like.
As a representative one displayed on the display unit 11, FIG.
As shown in FIG. 3B, there is a critical defect number management chart in which the inspection wafer is displayed on the horizontal axis and the number of fatal defects of each inspection wafer is displayed on the vertical axis.

The display unit 11 may be provided in a place different from the test data processing apparatus 1 and connected to the communication control unit 3 and the data input / output unit 4 via the network Nt.

The test data storage unit 21, the chip area coordinate data storage unit 23, and the determination data storage unit 24 may be the same storage medium.

Further, the test data processing device 1 may be connected to another system via the network Nt to transmit and receive data of the test data storage unit 21, the chip area coordinate data storage unit 23, and the determination data storage unit 24. .

Further, the inspection data processing apparatus of the present embodiment may be incorporated in the inspection apparatus WI without using the network Nt.

The arithmetic unit 9 and the main control unit 10 may be the same semiconductor device.

Further, the arithmetic unit 9, the main control unit 10, the search unit 5, the memory 6, the data input / output unit 4, and the communication control unit 3 may be the same semiconductor element.

Next, the data processing method of this embodiment will be described with reference to the flowchart of FIG.

First, in step S 101, the search unit 5 searches for inspection data such as an inspection process, defect coordinates, and defect size stored in the inspection data storage unit 21, and reads the inspection data into the memory 6.

Further, in step S 102, the search section 5 searches the chip area coordinate data stored in the chip area coordinate data storage section 23 and reads the chip area coordinate data into the memory 6.

As shown in the chip layout diagram of FIG. 3, for example, the coordinate data of the chip area is obtained by dividing the chip into a plurality of areas according to the width of the circuit pattern used.
Each region is represented by a rectangle, and each region is represented by the coordinates of the lower left and the coordinates of the upper right of each rectangle. For example, the area R0 has coordinates C01 and C02, and the area R10 has coordinates C101 and C101.
At 102, region R40 can be represented by coordinates C401 and C402.

As the chip area coordinate data, as shown in a chip area coordinate data diagram shown in FIG. 4, a lower left coordinate (x1, y1) and an upper right coordinate (x2, y2) of each area and an area class (X) indicating the type of the area. (Hereinafter, simply referred to as a class). The chip area coordinate data is written by manual input or by reading a file from CAD data. Also, the origin of the chip area coordinate data (for example,
C01) is converted so as to coincide with the chip coordinate origin of the inspection device W1.

Next, in step S103, step S1
The search unit 5 searches the judgment data storage means 24 for judgment data (judgment rule data) corresponding to each chip area of the inspection step read in 02 and reads it into the memory 6.

[0055] the decision data, as shown in FIG. 5, class defects corresponding to (class value C) size threshold R _C is written. The defect size threshold value _RC is, for example, a representative pattern width or a pattern space of a chip region corresponding to a class, or a design value of a pattern width or a pattern space of the next process, or a film thickness, or a local value such as a grain. Process information such as typical film thickness fluctuation. Since the typical pattern width or the design value of the pattern space differs depending on the chip area, the judgment data needs to be changed depending on the chip area.
To switch the judgment data. Of course, the chip area may be further subdivided, for example, subdivided into each pattern section, each space section, and the like, and a class value C may be assigned to each.

In FIG. 5, the defect size threshold R _C is shown.
The defect length threshold R _L and the defect area threshold R
_S, has been shown defects brightness threshold R _B.

Next, in step S104, the defect No. N
Is set to N = 1, the defect existence area is determined from the defect coordinates (x, y) read in step S102 in step S105, and the class value C corresponding to the defect existence area is written in the memory 6.

The determination of the defect existing area is made for each area No. by determining whether or not the following expression is satisfied.

X1 <x <x2 y1 <y <y2... Expression The class value C of the chip area is determined by overwriting in the order of the area numbers. The class value C is written into the memory 6 as the class value of the defect (that is, the class value C of the chip area having a small area is written into the memory 6 as the class value of the defect).

Next, in step S106, the defect N
o. From the defect size of N (for example, the defect length L _N ) and the defect size threshold value R _C (for example, the defect length threshold value R _L ) corresponding to the class value C, the “No. 1
Perform a criticality determination. Here, the defect length L _N may be a length when the defect is projected on the x-axis or the Y-axis. Alternatively, as the defect size, instead of the defect length L _N , an area S _{N of} the defect or an absolute value B _N of a difference in brightness from a normal portion of the defect may be used.

“First” fatality: L _N / R _C (here, R _C = R _L ): S _N / R _C (here, R _C = R _S ): B _N / R _C (here Then, R _C = R _B ) Expression If the judgment by the above expression is true, step S107
In the defect No. The multi-level classification number F of the fatal defect is written into the memory 6 corresponding to the N defect. If false, step S
108, the defect No. The multi-level classification number NF of the non-fatal defect is written into the memory 6 corresponding to the N defect.

Although the fatality of the above equation is a simple ratio value, the fatality is determined from the probe test results and the like so as to express the influence on the circuit pattern with higher accuracy as described later. Can be calculated using a function, a lookup table, or the like.

