A method for distinguishing the source of faults and defects in a chemical mechanical polishing layer

Abstract

The present invention provides a method for distinguishing the source of faults and defects in a chemical mechanical polishing layer, including obtaining the design information of a product chip and obtaining a first feature map according to the design information; obtaining the pattern of a wafer, and assigning the pattern of the wafer with the same coordinate origin as the first feature map, so as to obtain a second feature map; importing the second feature map into the database of an ADC server in the prior art to obtain defect coordinates, wherein the database classifies defect codes according to the types of defects generated by different process machines; when the ADC server classifies, it uses the defect coordinates to match the coordinates of the first feature map, so as to compare and obtain the position of the defect coordinates on the first feature map; determining the process machine corresponding to the defect coordinates, which can be accurately classified automatically in a cycle, excluding human intervention, helping engineers quickly and accurately lock the processing machine, preventing more affected batches from being generated, and improving the product yield.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and particularly to a method for distinguishing the fault defect sources of a chemical mechanical polishing layer. Background Art

[0002] As the manufacturing process becomes smaller and smaller, the loss rate of the product yield caused by the faults and defects of the chemical mechanical polishing (CMP) layer in the back-end process is close to 100%. Moreover, since there is more than one machine tool that can cause such defects, for example, the particles of the metal hard mask layer etching machine tool and the integrated etching machine tool can both cause pattern damage to the chemical mechanical polishing layer. In the actual operation process, engineers may notify the wrong machine tool or notify multiple machine tools to shut down, resulting in a loss of product yield or a loss of machine tool production capacity.

[0003] The existing detection and classification method is based on the automatic classification of an ADC (Auto Defect Classfication) server. The traditional ADC classification mainly establishes a method in the early stage, and then uses a large number of defect data as a database. In the later stage, when the SEM machine tool automatically identifies, based on the different gray scales of the defect and the background, the ADC server first defines the contour information of the defect, and then matches and classifies it with the information in the database; this existing technology cannot help engineers lock the defect source in the metal hard mask layer etching (Metal Hard Mask ETCH) or the integrated etching (AIO Etch). In the actual operation process, engineers may notify the wrong machine tool or notify multiple machine tools to shut down, resulting in a loss of product yield of some products or a loss of machine tool production capacity.

[0004] Disadvantages of the current technology:

[0005] (1) Due to the ever-changing morphology of the particles, the contours of the faults and defects of the chemical mechanical polishing layer are also various; the accuracy (purity) of the ADC classification can only reach about 90%.

[0006] (2) The current ADC technology can only simply classify the defects, and cannot classify the sources of the defects. Moreover, due to the chemical mechanical polishing (CMP) grinding process, part of the previous layer information will be lost in the pattern, which often brings trouble to engineers in making judgments. Summary of the Invention

[0007] In view of the above-mentioned disadvantages of the prior art, the purpose of the present invention is to provide a method for distinguishing the fault defect sources of a chemical mechanical polishing layer, which is used to solve the problems in the prior art that the accuracy of the ADC classification is low, and it can only simply classify the defects and cannot classify the sources of the defects.

[0008] To achieve the above and other related objectives, the present invention provides a method for distinguishing the source of faults and defects in a chemical mechanical polishing layer pattern, which at least includes:

[0009] Step 1: Obtain the design information of the product chip, and obtain a first feature map according to the design information, where different positions of the first feature map correspond to different process machines;

[0010] Step 2: Obtain the pattern of the wafer, where the pattern of the wafer has the same size ratio as the first feature map, and assign the pattern of the wafer with the same coordinate origin as the first feature map, so as to obtain a second feature map;

[0011] Step 3: Import the second feature map into the database of the ADC server, so that the ADC server obtains the defect coordinates in the second feature map, where the database classifies the defect codes according to the types of defects generated by different process machines;

[0012] Step 4: The ADC server classifies the defect coordinates, and uses the defect coordinates to match the coordinates of the first feature map, so as to compare and obtain the position of the defect coordinates on the first feature map;

[0013] Step 5: Determine the process machine corresponding to the defect coordinates according to the defect code in Step 3.

[0014] Optionally, the process machines in Step 1 include a metal hard mask layer etching machine and an integrated etching machine.

