JP3006166B2 - Data analysis method, inspection system, and method of manufacturing thin film product - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of inspecting a thin film product, for example, a semiconductor device, a thin film circuit board, a magnetic disk, a thin film transistor, etc., for an appearance defect such as a foreign matter, a defect or a film defect, and a method of inspecting the appearance defect. The present invention relates to a system for analyzing analysis data to analyze a failure occurrence situation and a failure occurrence cause, and a method of manufacturing a thin film product using the above inspection method and system.

[0002]

2. Description of the Related Art In the production of thin film products, the occurrence of foreign matter, defects, film defects, etc. (hereinafter collectively referred to as appearance defects) is strictly controlled because it causes product defects. Inspections are performed on all or all of the workpieces that run through the production line to process products (hereinafter abbreviated as product workpieces), and the number of appearance defects, type, size, etc. Recorded and managed according to certain management standards. If it is determined that the appearance defect is abnormal based on the management standards, a typical work is extracted, the components are analyzed for some or all of the appearance defects in the work, and the cause of the appearance defect from the components is analyzed. To explore. At this time, the case where the product work is directly inspected and / or analyzed, and the case where only the process which seems to have an abnormality (hereinafter referred to as dummy work) are inspected and / or analyzed.
Or it may be analyzed. Since the cause of the occurrence is considered to be maintenance failure of the device, the content of the device maintenance is examined. An example of the analysis is described in “Semicon NEWS” published by Marcom International, November 1988, pp. 31-39. As an apparatus used for analysis, an apparatus described in JP-A-60-218845 has been devised. Since such an analysis is performed for comparison between different types of data (between the appearance inspection and analysis data, and between the appearance inspection data and the device maintenance data), it is premised that the analysis is performed in advance and is performed for a specific lot, process, or device. On the other hand, appearance inspection, component analysis, and collection of equipment maintenance information had to be performed.

[0003]

However, appearance inspection, component analysis, and collection of equipment maintenance data are performed daily at the production site, and appearance inspection data and component analysis generated from unspecified lots, processes, and equipment every day. In order to perform an analysis using data and device maintenance data, key data required for collation between respective data was not clarified, and collation between respective data could not be performed. Further, means for collating each data has not been standardized. For this reason, it has been difficult to know the apparatus in a process in which an abnormality is considered to occur, and to compare maintenance data of the apparatus with appearance inspection data or analysis data of a product work or a dummy work processed by the apparatus. In addition, the date and time when the visual inspection and / or analysis was performed and the date and time when the device maintenance was performed are often not recorded, and the temporal order of the visual inspection, analysis, and device maintenance could not be determined. The causal relationship between the test results and the analysis results cannot be analyzed.

[0004] It is an object of the present invention to efficiently analyze the cause of appearance failure by comparing various data such as appearance inspection data, analysis data, and apparatus maintenance data in a time series.

[0005]