Further, after setting N = N + 1 in step S109, in step S110, the defect number is added to the inspection data.
It is determined whether there are N defects, and if there are, the process returns to step S105. If there is no defect, step S111
And the process ends, and steps S107 and S10
8, the multi-level classification number F of the fatal defect written in the memory 6
Then, the multi-value classification number NF of the non-fatal defect and the class value C written in the memory 6 in step S105 are written in the inspection data storage unit 21 in correspondence with the respective defect Nos.

FIG. 6 is a diagram showing the configuration of a system using the test data processing device according to the second embodiment of the present invention. In the figure, the same components as those in the first embodiment shown in FIG. 1 are denoted by the same reference numerals, and the description thereof will be omitted to avoid duplication.

The inspection data processing apparatus 1 of the present embodiment shown in FIG. 6 has a configuration in which a chip area image data storage means 22 is added to the configuration of the inspection data processing apparatus of FIG. The data storage means 22 is connected to the input / output interface 7 and can read and write data.

Next, the processing by the inspection data processing apparatus of the present embodiment will be described.

First, as preparatory work, a region image data creating process shown in the flowchart of FIG. 7 is performed. In the processing flow illustrated in FIG. 7, the search unit 5 searches the coordinate data of the chip area stored in the chip area coordinate data storage unit 23 in step S201, and reads the coordinate data into the memory 6.

Next, in step S202, step S2
01, the area image data is written with respect to the data of the first row (area No. n = 1) in the chip area coordinate data read in step S01.

That is, the lower left coordinates (x1, y1) and the upper right coordinates (x2, y2) of the chip area coordinate data of the first row (area No. n = 1) are converted into pixel coordinates by the following equation.

IX1 = Int (x1 / P) IX2 = Int (x2 / P) IY1 = Int (y1 / P) IY2 = Int (y2 / P) where P is a preset pixel pitch, Int Is a function that rounds down decimal places.

Next, in step S203, data of the class value C is written to each pixel IP (I, J) in the rectangular area represented by the coordinates (x1, y1) and the coordinates (x2, y2).

Further, in step S204, the area No. n
Is set to n = n + 1, and in step S205, the chip area coordinate data is set to the area No. It is determined whether there is n data. If there is no data, the writing is completed. If there is data, the process returns to step S202.

The processing in steps S202 to S204 is performed by the arithmetic unit 9, and the determination in step S205 is performed by the main control unit 10. If the main control unit 10 determines that the writing has been completed, the area image data IP
(I, J) (that is, the image data as information representing the area to which the class value C is added) is stored in the chip area image data storage unit 22 via the data input / output unit 4 and the input / output interface 7. Is stored.

The chip area image data storage means 22
When writing to the chip area, the storage capacity of the chip area image data storage means 22 may be saved by processing such as compressing and writing the chip area image data and decompressing again when reading.

Next, the inspection data processing apparatus 1 of the present embodiment
Performs a criticality determination process according to the flowchart of FIG.

In the processing flow shown in FIG. 8, first, in step S301, the search unit 5 searches the area image data stored in the chip area image data storage unit 22 and reads it into the memory 6.

Next, in step S302, the search unit 5 searches the inspection data stored in the inspection data storage unit 21 for the inspection process, defect coordinates, defect size, and the like.
Read in.

Further, in step S303, step S
The search unit 5 searches the determination data (judgment rule data) corresponding to the inspection process read in 302 from the determination data storage means 24 and reads it into the memory 6. As shown in FIG. 5, the defect size threshold value RC corresponding to the class value _C is written in the determination data.

Next, in step S304, the defect No. N is set to N = 1, and in step S305, the defect No. The defect coordinates (x, y) of N are calculated by the following equations using pixel coordinates (KX,
KY).

KX = Int (x / P) KY = Int (y / P) Expression where P is a preset pixel pitch and has the same value as the expression.

Further, in step S306, step 3
01, the pixel coordinates (KX, K
The class value C, that is, the defect existing area is determined from the value IP (KX, KY) of Y) and is written to the memory 6.

Next, in step S307, the defect N
o. From the defect size of N and the defect size threshold value R _C corresponding to the class value C, a “first” fatality determination represented by the above equation is performed.

[0083] If the expression is true, in step S308, the defect No. N
The multi-level classification number F of the fatal defect corresponding to the defect
Write to. If false, the defect N is determined in step S309.
o. Multi-level classification number NF of non-fatal defect corresponding to N defect
Is written to the memory 6.

Further, after setting N = N + 1 in step S310, the defect number is added to the inspection data in step S311. N
It is determined whether or not there is a defect.
Return to 305. If there is no defect, the process ends in step S312, and steps S308 and S309.
The multi-level classification number F of the fatal defect written in the memory 6 and the multi-level classification number NF of the non-fatal defect and the class value C written in the memory 6 in step S306 correspond to each defect No. The data is written in the inspection data storage unit 21.

The use of the chip area image data of the present embodiment has an advantage that a defect existing area can be identified instantaneously.

FIG. 9 is a diagram showing the configuration of a system using the test data processing device according to the third embodiment of the present invention. In the figure, the same reference numerals are given to the components of the first embodiment of FIG. 1 and the second embodiment of FIG. 6, and the description thereof will be omitted to avoid duplication.