[0015] Optionally, the pattern of the wafer in Step 2 is obtained by a scanning machine or imported from external information.

[0016] Optionally, the defect coordinates in Step 3 are automatically recognized by an SEM machine.

[0017] Optionally, when the ADC in Step 3 does not detect a defect on the wafer, the process machine continues to work.

[0018] Optionally, the design information in Step 1 is the layout of the product chip, and the first feature map is obtained by CAD drawing according to the layout.

[0019] Optionally, the second feature map in Step 3 is input into the ADC server in text form.

[0020] As described above, the method for distinguishing the source of faults and defects in the chemical mechanical polishing layer of the present invention has the following beneficial effects:

[0021] It can be automatically and accurately classified by a loop, excluding human intervention, and making up for the drawback that the previous ADC classification accuracy could not reach 100%; it can help engineers quickly and accurately lock the processing machine, prevent more affected batches from occurring, and improve the product yield; it can avoid the downtime inspection of multiple processing machines together, resulting in the loss of the production capacity of the semiconductor factory; it can more accurately monitor the working state of the processing machine, which is conducive to improving the process stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It shows a schematic flow chart of the method of the present invention;

[0023] Figure 2 It shows a schematic diagram of the pattern developed by lithography of the hard metal mask layer in an embodiment of the present invention;

[0024] Figure 3 It shows a schematic diagram of the pattern developed by lithography of the metal through hole in an embodiment of the present invention;

[0025] Figure 4 It shows a schematic diagram of assigning corresponding coordinates to the pattern in an embodiment of the present invention;

[0026] Figure 5 It shows a schematic diagram of matching the coordinates of the first feature map in an embodiment of the present invention;

[0027] Figure 6 It shows a schematic diagram of confirming the defect position in an embodiment of the present invention;

[0028] Figure 7 It shows a schematic diagram of confirming the position of the faulty machine causing the defect in an embodiment of the present invention.

[0029] Among them, 101 - the first feature map, 102 - the second feature map, 103 - the defect coordinates. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] The following specific examples illustrate the embodiments of the present invention. Those skilled in the art can easily understand the other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0031] Please refer to Figure 1 , the present invention provides a method for distinguishing the source of fault defects in the chemical mechanical polishing layer pattern, which at least includes:

[0032] Step 1: Obtain the design information of the product chip, and obtain the first feature map 101 according to the design information, where different positions of the first feature map 101 correspond to different process machines, and different process machines generate different types of defects;

[0033] Specifically, the design information of the chip includes the design layout of the chip. According to each part in the design layout, the first feature map 101 is drawn by CAD. Each part of the device corresponds to a different processing machine tool. During processing, particles will cause damage to this part during chemical mechanical polishing. A large number of defect data caused by different processing machine tools are used as a database.

[0034] Please refer to Figure 2 and Figure 3 , in a possible embodiment, the information related to the product chip is obtained and the feature map is drawn by CAD, where Figure 2 is the pattern developed by metal hard mask layer lithography (M*_PH), Figure 3 is the pattern developed by metal via lithography (V*_PH). It should be understood that more types of processing machine tools can also be used here.

[0035] Step 2: Obtain the pattern of the wafer. After the pattern of the wafer is scaled and processed, it has the same size ratio as the first feature map 101. The pattern of the wafer here can be obtained by a scanning machine tool or through external data input. The pattern of the wafer is given the same coordinate origin as the first feature map 101, so as to obtain the second feature map 102. The coordinate axes of the second feature map 102 are the same as those of the first feature map 101. The first feature map 101 can be used as a reference system, so as to obtain the position corresponding to the design layout of the product chip on the second feature map 102;

[0036] Please refer to Figure 4 , in a possible example, select the same coordinate origin as the scanning machine tool and assign corresponding coordinates to the pattern.

[0037] Step 3: Import the second feature map 102 into the database of the ADC server in the prior art in text form, so that the SEM machine tool in the ADC server can automatically identify it. The database classifies defect codes according to different defect types generated by different processing machine tools;

[0038] Specifically, when the SEM machine tool automatically identifies, based on the different gray levels of the defect and the background, the ADC server first defines the contour information of the defect, and then matches it with the information in the database for classification, so as to obtain the defect coordinates 103 at the defective part in the second feature map 102. When the ADC does not detect a defect on the wafer, the processing machine tool continues to work.