In order to achieve the above object, the present invention provides a data analysis method for analyzing a cause of a defect occurring in a work in a process of processing the work. The inspection result is output in association with information on a time at which the work was inspected and information on a maintenance time of a processing apparatus that processed the work. In order to achieve the above object, according to the present invention, in a data analysis method for analyzing a cause of a defect generated in a work in a step of processing the work, an inspection result obtained by inspecting the work is used. Is output in association with the information on the inspection time and the information on the result of analyzing the defect generated in the work. Further, in order to achieve the above object, in the present invention,
In a data analysis method for analyzing a cause of a defect occurring in a work in a process of processing the work, an inspection result obtained by inspecting the work is converted into information on a time when the work is inspected, The information is output in association with the information on the time when the apparatus is maintained and the information on the result of analyzing the defect generated in the work. Further, in order to achieve the above object, according to the present invention, in an inspection system including an inspection device for inspecting a workpiece having undergone a predetermined processing step and an analysis device for analyzing data of an inspection result of the inspection device, The analysis device obtains an inspection result obtained by inspecting the work with the inspection device, and relates information on a time when the inspection is performed on the work with the inspection device and a time when the processing device that processes the work in the predetermined processing step is maintained. A processing unit for performing a process of associating with the information; displaying the inspection result, the information on the time when the work was inspected, and the information on the time when the processing device was maintained on the same screen on the same screen; Display means. Further, in order to achieve the above object, the present invention includes an inspection device that inspects a workpiece that has undergone a predetermined processing step to detect a defect, and an analysis device that analyzes data of a result of inspection performed by the inspection device. In the inspection system, the analysis device analyzes the inspection result obtained by inspecting the work with the inspection device, analyzes information regarding a time when the inspection is performed on the work by the inspection device, and analyzes a defect of the work detected by the inspection device. Processing means for performing a process of associating with the result information, the inspection result processed by the processing means and the information relating to the time when the work was inspected and the information of the result of analyzing the defect of the work are displayed on the same screen. And a display means for displaying the information. Further, in order to achieve the above object, in the present invention,
In an inspection system including an inspection device that inspects a workpiece that has undergone a predetermined processing process to detect a defect and an analysis device that analyzes data of a result of inspection performed by the inspection device, the analysis device includes: Inspection results obtained by inspecting, the information on the time when the work was inspected by the inspection device, the information on the time when the processing device that processes the work in the predetermined processing step was maintained, and the inspection device detected by the inspection device Processing means for performing a process of associating the work with the information of the result of analyzing the defect of the work; information on the inspection result and the time when the work was inspected, which was processed and associated by the processing means; and information on the time when the processing apparatus was maintained. And display means for displaying, on the same screen, information on the result of analyzing the work defect. Further, in order to achieve the above object, according to the present invention, in a thin film product manufacturing method for manufacturing a thin film product by sequentially processing the work in a plurality of processing steps, the appearance of the work sequentially processed in a predetermined processing step is inspected. Then, the result of the inspection of the appearance is stored in association with the information on the inspection time and the information on the maintenance time of the processing device for sequentially processing the workpieces in the predetermined processing step, and the associated stored appearance is inspected. Information on the result of the inspection and the time when the inspection was performed and information on the time when the processing apparatus was maintained are displayed on the screen, and when a defect occurs in the appearance of the workpiece processed in the predetermined processing step, the information is displayed on the screen. The cause of the occurrence of a defect in the predetermined processing step is analyzed based on the obtained information. Further, in order to achieve the above object, according to the present invention, in a method of manufacturing a thin film product in which a work is sequentially processed in a plurality of processing steps to manufacture a thin film product, the appearance of the work processed in a predetermined processing step is inspected. And detects the defect of the work, stores the result of the appearance inspection in association with the information on the inspection time and the information of the result of analyzing the defect generated in the work, and inspects the associated stored appearance. The result of the inspection, the information on the inspection time, and the information on the result of analyzing the defect generated in the work are displayed on the screen, and when a defect occurs in the appearance of the work processed in the predetermined processing step, the screen is displayed on the screen. The cause of the occurrence of a defect in the predetermined processing step is analyzed based on the information displayed in (1).

[0006]

As described above, it is possible to efficiently analyze the cause of the appearance defect by comparing various data such as the appearance inspection data, the analysis data, and the apparatus maintenance data in a time series. For example, by giving the inspection date and time to the appearance inspection data, the analysis date and time to the analysis data, and the maintenance date and time to the device maintenance data, the order in which each data occurred can be understood so that the causal relationship between the device maintenance and the appearance inspection and analysis can be grasped. Now you can. In addition, means for arranging appearance inspection results in chronological order for each process to determine whether the appearance defect is within the control standards, and a means for determining the appearance defect on the workpiece processed in the corresponding process during the period when the appearance defect is not within the control standards. By providing the means for displaying the analysis result of (1), it is possible to search for the analysis result of the appearance defect indicating that the appearance defect is abnormal. Means for arranging the appearance inspection results in time series for each process to determine whether or not the appearance defect is within the control standards, means for knowing a device to perform processing in this process, and a device for immediately preceding this period in the corresponding process By providing the means for displaying the contents of the maintenance, it is possible to analyze the causal relationship between the result of the apparatus maintenance and the occurrence of the appearance defect.