The inspection data processing device 1 of the present embodiment has the functions of the inspection data processing device of the second embodiment,
It has a function of selecting a defect to be reviewed and analyzed. The inspection data processing apparatus 1 according to the present embodiment is configured as shown in FIG.
The review / analysis target data storage means 25 is added to the configuration of the inspection data processing apparatus of the above. The review / analysis target data storage means 25 is connected to the input / output interface 7 so that data can be read and written. It is possible.

As shown in FIG. 9, the inspection data processing apparatus 1 according to the present embodiment, in addition to the above-described inspection apparatus WI, inspects an enlarged image of a metal microscope, a scanning electron microscope, etc. It is also connected via a network Nt to a review station (review / analysis device) RS for classifying defects.

The review station RS is a
The coordinate data of the defect to be reviewed is read from the analysis target data storage means 25, and the stage is moved to the defect position based on the coordinate data. The image detected by the review station RS is stored in the inspection data storage unit 21.

The inspection data storage section 21, the chip area image data storage section 22, the chip area coordinate data storage section 2
3. The judgment data storage unit 24 and the review / analysis target data storage unit 25 may be the same storage medium.

The display unit 11 may be provided at a place different from the inspection data processing device 1 and connected to the communication control unit 3 and the data input / output unit 4 via the network Nt.

Further, the inspection data processing apparatus 1 is connected to another system via the network Nt, and the inspection data storage unit 21, the chip area coordinate data storage unit 22, the chip area coordinate data storage unit 23, the judgment data storage unit 24, The data of the review / analysis target data storage means 25 may be transmitted and received.

Further, the inspection data processing apparatus 1 of the present embodiment
May be incorporated into the main body of the inspection apparatus WI without using the network Nt.

Next, the inspection data processing apparatus 1 of the present embodiment
The process of selecting a defect to be reviewed / analyzed will be described with reference to the flowchart of FIG.

Prior to the execution of the processing flow shown in FIG. 10, the criticality determination process described with reference to FIG. 2 or FIG. 8 is performed, and the above-described multi-value classification number and class value C are written in the inspection data storage unit 21. It is assumed that

First, in step S401, an inspection process for review / analysis and the number of selected defects _Nmax for review / analysis are input from the display unit 11 (the display unit 11 is assumed to have an input function unit). Are stored in the memory 6 via the data input / output unit 4.

Next, in step S402, from the inspection process input in step S401, the retrieval unit 5 retrieves necessary data such as defect coordinates and multi-level classification numbers stored in the inspection data storage unit 21, and the memory 6 Read in.

Next, in step S403, a defect size threshold value R corresponding to the defect size and the class value C is set for each defect.
_C , the size threshold ratio (the value corresponding to the multi-valued classification number) shown in the above equation is in descending order.
That is, the defects are rearranged in descending order of fatality.

Next, in step S404, N _max defects are selected from the top in the order after the processing in step S403, and the selected defects are written to the review / analysis target data storage means 25 in step 405.

In step 401, the order of the class to be prioritized in the review / analysis selection is input, and step S403 is performed.
The defects may be rearranged in the order of priority classes. This makes it possible for the review / analysis operator to automatically select a defect in the chip area (class) that the user wants to review first.

In step S401, the maximum number of reviews / analysis selections per chip in the wafer, _Icmax, is input. If the number of selections per chip exceeds _Icmax at the time of defect selection in step S404, _Icmax is set. A process may be added so as not to select a defect exceeding the number. In this way, even when defects are concentrated on the same chip, it is possible to select a defect to be reviewed and analyzed from the entire wafer.

Here, in the above description, the defect to be reviewed / analyzed is selected based on the criticality. However, the selection condition is set externally or in advance, and the defect meeting this selection condition is reviewed / analyzed. It is also effective to extract as a power target.

In this case, for example, as the selection conditions, a) defect size b) defect coordinates c) fatality d) chip identification coordinates indicating which chip on the wafer e) chip coordinates indicating the area of each chip, etc. Any one of these, or
Any combination of a plurality of conditions can be used as a condition for selecting a defect to be reviewed and analyzed.

The selection condition here is basically that the fatality of a defect (including a foreign substance) is determined by the region of the wafer. FIG. 17 shows an example.

By the way, semiconductor products are formed in each process.
It is manufactured by repeating exposure and etching, and inspection is also performed in each step.

Here, FIG. 11 shows the transition of the number of defects detected in each process, which was examined by the inventors of the present invention, for each generation process. It can be seen that it is detected again. Therefore, as the process progresses, the number of defects already reviewed / analyzed out of the number of defects to be reviewed / analyzed increases, indicating that time efficiency is poor.

Therefore, as shown in FIG. 12, the coordinate data 211 of the defect in the current process on the semiconductor wafer W is obtained.
The search unit 5 retrieves the coordinate data 212 of the defect in the previous process from the inspection data storage unit 21 and reads the coordinate data 212 into the memory 6. 211 and the coordinate data 21 of the defect in the previous step
2. The coordinate data of the defect of the current process read from the memory 6 is read from the memory 6 by aligning the coordinate system with the coordinate system 2 and reading the coordinate data of the defect of the current process which is coincident or located at a position closer than a predetermined allowable value. , A process of storing only the coordinate data 213 of the defect newly found in the current process in the inspection data storage unit 21 (or the review / analysis target data storage unit 25) may be added.

