Method for filtering SiGe control defect signal

Abstract

The invention discloses a method for filtering SiGe control signals, which comprises the following steps of: firstly, summarizing a large number of feature correction values of control defects into a database, and carrying out visual matching on defect scanning patches and database patches; linkage is carried out on the defect coordinates and the GDS layout, secondary screening is carried out on the coordinates, and if the coordinates meet the standard, the signals are determined as SiGe control signals, and the SiGe control signals are filtered. According to the defect position coordinate and the patch image, the SiGe control signal can be accurately identified, the maintenance burden of the process parameters is reduced, the interference of the front layer is eliminated, the maintenance of the process trend of the current station is facilitated, the working state of the process machine table of the current station can be more accurately monitored, and the stability of the process is facilitated.

Description

Technical Field

[0001] This invention relates to the field of semiconductor device design and manufacturing, and in particular to a method for accurately filtering out SiGe concave defect signals. Background Technology

[0002] Due to the characteristics of SiGe epitaxial processes and the existence of equipment condition variations, defects in bulk SiGe growth are common and highly likely to occur in the long channel region defined by DPL (carrier diffusion distance). Although these defects have a minor impact on yield, the defect scanning signal (referring to the SiGe concave signal generated by the SiGe growth defect) will be exceptionally strong during the inspection process, and can be captured at multiple subsequent scanning inspection stations.

[0003] The main methods currently used in the industry for controlling machine defect scanning signals are: By analyzing historical data or setting a signal control threshold based on experience, defects below the signal control threshold are identified as abnormal defects.

[0004] After the signal control threshold is controlled, the defect signal is divided into different bins according to the scanning channel. In each bin, ST-NEF is used to select two parameters, such as the total energy and intensity of the defect signal. Defects that are lower than both control thresholds at the same time are identified as abnormal defects.

[0005] The commonly used methods in the industry for controlling machine defect scanning signals cannot completely filter out the scanned models. This poses a significant challenge to subsequent site monitoring for capturing actual defects and maintaining trend charts because the aforementioned control methods have the following drawbacks: (1) For SiGe concave signals, the signal is too strong. If you want to filter it, for most control parameters, the control value is far greater than the BKM value, which will cause false damage to a large number of real defects. (2) Pure parameter values ​​cannot fully summarize the characteristics of defects, and different defects may have similar numerical performance. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method for accurately filtering out SiGe concave signals.

[0007] To address the above problems, this invention provides a method for filtering out SiGe concave defect signals, comprising: First, the numerous feature correction values ​​collected for concave defects are summarized into a defect patch database; When the machine scans the wafer for defects, if a SiGe concave defect is detected and the defect signal formed by the defect has reached the preset defect signal trigger threshold, it can be identified as a pending SiGe concave defect signal. This pending SiGe concave defect signal is then visually searched and matched with the defect patch database to find the corresponding database patch. Next, the SiGe concave defects are located on the wafer to form defect coordinate data on the wafer. This defect coordinate data is then linked with the GDS layout data and compared and searched in the GDS layout to find the corresponding position. Through secondary screening using coordinates, those that meet the criteria are identified as SiGe concave defect signals and filtered out.

[0008] The concave defect mentioned above is a defect that occurs during SiGe growth in the SiGe epitaxial process, and is a SiGe concave signal generated by a defect scanning machine.

[0009] Furthermore, the numerous feature correction values ​​of the concave defects are concave defect correction data collected and organized during historical processes. These data, after being previously determined to be genuine SiGe concave defect signals, are then organized to form a defect patch database.

[0010] Furthermore, the GDS layout data is the processing data that the IC design end transfers to the IC manufacturing plant after the layout design verification is completed. It includes graphic information of each layer of the IC layout, contact holes, each injection area or structure, and interconnects.

[0011] Furthermore, the aforementioned coordinate secondary filtering involves comparing the database patch found in the defect patch database with the corresponding graphic obtained by searching the GDS layout based on the defect coordinate data. If the comparison results are consistent, the pending SiGe concave defect signal can be determined to be a genuine SiGe concave defect signal and filtered out. If the comparison results are inconsistent, the pending SiGe concave defect signal can be determined to be a false SiGe concave defect signal and should be retained.