[0007]

An embodiment of the present invention will be described with reference to the drawings.

Embodiment 1 FIG. 1 shows an outline of a data structure and data processing contents in the present invention. The appearance inspection data group 1 is data relating to the appearance inspection data. The appearance inspection data group 1 includes a lot number, a work number, a process name, a kind, an inspection date and time, and an appearance inspection result. The appearance inspection result includes the position, size, and type of the appearance defect on the work. In the visual inspection data group 1, the lot number and the process name are used as a data identifier. The content of the data identifier does not exist in the same data group. Usually, the type, lot number and work number are not assigned to the dummy work, so when managing the appearance inspection data on the dummy work, the type should be “dummy” and the lot number and work number should be serial numbers in the dummy work. And make it easy to distinguish it from product work. The analysis data group 2 is a data group related to the analysis data. Analysis data group includes lot number, work number, process name,
Includes analysis date and time, analysis result. The analysis result consists of the position of the appearance defect and the component spectrum. Instead of the component spectrum, a component element or a component group obtained by analyzing the component spectrum may be used. The analysis data group is a data identifier by combining the process name and the analysis date and time. The process equipment corresponding data group is data relating to the process equipment corresponding data. In the group of process apparatus corresponding data, the process name becomes a data identifier. The device maintenance data group is data relating to device maintenance data. The device maintenance data group includes a device name, a maintenance date and time, and maintenance contents. The name of the maintenance worker may be included. The device maintenance data group is used as a data identifier by combining the device name and the maintenance date and time. The process 100 searches the appearance inspection data group 1 for appearance inspection data for a desired step and a desired inspection period,
The visual inspection results are arranged in chronological order and displayed. At this time, the management standard is also displayed, and it is easy to determine whether the inspection result is within the management range. When there is a period in which the inspection result is out of the management range, the analysis operator specifies the period.
The process 100 passes the process name, the inspection period, and this period to the process 101 or 102. In process 101, process 1
The analysis data group corresponding to the process name and the period passed from 00 is searched for the analysis data group, and the analysis data is analyzed and displayed. The process 101 transfers the process name and the period to the process 102 as necessary. In the process 102, the process name and the period are delivered from the process 100 or the process 101. Based on the process name, the device maintenance data group is searched to extract the device name corresponding to the process name. Further, the process 102 extracts and displays the maintenance content corresponding to the device name and the period.

The visual inspection data group 1 is stored in the order of the inspection date and time, the analysis data group is stored in the order of the analysis date and time, and the processing apparatus correspondence data group is stored in the order of the oldest or newest according to the maintenance date and time.

<Embodiment 2> The appearance inspection data group 1, the analysis data group 2, the process apparatus correspondence data group 3,
The respective structures of the device maintenance data group 4 will be described. FIG. 2 shows the structure of the appearance inspection data group 1. The appearance inspection data group 1 includes a lot data table 501, a work data table 502, and a failure data table 503. The lot data table 501 has a lot number, a process name, a lot number + a process name, a product type, and an inspection date and time. The lot data table has the number of lots whose appearance has been inspected.
The lot number + process name is a data identifier in the visual inspection data group 1. One lot data table has work data tables as many as the number of works constituting the lot. The work data table 502 has a lot number + process name, a work number, a lot number + work number + process name, the total number of appearance defects, and the number of types of appearance defects. For each work data table, there are defective data tables for the number of appearance defects on the work. The defect data table 503 has lot number + work number + process name, defect coordinates, size, and type.