By doing so, it is possible to prevent the defect already reviewed in the previous step from being reviewed again. Of course, the defect that occurred in the previous process
In order to find out what has happened in the current process, the defect including the defect generated in the previous process may be extracted and subjected to the criticality determination and review processing.

Further, according to the study by the present inventors, for example, the wafer W detected by the inspection device WI is, for example,
When there is a large defect such as a flaw, it is detected as a group of many defects, that is, a dense defect. Even when reviewing these dense defects, the inspection data storage means 21 is registered as individual defects, so it is necessary to review individual defects constituting a large defect one by one. There are inefficiencies.

Therefore, the coordinate data 214 of a defect newly found in the current step after the defect deletion in the previous step, which is stored in the review / analysis target data storage unit 25, is read into the memory 6, and FIG. As shown in the figure, a cluster multi-value classification number indicating that each defect having a close coordinate belongs to the same dense defect (cluster) is given to each defect of the coordinate data 214 of the defect newly found in the current process. The data is stored in the review / analysis target data storage unit 25. Further, the number of defects corresponding to the number of reviews / analysis or the ratio of review / analysis in one cluster set in the display unit 11 in advance is selected from the same cluster, and the coordinate data of the defect after the selection in the cluster is reviewed / reviewed. Processing for storing in the analysis target data storage unit 25 is added. Further, after the review by the review station RS, the review / analysis multi-value classification number is given to each defect to be reviewed, and the defect is transferred to the inspection data processing apparatus 1. Of the coordinate data 214 of the defect newly found in the process of step 2
5, and among the defect coordinate data newly found in the current step with the cluster multi-level classification number, all of the defects that match the cluster multi-level classification number of the coordinate data of the defect selected within the cluster. Review and analysis
A review / analysis multi-value classification number of the coordinate data of the defect after selection in the cluster may be added.

By doing so, even if there is a cluster which is a large defect such as a flaw, the review / analysis of the entire cluster can be performed simply by reviewing / analyzing one to several defects constituting the cluster. It becomes possible to assign a value classification number.

In the cluster of the above-described example, the dense defects having a short distance are grouped into one group. However, the physical characteristics of the defects detected by the inspection apparatus W1, such as the size, brightness, color, and shape of the defects, are set as follows. Taken into the inspection data storage means 21,
Grouping may be performed based on these (by one of these or any combination of these conditions).

In this way, by grouping defects having the same physical characteristics other than dense defects, one to several defects in the group can be reviewed and processed by the same processing as described above. Simply by performing analysis, a review / analysis multi-value classification number for the entire group can be assigned.

According to the study by the present inventors, when the detection sensitivity of the inspection apparatus WI differs between the steps,
For example, when the detection sensitivity in the preceding step A is lower than the detection sensitivity in the subsequent step B, the defect that has already occurred in the step A is first detected in the subsequent step B. At this time, as shown in FIG. 14, the cluster 216 in the process B is detected to be wider than the cluster 217 in the process A. Here, when the detected defect is deleted in the previous process described with reference to FIG.
Is recognized as a newly generated cluster, the originally identical cluster 217 and cluster 218 are divided, and different cluster multi-valued classification numbers are assigned. As a result, the process of allocating the review / analysis multi-valued classification numbers of one to several defects in the cluster to all the defects in the cluster described with reference to FIG.
The problem arises that it has to be performed separately.

Therefore, based on the coordinates of the individual defects included in the clusters of the process A and the process B, the area where the cluster exists is determined as, for example, the maximum value / minimum value of the defect coordinates in the cluster, or the entire wafer is several tens to When divided into blocks each having a size of about several hundred mm ² , this is expressed as an address of a block in which an intra-cluster defect exists, and there is a portion where the region where the cluster exists in the process A coincides with the region where the defect exists in the cluster in the process B Is the same review and defect as the defect included in the process A cluster for all the defects included in the process B cluster.
A process of giving an analysis multi-value classification number may be added.

As a result, since the sensitivity of the inspection apparatus WI is different between the processes, even when a cluster which should be detected in the previous process is detected in the subsequent process, the defect of the newly detected cluster is not affected. Thus, it is possible to assign a review / analysis multi-value classification number of the previous cluster generation process without review / analysis.

FIG. 19 is a diagram showing the transition of the number of detected defects according to the progress of the semiconductor manufacturing process. As shown in the figure, although the total number of defects tends to increase as the process progresses as a whole, the sensitivity of the inspection apparatus WI differs between processes as described above, or
The type (category) of the defect is the attachment of foreign matter such as dust, and the foreign matter is washed away by washing in the next step, so that the number of detected defects in the next step is smaller than the total number of detected defects in the previous step. May decrease.

Next, another example of processing for selecting a defect to be reviewed / analyzed by the inspection data processing apparatus 1 according to the third embodiment will be described. Although the components shown in FIG. 9 have already been described, in the inspection data processing apparatus 1 after the above-described review, the defect category information manually added by the operator during the review and the information on the review station RS The defect category information automatically added by the software or the like is stored in the inspection data storage unit 21 one by one or collectively at the end of the review.

Next, another process of selecting a defect to be reviewed / analyzed according to this embodiment will be described with reference to the flowchart of FIG.