[0012] Furthermore, the aforementioned defect signal trigger threshold is a criterion used by the machine to determine whether a signal is a defect among the many signals scanned. Defect signals exceeding the criterion will be marked as defect signals, while signals below the criterion will be discarded and not identified as defect signals.

[0013] This invention organizes and summarizes a large number of feature patches of concave defects into a defect patch database, and then visually matches the defect scan patch signals with the data in the defect patch database; next, it links the coordinates of the defects with the GDS layout; through secondary coordinate filtering, those that meet the criteria are identified as genuine SiGe concave defect signals and are filtered out, while others are retained. Accurate classification of scanned and identified SiGe concave defect signals enables precise determination of the product's condition. Attached Figure Description

[0014] Figure 1 This invention relates to a method for filtering out SiGe concave signals. Detailed Implementation

[0015] The following detailed description, in conjunction with the accompanying drawings, provides specific embodiments of the present invention and clearly and completely describes the technical solutions of the present invention. However, the present invention is not limited to the following embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The advantages and features of the present invention will become clearer from the following description and claims. It should be noted that the accompanying drawings are all in a very simplified form and use non-precise ratios, and are only used for the purpose of conveniently and clearly illustrating the embodiments of the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] It should be understood that the present invention can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein. Rather, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In the drawings, for clarity, the dimensions and relative dimensions of layers and regions may be exaggerated, and the same reference numerals denote the same elements throughout. It should be understood that when an element or layer is referred to as “on,” “adjacent to,” “connected to,” or “coupled to” other elements or layers, it may be directly on, adjacent to, connected to, or coupled to other elements or layers, or there may be intervening elements or layers. Conversely, when an element is referred to as “directly on,” “directly adjacent to,” “directly connected to,” or “directly coupled to” other elements or layers, there are no intervening elements or layers. It should be understood that although the terms first, second, third, etc., may be used to describe various elements, components, regions, layers, and / or portions, these elements, components, regions, layers, and / or portions should not be limited by these terms. These terms are used only to distinguish one element, component, region, layer, or part from another element, component, region, layer, or part. Therefore, without departing from the teachings of this invention, the first element, component, region, layer, or part discussed below may be referred to as the second element, component, region, layer, or part.

[0017] This invention provides a method for accurately filtering out SiGe concave signals. Based on the defect location coordinates and patch images, the SiGe concave signals are accurately identified and automatically filtered to reduce the impact on the capture of actual defects at the site.

[0018] The technical method involves collecting historical data, summarizing and organizing a large number of feature patches with concave defects into a database, and then visually matching the defect scanning patches with the database patches. Next, the defect coordinates are compared with the GDS layout, and through secondary coordinate filtering, those conforming to the standard are identified as SiGe concave signals and filtered out. GDS layout data is the processing data transferred from the IC design end after layout design verification and handed over to the IC manufacturing plant. It contains various structural features of the layout, including processing information such as layout layers, interconnections, contact holes, and injection areas, used for chip manufacturing.

[0019] For example, such as Figure 1 As shown, a patch database is formed based on historical scan data. This patch database contains a large number of SiGe concave signals from historical scans, which are classified and organized according to threshold control to form the patch database.

[0020] During a wafer defect scan, the equipment detects a defect whose signal has reached a pre-set defect signal trigger threshold. This threshold can be preliminarily identified as a SiGe concave signal, forming defect patch data. The defect signal trigger threshold is a criterion used by the equipment to determine whether a signal is a defect among the many signals scanned. Signals exceeding this threshold are identified as defects, while signals below it are discarded. This criterion is established based on historical data analysis and verification.

[0021] Then, this defect patch data is searched and compared in the formed defect patch database to find the database patch data that is closest to it, and it is recorded as database patch data A.