FIG. 3 shows the structure of the analysis data group 2. The analysis data group 2 has an analysis work data table 504 and an analysis data table 505. The analysis work data table 504 includes a process name, a lot number, a work number, an analysis date and time,
It has a process name + analysis date and time. The process name + analysis date / time is a data identifier in the analysis data group 2. The number of analysis work data tables is equal to the number of works analyzed. One analysis work data table has analysis data tables for the number of appearance defects analyzed on the work. The analysis data table 505 has a process name + analysis date / time, defective coordinates, and a component spectrum. The component spectrum may be held as image data as shown in FIG. The coordinate system used for analysis and the coordinate system used for appearance inspection are shared.

FIG. 4 shows the structure of the data group 3 corresponding to the process equipment. The process device correspondence data group is a process device correspondence table 50.
Have 6. The process device correspondence table 506 has a process name and a device name. The process device correspondence table has the same number as the number of processes to be managed in the present invention. A plurality of device names may correspond to one process name. The process name is a data identifier in the group of process device data.

FIG. 5 shows the structure of the device maintenance data group 4. The device maintenance data group has a maintenance data table 507. The maintenance data table has a device name, a maintenance date and time, a device name + maintenance date and time, a maintenance content, and a maintenance worker. The device name + maintenance date is a data identifier in the device maintenance data group.

<Embodiment 3> Processing 10 described in Embodiment 1
The analysis and display of data at 0 will be described in detail. The analysis of the data in process 100 aims to determine whether the number of appearance defects is within the management range. FIG. 6 shows the output format (transition defect diagram), and FIG. 7 shows the processing contents. In the process 100, the analysis operator inputs a desired process name and an inspection period to the present system (step 100).
1). Next, data of the appearance inspection result corresponding to the process name and the period is retrieved from the appearance inspection data group 1 (step 10).
02). Next, lot numbers are arranged in chronological order according to the inspection date and time, and an appearance defect transition diagram is drawn with the horizontal axis 11 and the vertical axis 12 being the number of appearance defects (step 1003). FIG. 6 shows the appearance transition chart. The number of appearance defects on the vertical axis 12 indicates an average value in a lot per work. Next, the control upper limit 13 is also described (step 1004), and it is determined whether there is a lot exceeding the control upper limit (step 1005). If there is, the inspection period and the inspection date and time of the lot are output together with the process name (step 1006), and the process ends. If not, the process ends. Here, when there are a plurality of lots, the inspection date and time of the first lot and the last lot exceeding the upper management limit are output. Hereinafter, when simply referred to as a period, it is a period between the inspection date and time of the first lot and the last lot exceeding the upper management limit.

<Embodiment 4> Processing 10 described in Embodiment 1
1 will be described in detail. The analysis of the data in the process 101 is to analyze the analysis data of the appearance defect in the lot (including the lot including the dummy work) processed in the same process as the lot or the lot exceeding the upper control limit in the process 100 in the same period. It is in. It is another object of the present invention to analyze whether there is a change in a component having poor appearance before and after a lot exceeding a control upper limit value.

FIG. 8 shows the contents of the process 101. Processing 10
The process name and the period output at 0 are read (step 10
07). Next, an analysis spectrum of the appearance defect corresponding to the process name and the period is retrieved from the analysis data group 2, and the analysis spectrum is displayed (step 1008). When there are a plurality of analysis spectra, the data may be output in the order in which the data was searched. FIG. 9 shows an output example of the analysis spectrum. The analysis spectrum is a mass spectrum. The horizontal axis 14 is the amount obtained by dividing the mass m by the electric charge z, and the vertical axis 15 is the intensity, which is a count per unit time. Next, in the process, analysis data before the inspection date and time is searched and displayed (step 1009), and the process ends. The display format is as shown in FIG. The analysis operator analyzes how the components of the appearance defect have changed during the period and before the period, and examines the components of the appearance defect that caused the number of the appearance defects to exceed the upper management limit.