It is assumed that defect category information is stored in the inspection data storage unit 21 before this processing. In this case, it is assumed that the inspection device W1 has a function of automatically determining the category of the defect based on the color, shape, and the like of the defect (for example, the review / analysis by the review station RS is fed back to the inspection device W1). By doing so, it is assumed that the inspection apparatus W1 has a function of roughly determining the category of a defect.) Of course, once the review is performed, the processing flow of FIG. 20 may be executed again for the same wafer in the same process.

First, in step S501, an inspection process to be reviewed / analyzed and the number _Nmax of defects selected for review / analysis are input from the display unit 11 (the display unit 11 is assumed to have an input function unit). , Data input / output unit 4
Through the memory 6.

Next, in step S502, the search unit 5 searches the memory 6 for the defect coordinates, defect size, and class value C stored in the inspection data storage unit 21 from the inspection process input in step S501. .

Further, in step S503, the judgment data (judgment rule data) corresponding to each class value C read in step S502 is stored in the judgment data storage unit 24.
Is retrieved by the retrieval unit 5 and read into the memory 6.

Next, in step S504, for each defect, a defect size threshold R corresponding to the defect size and the class value C is set.
A size threshold ratio expressed by the above equation, which is a ratio to _C , is obtained, and this is defined as a first fatality.

Next, in step 505, the first fatality obtained in step 504 and the inspection data
The second fatality is determined based on the predetermined determination rule data using the defect category read from.

Then, in step 506, the defects are rearranged in descending order of the second criticality obtained in step 505, and then in step S507, the defects are sorted in step S5.
06, select N _max defects from the top in the order after processing,
In step 508, the selected defect is written to the review / analysis target data storage unit 25.

Here, the judgment rule data used in the processing of the above step 505 is, for example, such that a predetermined weight is applied to the size threshold ratio represented by the above equation according to each defect category. I do.

The fatality (first and second fatality) is as follows.
As will be described later, it is also possible to calculate this from a probe test result using a function, a look-up table, or the like so as to more accurately represent the influence on the circuit pattern.

The second criticality obtained in the above processing flow is stored in the inspection data storage unit 21 together with the defect information.

Therefore, the search unit 5 determines that the first fatality and the second
Using the defect information such as the criticality of the defect and the size of the defect as a retrieval key, data can be retrieved and significant information can be extracted. The display unit 11 is connected to the data input / output unit 4, and includes an inspection data storage unit 21, a chip area image data storage unit 22, a chip area coordinate data storage unit 23, and a determination data storage unit 2.
4. The data in the review / analysis target data storage unit 25 is displayed. For example, image data of a detected defect can be displayed.

Therefore, based on the extracted data,
The display unit 11 can selectively display a distribution map of highly critical defects or a defect image. The communication control unit 3 can output information to the outside via the network Nt. The image of the defect to be displayed has improved criticality accuracy based on the use of the defect category, and provides more useful data. In addition, in the case of a categorization in which the review work is performed automatically using the detected images, only the more fatal images are extracted and displayed to the worker, thereby preventing unnecessary review work. it can. This leads to a reduction in work time and work load.

As another effect, the inspection conditions of the inspection device WI can be set and the inspection conditions can be compared so that the number of defects with high fatality can be ensured. I can plan. For example, the comparison result of the detected defects is displayed on the display unit 11 as a Venn diagram.
Is output to Also, a plurality of inspection devices WI are connected,
When used, these machine differences can be visually grasped using a Venn diagram or the like. Further, when different types of inspection devices, for example, a foreign material inspection device and a defect inspection device, are used, the difference between these performances can be objectively evaluated using a criticality scale, and the inspection devices can be used properly.

Further, when the wafer W is formed of a film forming apparatus, an exposure apparatus,
When it is processed with an etching apparatus and finally reaches an electrical inspection called a probe inspection, electrical characteristics and the like of each chip are determined. Therefore, if the probe test data is compared with the data stored in the test data storage unit 21 to determine the difference in the fatality, the process of deriving the fatality can be corrected, so that more accurate fatality determination can be performed.

As described above, the inspection data storage unit 21 can be used as an important defect database to search its contents and edit data. By confirming these results on the display unit 11, it is possible to take measures for the process and the manufacturing apparatus. It can be done quickly.

In this case, it is also effective to set the selection conditions from the outside or to set the conditions internally beforehand, and to extract defects that meet the selection conditions.

For example, selection conditions include: f) defect size g) defect coordinates h) fatality i) chip identification coordinates indicating which chip on a wafer j) chip coordinates indicating an area within each chip k) defect There are categories. The selection condition here is basically that the fatality of a defect (including a foreign substance) is determined by its category and the area of the wafer.

In addition, the inspection data processing apparatus 1 according to the present embodiment for performing defect fatality determination, selecting a defect to be reviewed / analyzed, and editing / searching inspection data is provided to the inspection apparatus WI without using the network Nt. May be incorporated. In this case, the details of the defect information can be grasped on the inspection apparatus WI in a state where the wafer W is mounted.