[0039] It should be understood that the ADC server here can also be replaced by other detection and classification devices.

[0040] Step 4: When classifying, the ADC server uses the defect coordinates 103 at the defective locations in the second feature map 102 to match the coordinates of the first feature map 101, thereby obtaining the position of the defect coordinates 103 on the first feature map 101 through comparison.

[0041] Please refer to Figure 5 and Figure 6 , in a possible embodiment, when the ADC server identifies, it matches the defect coordinate 103 information with the coordinates of the CAD first feature map 101 to confirm that the defect is in the pattern developed by the lithography of the metal hard mask layer (M*_PH).

[0042] Step 5: According to the defect code in Step 1, determine the process machine tool corresponding to the defect coordinates 103, which can accurately determine the process machine tool where the failure occurs and can accurately lock the faulty machine tool.

[0043] Please refer to Figure 7 , in a possible embodiment, the position of the faulty machine tool causing the defect is confirmed, and it is in the metal hard mask layer etching machine tool.

[0044] It should be noted that the diagrams provided in this embodiment only illustrate the basic concept of the present invention in a schematic manner. Therefore, only the components related to the present invention are shown in the diagrams, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0045] In summary, the present invention can perform cyclic automatic and precise classification, eliminate human intervention, and make up for the disadvantages of low ADC classification accuracy in the past; it can help engineers quickly and accurately lock the processing machine tool, prevent more impacts on batches from occurring, and improve the product yield rate; it can avoid multiple processing machine tools from crashing and being inspected together, resulting in losses in the production capacity of the semiconductor factory;

[0046] It can monitor the working status of the processing machine tool more precisely, which is beneficial to improving the stability of the process. Therefore, the present invention effectively overcomes various disadvantages in the prior art and has high industrial utilization value.

[0047] The above embodiments only illustrate the principle and efficacy of the present invention by way of example, rather than limiting the present invention. Any person familiar with this technology can modify or change the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or changes completed by those with ordinary knowledge in the technical field without departing from the spirit and technical idea disclosed by the present invention should still be covered by the claims of the present invention.

Claims

1. A method for distinguishing the source of faults and defects in a chemical mechanical polishing layer, characterized in that, it at least includes: Step 1: Obtain the design information of the product chip. The design information of the chip includes the design layout of the chip. According to each part in the design layout, a first feature map is drawn through CAD. Different positions of the first feature map respectively correspond to different process machines; Step 2: Obtain the pattern of the wafer. The pattern of the wafer has the same size ratio as the first feature map. The pattern of the wafer is given the same coordinate origin as the first feature map, so as to obtain a second feature map; Step 3: Import the second feature map into the database of the ADC server, so that the ADC server obtains the defect coordinates in the second feature map. The database classifies the defect codes according to the defect types generated by different process machines; Step 4: The ADC server classifies the defect coordinates, and uses the defect coordinates to match the coordinates of the first feature map, so as to compare and obtain the position of the defect coordinates on the first feature map; Step 5: Determine the process machine corresponding to the defect coordinates according to the defect code in Step 3.

2. The method for distinguishing the source of faults and defects in a chemical mechanical polishing layer according to claim 1, characterized in that: The process machines in Step 1 include a metal hard mask layer etching machine and an integrated etching machine.

3. The method for distinguishing the source of faults and defects in a chemical mechanical polishing layer according to claim 1, characterized in that: The pattern of the wafer in Step 2 is obtained through a scanning machine.

4. The method for distinguishing the source of faults and defects in a chemical mechanical polishing layer according to claim 1, characterized in that: The defect coordinates in Step 3 are automatically identified by an SEM machine.

5. The method for distinguishing the source of faults and defects in a chemical mechanical polishing layer according to claim 1, characterized in that: When no defects occur on the wafer are detected by the ADC in Step 3, the process machine continues to work.

6. The method for distinguishing the source of faults and defects in a chemical mechanical polishing layer according to claim 1, characterized in that: The second feature map in Step 3 is input into the ADC server in text form.

Images