<Embodiment 5> Processing 10 described in Embodiment 1
The data analysis and display in 2 will be described in detail. The purpose of the data analysis in the process 102 is to analyze what kind of maintenance has been performed in the apparatus constituting the process immediately before the number of appearance defects exceeds the management upper limit in the process 100. If the time when the number of appearance defects increases and the time when maintenance is performed coincide with each other, the analysis worker reviews the maintenance items, considering that maintenance is the cause of the increase in appearance defects.

FIG. 10 shows the contents of the process 102. Processing 1
The process name, inspection period and appearance inspection result output at 00 are read (step 1010). Next, a device name corresponding to the process name is searched from the process device corresponding data group (step 1).
011). When there are a plurality of device names, the analysis operator selects only one (step 1012). Next, a maintenance content and a maintenance worker name corresponding to the inspection period and the device name output in the process 100 are searched from the device maintenance data group (step 1013). Next, an analysis result is output using the appearance failure device maintenance correspondence graph shown in FIG.
(Step 1014) The process ends. The horizontal axis 16 indicates the date and time, and shows the appearance inspection date and time and the maintenance date and time. Vertical axis 17 of transition chart
Is the number of appearance defects, and is the average value in a lot per work as in FIG. The time when the maintenance was performed is shown as a skewer 17 below the transition diagram. The maintenance content 19 is shown beside the skewer 18. When there are a plurality of maintenance contents, one skewer is prepared for one maintenance
It is sufficient to write them together as shown in FIG. A skewer 17-shaped dot 120 indicates the maintenance date and time when the maintenance content 18 was performed. The analysis operator can compare the change in the appearance defect and the maintenance content with reference to FIG. In the example shown in FIG. 11, it can be seen that the number of appearance defects has increased after the time point 21 when the bulb was replaced.

[0019]

According to the present invention, the appearance inspection result and the analysis result,
The result of the appearance inspection can be easily compared with the contents of the device maintenance, and the cause of the increase in the appearance defect can be easily analyzed. Further, by managing the data in the data format shown in the present invention, the redundancy of the data amount is eliminated, and each data item in the appearance inspection result, the analysis result, the device maintenance content, and the process device corresponding data is efficiently processed. You can now refer to each other. Further, by managing data for associating the process with the apparatus in a table different from the apparatus maintenance data, the data can be easily changed, added, or updated.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing an outline of processing contents and data items of the present invention,

FIG. 2 is an explanatory view showing the structure of a visual inspection data group;

FIG. 3 is an explanatory view showing a structure of an analysis data group;

FIG. 4 is an explanatory diagram showing the structure of a process apparatus corresponding data group;

FIG. 5 is an explanatory diagram showing the structure of a device maintenance data group;

FIG. 6 is an explanatory diagram showing a transition diagram of appearance defects.

FIG. 7 is a flowchart showing details of a process 100;

FIG. 8 is a flowchart showing details of a process 101;

FIG. 9 is an explanatory diagram showing a display example of an analysis spectrum;

FIG. 10 is a flowchart showing details of a process 102;

FIG. 11 is an explanatory view showing a display example of an appearance defective device maintenance correspondence graph.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Appearance inspection data group, 2 ... Analysis data group, 3 ... Process apparatus corresponding data group, 4 ... Equipment maintenance data group, 100 ... Process of searching and analyzing appearance inspection data, 101 ... Process of searching and analyzing analysis data , 102: A process of searching for a device name from a process name, and further searching and analyzing device maintenance data.

Claims (11)