As described above, the inspection data processing apparatus 1 of the present embodiment uses the input inspection data having chip identification coordinates, chip coordinates, defect coordinates, fatality, defect size, defect category, and the like. The review / analysis order and the review / analysis target can be edited. Therefore, a review / analysis system capable of controlling the review / analysis order and the review / analysis target can be constructed. The review station RS in this review and analysis system is SE
It may be M. In addition, by providing an editable viewer, such as browsing the images detected by the review station RS and numbering, the user can freely edit the review results and use it for report creation. Further, even when a cross-sectional photograph of a defective portion is taken by an SEM, data can be more efficiently narrowed down by using the above-described inspection data processing method. For example, a cross-sectional photograph is obtained by moving a stage to a defective portion in a FIB (Focued Ion Beam) analyzer, irradiating the defective portion with FIB or the like, shaving it in half, and taking a SEM photograph so that the cross section can be observed. can get. Of course, the present invention can be applied to effective use of an analyzer such as an elemental analyzer.

Here, an example of determining the inspection conditions of the inspection apparatus W1 will be described with reference to FIG. FIG. 18 is a graph plotting the defect area on the horizontal axis and the number of detected defects for each defect area on the vertical axis.

In general, the smaller the defect area, the larger the number of detected defects. However, the inventors of the present application have observed in detail that it has been confirmed that the smaller the defect area is, the smaller the number of detected defects is. In FIG. 18, the number of detected defects decreases below 0.25 μm square. This means that the defect detection sensitivity of the inspection apparatus W1 gradually decreases with respect to minute defects, which means that inspection reproducibility cannot be ensured for these minute defects. In addition, it is highly likely that these non-reproducible defects are actually erroneous detections of normal patterns by the inspection apparatus W1.

Therefore, a detection defect having a small number of detections and having no reproducibility can be regarded as an erroneous detection, and it is possible to set an inspection condition in which this is not detected. For example, by outputting a defect having a defect size equal to or larger than a predetermined value, for example, a defect having a size of 0.25 μm square or more, such erroneous detection can be eliminated. Also, even if the binarization threshold value at the time of comparison is controlled, substantially the same effect can be obtained. As described above, by having the area data and the size data of the detected defect in the semiconductor pattern and the data of the number of defects, it is possible to determine the inspection condition from the data of the number of different sizes.

In the above-described embodiment, a semiconductor wafer has been described as an example of an object to be inspected.
The present invention can be applied to inspection of other inspection objects such as a liquid crystal substrate having a circuit pattern. It goes without saying that various modifications can be made without departing from the spirit of the present invention.

[0143]

The effects obtained by the representative inventions among the inventions disclosed in the present application are listed below.

(1) Criticality of a defect detected by the inspection device can be automatically determined without re-enlarging an enlarged image with a metallographic microscope or a scanning electron microscope. As a result, it is possible to determine the fatality of all the detected defects. Therefore, as shown in FIG.
It is possible to accurately detect abnormalities correlated with the yield, and to take measures for improving the yield at an early stage.

(2) In addition, it becomes possible to automatically select defects to be reviewed in descending order of fatality, thereby enabling efficient review and analysis.

(3) Further, there is no waste in selecting defects that have already been reviewed in the previous step again, or there is no waste in selecting a plurality of defects belonging to the same density and having the same classification as objects to be reviewed. Defects generated in the previous process due to the congestion defects generated in the previous process can be omitted from the review target.

(4) Further, (1), (2),
According to (3), it is possible to automate the criticality judgment of the review work, and to improve the throughput by increasing the efficiency of the review classification work for classifying foreign matters, pattern defects, and the like.

(5) Further, by using the defect category, the accuracy of the defect fatality determination can be further improved, and more accurate management based on the number of fatal defects can be realized.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing a configuration of a system using a test data processing device according to a first embodiment of the present invention.

FIG. 2 is a flowchart illustrating a processing flow of inspection data according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram showing an example of chip area data used in the embodiment of the present invention.

FIG. 4 is an explanatory diagram showing an example of chip area coordinate data used in the embodiment of the present invention.

FIG. 5 shows a class value C used in an embodiment of the present invention.
FIG. 7 is an explanatory diagram showing an example of criticality determination data (defect size threshold value) corresponding thereto.

FIG. 6 is an explanatory diagram showing a configuration of a system using a test data processing device according to a second embodiment of the present invention.

FIG. 7 is a flowchart illustrating a processing procedure for creating area image data used in the second embodiment of the present invention.

FIG. 8 is a flowchart illustrating a processing flow of inspection data according to the second embodiment of the present invention.

FIG. 9 is an explanatory diagram showing a configuration of a system using a test data processing device according to a third embodiment of the present invention.

FIG. 10 is a flowchart illustrating an example of a processing flow of inspection data according to the third embodiment of the present invention.

FIG. 11 is an explanatory diagram showing the transition of the number of defects for each generation step.

FIG. 12 is an explanatory diagram schematically showing a defect in a current process, a defect in a previous process, and a state after deleting a defect in a previous process from a defect in the current process.

FIG. 13 is an explanatory diagram showing an example of a cluster defect and a defect to be reviewed / analyzed in the cluster.

FIG. 14 is an explanatory diagram schematically showing a cluster of a current process B, a cluster of a process B before the cluster, and a cluster after a defect of a previous process is deleted from a defect of the current process.

FIG. 15 shows the correlation between the total number of defects and the yield, and
It is explanatory drawing which shows management by the total defect number.

FIG. 16 is an explanatory diagram showing a state of correlation between the number of fatal defects and a yield, and management based on the number of fatal defects.

FIG. 17 is an explanatory diagram showing an example of an important defect extraction condition according to the embodiment of the present invention.