[0022] Simultaneously, the defect is located on the wafer, its coordinates on the wafer or layout are determined, for example, coordinates X=123, Y=456. Then, the layout shape or structural information at that coordinate location is searched in the GDS data corresponding to that wafer layout. By searching in the GDS layout data, it is found that the coordinates are located in the overlapping area of ​​the DPL mask and the SiGe mask, such as... Figure 1 The GDS map is shown on the right side of the middle section.

[0023] The next step involves information comparison. The previously found database patch data A is compared with the graphic structure of the GDS layout coordinates. By comparing the defect patch database with the GDS layout information of the defect location, if the two match, the defect is determined to be a genuine defect, a real SiGe concave signal, and is filtered out. If the two do not match, the defect is determined to be a false defect, a false SiGe concave signal, and is retained in the signal set scanned by the machine.

[0024] Therefore, the above method can accurately determine whether a SiGe defect is a concave and effectively identify the signal, so as to facilitate the capture of true defects and the maintenance of process trends at subsequent stations.

[0025] The advantages of the method of the present invention are as follows: 1. Effectiveness: Based on the defect location coordinates and patch image, the SiGe concave signal is accurately identified and automatically filtered to reduce the impact on the capture of real defects at the current station.

[0026] 2. It can make the control conditions of the process parameters at the station return to the BKM value, reducing the maintenance burden of process parameters.

[0027] 3. Economic efficiency: Eliminating interference from the previous layer is beneficial for maintaining the process trend of the station, and allows for more accurate monitoring of the working status of the process equipment at the station, which is conducive to improving the stability of the process.

[0028] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for filtering SiGe concave defect signals, characterized in that: firstly, a large number of characteristic correction values of collected concave defects are summarized into a defect patch database; when a wafer is scanned by a machine, a SiGe concave defect is scanned, and a defect signal formed by the defect has reached a pre-set defect signal trigger threshold, and can be identified as a pending SiGe concave defect signal; the pending SiGe concave defect signal is visually searched and matched with the defect patch database to find a corresponding database patch; the SiGe concave defect is positioned on the wafer to form defect coordinate data on the wafer, and the defect coordinate data is linked with GDS layout data to find a corresponding position in the GDS layout; and through secondary screening of the coordinates, if the standard is met, the SiGe concave defect signal is identified, and it is filtered. The concave defect is a defect occurring in SiGe growth in SiGe epitaxy process, and a SiGe concave signal is generated by a defect scanning machine. The large number of characteristic correction values of the concave defect are concave defect correction data collected and sorted in historical process, and are determined as real SiGe concave defect signals through previous determination, and are sorted to form a defect patch database. The GDS layout data is processing data of an IC design end after layout design verification is completed and is given to an IC manufacturing factory, and contains graphic information of IC layout layers, contact holes, injection zones or structures, and interconnection lines. The secondary screening of the coordinates is to compare the database patch found in the defect patch database with corresponding graphics obtained by searching in the GDS layout according to the defect coordinate data, if the comparison result is consistent, the pending SiGe concave defect signal is determined as a real SiGe concave defect signal, and is filtered; and if the comparison result is inconsistent, the pending SiGe concave defect signal is determined as a false SiGe concave defect signal, and should be retained.

2. The method of filtering SiGe concave defect signals of claim 1, wherein: The defect signal trigger threshold is a determination criterion for determining whether a signal scanned by a machine is a defect, and a defect signal exceeding the determination criterion is marked and identified as a defect signal, and a signal lower than the determination criterion is discarded and not identified as a defect signal.

3. The method of filtering SiGe concave defect signals of claim 1, wherein: Through accurate classification of scanned and determined SiGe concave defect signals, the state of a product can be accurately determined and classified.

4. The method of claim 1, wherein the SiGe concave defect signal is filtered out by: ​ ​ 5. The method of filtering SiGe concave defect signals of claim 1, wherein: ​ 6. The method of filtering SiGe concave defect signals of claim 1, wherein: ​ 7. The method of filtering SiGe concave defect signals of claim 1, wherein: ​

Images