(57) [Claims] In the process of processing a work, the work
What is claimed is: 1. A data analysis method for analyzing a cause of a defect that has occurred, comprising: a test result obtained by inspecting a work; information on a time when the work was inspected; and information on a time when a processing device that processed the work was maintained. Associated with
Data analysis method and outputting. 2. An inspection result obtained by inspecting the work , and
Information on the time when the work was inspected and processing of the work
Information about when the maintenance
2. The data analysis method according to claim 1 , wherein is displayed on the same screen . 3. A process for treating a work,
What is claimed is: 1. A data analysis method for analyzing a cause of an occurrence of a defect , wherein the inspection result obtained by inspecting the work is obtained by comparing information on a time at which the work was inspected and a defect generated in the work.
A data analysis method characterized by outputting in association with information of an analysis result . 4. A process for treating a work,
A data analysis method that analyzes the cause of the failure that has occurred.
Inspection results obtained by inspecting the work
Information about the inspected time and the processing equipment that processed the workpiece.
Information about when the equipment was maintained, and
A data analysis method characterized by outputting a generated defect in association with information of a result of analysis. 5. A test system comprising a predetermined processing step inspecting a workpiece passing through the analyzing data of the inspection apparatus and a result of the inspection by the inspection device <br/> analyzing device, the analyzing device , inspection results obtained by inspecting the work in the test devices
And associated with the information relating to the inspection device time maintains a processing device for processing the workpiece the workpiece with information relating to time examined at the predetermined treatment process with
Processing means for performing the following processing, and the processing means
The inspection result and the time when the work was inspected.
Information and information about when the processor was maintained
A display means for displaying the information on the same screen . 6. A work which has passed through a predetermined processing step is inspected and inspected.
Inspection device for detecting goodness and data of the result of inspection by the inspection device
An analysis device for analyzing the data, wherein the analysis device inspects a workpiece with the inspection device.
The obtained inspection result, information on the time when the work was inspected by the inspection device and the work detected by the inspection device
Performs a process of associating the defect with the information of the analysis result
Processing means and the inspections processed and associated by the processing means
Information on the results and the time when the work was inspected and the work
Displays the information of the result of analyzing the faults on the same screen on the same screen
An inspection system, comprising: a display unit for performing inspection. 7. Inspection of a workpiece which has passed a predetermined processing step
Inspection device for detecting goodness and data of the result of inspection by the inspection device
An inspection system equipped with an analyzer for analyzing data
The analysis device inspects the workpiece with the inspection device.
The obtained inspection results were used to inspect the work with the inspection device.
Information on the time and the work in the predetermined process
Information on when the processing equipment to be processed was maintained
And the defect of the work detected by the inspection device was analyzed.
Processing means for performing processing associated with the result information;
The inspection result and the work,
And information on the time when the
Information on the timing of the
Display means for displaying information on the same result on the same screen.
An inspection system characterized by: 8. The inspection device according to claim 1, wherein the workpiece is a semiconductor wafer,
Circuit pattern formed on semiconductor wafer by inspection device
Defect inspection device for inspecting defects
The inspection system according to claim 5 . 9. The inspection result is the number of appearance defects of the work.
The method according to any one of claims 5 to 7, wherein
Inspection system. 10. A workpiece is sequentially processed in a plurality of processing steps.
A method of manufacturing a thin film product for manufacturing a thin film product, the method comprising:
Inspection of the appearance of the workpieces sequentially processed in the processing steps of
The result of the inspection was compared with the information on the inspection time and the previous
A processing device that sequentially processes workpieces in a predetermined processing process is provided.
Remembers information related to the time of maintenance
And inspecting the appearance stored in association with the inspection result.
Information about the time spent and when the processor was maintained
Information on the screen and the predetermined processing step
When a defect occurs in the appearance of the workpiece processed in
The predetermined processing step based on the information displayed on the screen
Of thin film characterized by analyzing the cause of failure occurrence in
Product manufacturing method . 11. A method of processing a workpiece sequentially in a plurality of processing steps.
A method of manufacturing a thin film product for manufacturing a thin film product, the method comprising:
Inspection of the appearance of the workpiece processed in the
Was detected, and the result of inspecting the appearance was inspected as described above.
Analyzed information about time and defects that occurred on the work
And store the information in association with the result information.
Information on the results of inspection of the
On the screen with information on the results of analyzing defects
Display and display the appearance of the workpiece processed in the predetermined processing step.
When a defect occurs, the information displayed on the screen
Analyze the cause of failure occurrence in the predetermined process based on
A method of manufacturing a thin film product .