FIG. 18 is an explanatory diagram showing the distribution of the defect area and the number of defects.

FIG. 19 is an explanatory diagram showing transition of the number of defects detected according to the progress of the manufacturing process.

FIG. 20 is a flowchart illustrating another example of the processing flow of inspection data in the third embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 inspection data processing device 3 communication control unit 4 data input / output unit 5 search unit 6 memory 7 input / output interface 8 program storage unit 9 calculation unit 10 main control unit 11 display unit 21 inspection data storage unit 22 chip area image data storage unit 23 Chip area coordinate data storage unit 24 Judgment data storage unit 25 Review / analysis target data storage unit Nt Network WI inspection device RS Review station

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Kenji Oka 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture F-term in Hitachi, Ltd. Production Engineering Research Laboratory F-term (reference) 2G014 AA02 AA03 AA25 AB59 AC09 AC11 AC15 2G051 AA51 AA56 AA73 AB02 EA14 EA20 EB01 EB02 EC01 ED07 FA10 4M106 AA01 CA38 CA39 CA40 DB01 DB04 DH07 DJ11 DJ18 DJ20 DJ21 DJ23 DJ26 DJ27 DJ28 DJ38

Claims (31)

[Claims] 1. Area coordinate data on a surface of an object to be inspected having a circuit pattern, criticality determination data predetermined for each area of the surface of the object to be inspected, coordinate data and characteristics of a detected defect Obtaining the quantity data; determining the fatality of the defect from the fatality determination data corresponding to the defect existing area and the feature amount data of the defect; and expressing a degree of the fatality of each defect. Writing information to an inspection database. 2. The inspection data processing method according to claim 1, wherein the area where the defect exists is determined by dividing the area of the inspection object surface into pixels and assigning a class value corresponding to the area to each pixel. Generating area image data, converting the coordinate data of the detected defect into an address of a pixel in the area image, and identifying a defect existing area from the class value of the pixel at the pixel address corresponding to the detected defect. And a step of processing the inspection data. 3. Area coordinate data on a surface of an object to be inspected having a circuit pattern, fatality determination data predetermined for each region of the surface of the object to be inspected, coordinate data and characteristics of a detected defect. Obtaining quantity data; determining the fatality of the defect from the criticality determination data corresponding to the defect existing area and the size feature amount of the defect; and determining the criticality of the defect based on the determined criticality. Selecting the defects to be reviewed / analyzed in descending order of the degree of fatality until the total number of defects to be reviewed / analyzed reaches a preset total number of defects. 4. The inspection data processing method according to claim 3, wherein the area where the defect exists is determined by dividing the area of the inspection object surface into pixels and assigning a class value corresponding to the area to each pixel. Generating area image data, converting the coordinate data of the detected defect into an address of a pixel in the area image, and identifying a defect existing area from the class value of the pixel at the pixel address corresponding to the detected defect. And a step of processing the inspection data. 5. The inspection data processing method according to claim 1, wherein the coordinate data of a detected defect in a step before or a step before the current step to be inspected, Comparing the coordinate data of the detected defect in the step with the coordinate data of the detected defect, and selecting only the detected defect in the current step which does not match the inspection data. 6. The inspection data processing method according to claim 1, further comprising the step of: classifying the defect data into a dense defect and a random defect based on coordinate data of the detected defect; Selecting the defect of the inspection data. 7. The inspection data processing method according to claim 1, wherein a previous step or a previous step in a current step to be inspected in a manufacturing process of the inspection object. An inspection data processing method, comprising the step of classifying a detected defect that does not match the detected defect in (1) into a dense defect and a random defect from the coordinate data of the detected defect. 8. The inspection data processing method according to claim 7, wherein the coordinate data of the detected defect included in the congestion defect in the previous step or the previous step and the congestion defect in the current step are determined. Compared with the coordinate data of the detected defects included, if one or more detected defects match, the dense defect in the previous step or the previous step and the dense defect in the current step described above and A step of classifying all defects included in the inspection data into the same classification. 9. Area coordinate data on a surface of an object to be inspected having a circuit pattern, fatality determination data predetermined for each region of the surface of the object to be inspected, coordinate data and characteristics of a detected defect. Obtaining quantity data and category data of the defect; determining a first fatality of the defect from the fatality determination data corresponding to the area where the defect is present and the feature amount data of the defect; Determining a second criticality of the defect from the determined first criticality and the category data of the defect; and writing information indicating the degree of the second criticality of the determined defect to an inspection database. And a step of processing the inspection data. 10. The inspection data processing method according to claim 1, wherein the feature data of the detected defect is defect size data represented by a length or an area of the defect. Inspection data processing method. 11. The inspection data processing method according to claim 1, wherein the detected defect size data and the defect number data of the detected defect on the surface of the inspection object are used to determine the number of detected defect size data. An inspection data processing method comprising: determining the size of a defect to be output as a detected defect from the size-separated number data; and outputting a defect having a size equal to or larger than a predetermined size as a detected defect. 12. The inspection data processing method according to claim 1, wherein the coordinate data of the area of the surface of the inspection object, the coordinate data of the detected defect, and the feature amount of the detected defect written in the inspection database. An inspection data processing method, wherein inspection data is searched or edited based on data, fatality of detected defects, and categories of detected defects. 13. The test data processing method according to claim 1, wherein a fatality determination procedure is corrected based on a final probe test result. 14. A first storage means for storing coordinate data and feature amount data of a defect detected by an inspection apparatus for inspecting an inspection object surface having a circuit pattern, and storing area coordinate data on the inspection object surface. A second storage unit for storing, and a third storage unit for storing predetermined fatality determination data according to each region of the object surface to be inspected.
Storage means, first control means for identifying a defect existing area, and second control for judging the fatality of a defect from the characteristic amount of the defect and the fatality determination data corresponding to the defect existing area. And a data input / output unit for writing information indicating the degree of fatality of the defect into the first storage unit. 15. The test data processing apparatus according to claim 14, further comprising a communication control unit that exchanges data with the test apparatus via a network. 16. The inspection data processing apparatus according to claim 14, wherein the area of the inspection object surface is divided into pixels, and a class value corresponding to the area is assigned to each pixel, thereby generating area image data. A first calculating means, a fourth storing means for storing the area image data generated by the first calculating means, and a second calculating means for converting the coordinate data of the detected defect into an address of a pixel in the area image. An inspection data processing apparatus comprising: 17. A first storage unit for storing coordinate data and feature amount data of a defect detected by an inspection apparatus for inspecting an inspection object surface having a circuit pattern, and storing area coordinate data on the inspection object surface. Second storage means for storing; third storage means for storing predetermined fatality determination data in accordance with each area of the object surface to be inspected; Control means; second control means for judging the fatality of the defect from the feature amount of the defect and the fatality judgment data corresponding to the area where the defect is present; a review / analysis object based on the judged fatality An inspection data processing apparatus comprising: a third control unit that selects a defect to be a defect in descending order of the degree of criticality until the total number of defects to be reviewed and analyzed reaches a preset total number. 18. The test data processing device according to claim 17, further comprising a communication control unit that exchanges data with the test device via a network. 19. The inspection data processing apparatus according to claim 17, wherein the area of the inspection object surface is divided into pixels, and a class value corresponding to the area is assigned to each pixel, thereby generating area image data. A first calculating means, a fourth storing means for storing the area image data generated by the first calculating means, and a second calculating means for converting the coordinate data of the detected defect into an address of a pixel in the area image. An inspection data processing apparatus comprising: 20. The inspection data processing apparatus according to claim 14, wherein coordinate data of a detected defect in a step before or up to a current step to be inspected, An inspection data processing apparatus comprising: fourth control means for comparing the coordinate data of the detected defect in the step with the coordinate data of the detected defect and selecting only the detected defect in the current step which does not match. 21. The inspection data processing apparatus according to claim 14, further comprising: fifth control means for classifying the defect data into a dense defect and a random defect based on the coordinate data of the detected defect. Inspection data processing device. 22. The inspection data processing apparatus according to claim 14, wherein coordinate data of a detected defect in a process before or up to a current process to be inspected, A fourth control means for comparing the coordinate data of the detected defect in the step with the coordinate data of the detected defect selected by the fourth control means; An inspection data processing apparatus comprising: fifth control means for classifying the defect into a cluster defect and a random defect. 23. The method according to claim 22, wherein the coordinate data of the detected defect included in the dense defect in the previous step or the previous step and the detected data of the detected defect included in the dense defect in the current step are set. Compared with the coordinate data, if one or more detected defects coincide with each other, all the defects included in the dense defect in the previous step or the previous step and the dense defects in the current step described above are included. An inspection data processing apparatus comprising: sixth control means for classifying defects into the same classification. 24. A storage means for storing coordinate data and characteristic amount data of a defect detected by an inspection apparatus for inspecting an inspection object surface having a circuit pattern, and a storage for storing area coordinate data on the inspection object surface. Means, storage means for storing predetermined fatality determination data according to each area of the object surface to be inspected, storage means for storing category data of detected defects, and identification of a defect existence area Control means; control means for determining a first fatality of the defect from the feature amount of the defect and the fatality determination data corresponding to the area where the defect exists; and the determined first fatality and the category data. Control means for judging the first fatality of the defect from the above, and data input / output means for writing information indicating the degree of the second fatality of the judged defect to the storage means. Test data processing apparatus characterized by. 25. The inspection data processing apparatus according to claim 14, wherein the feature data of the detected defect is defect size data represented by a length or an area of the defect. Inspection data processing device. 26. The inspection data processing apparatus according to claim 14, wherein: means for searching and editing inspection data; means for displaying the searched and edited data; and data for searching and editing. And a means for externally outputting the data. 27. The inspection data processing apparatus according to claim 14, wherein an input to a review / analysis apparatus can be controlled using the inspection data. apparatus. 28. The inspection data processing device according to claim 27, wherein the review / analysis device is an SEM. 29. The inspection data processing apparatus according to claim 27, further comprising means for controlling a review / analysis order in said review / analysis apparatus and for editing a review / analysis object. apparatus. 30. The inspection data processing device according to claim 27, wherein an image detected by the review / analysis device is browsed,
An inspection data processing device comprising means for performing editing such as numbering. 31. The test data processing apparatus according to claim 14, further comprising means for correcting a fatality determination procedure based on a final probe test result. Inspection data processing device.