WO2018010272A1 - 在自动缺陷分类流程中管理缺陷的方法及系统 - Google Patents

在自动缺陷分类流程中管理缺陷的方法及系统 Download PDF

Info

Publication number
WO2018010272A1
WO2018010272A1 PCT/CN2016/097168 CN2016097168W WO2018010272A1 WO 2018010272 A1 WO2018010272 A1 WO 2018010272A1 CN 2016097168 W CN2016097168 W CN 2016097168W WO 2018010272 A1 WO2018010272 A1 WO 2018010272A1
Authority
WO
WIPO (PCT)
Prior art keywords
defect
image
inspection
background region
processor
Prior art date
Application number
PCT/CN2016/097168
Other languages
English (en)
French (fr)
Inventor
马卫民
吴筱美
张兆礼
Original Assignee
东方晶源微电子科技（北京）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 东方晶源微电子科技（北京）有限公司 filed Critical 东方晶源微电子科技（北京）有限公司
Priority to US15/285,955 priority Critical patent/US9928446B2/en
Publication of WO2018010272A1 publication Critical patent/WO2018010272A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G06T7/001—Industrial image inspection using an image reference approach
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/903—Querying
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30148—Semiconductor; IC; Wafer

Definitions

  • the present invention relates to the use of simulation techniques, such as lithographic simulation, to enhance the function of automatic defect classification in an automatic defect classification (ADC) process.
  • ADC automatic defect classification
  • ADC automatic defect classification
  • the method includes:
  • the defect is defined as a system defect based on a determination that the background region matches the defect image.
  • Also disclosed herein is a system for managing defects in an automated defect classification process, the system comprising:
  • a defect inspection module for receiving a defect record based on the inspection of the target sample
  • Design a database for storing wafer or mask design data
  • a lithography analysis module configured to extract, from a design database, an area around the location of a defect in the defect record, and associated design data associated therewith, and, by the processor, based on the association associated with the area Performing lithography simulation on relevant design data to determine a simulated image of a background region;
  • An automatic defect classification module configured to compare, by the processor, a simulated image of the background region with a defect image of the defect in the defect record to determine whether the background region matches the defect image, and based on The determination that the background area matches the defect image defines the defect as a system defect.
  • non-volatile computer readable medium storing a set of instructions that, when executed by a processor of a computer system, are operable by the processor to manage in an automated defect classification process Defect, the non-transitory computer readable medium includes instructions to:
  • the defect is defined as a system defect based on a determination that the background region matches the defect image.
  • Also disclosed herein is a system for managing defects in an automated defect classification process, the system comprising:
  • a memory coupled to the processor, the memory configured to store a set of instructions that, when executed by the processor, are operable by the processor to:
  • the defect is defined as a system defect based on a determination that the background region matches the defect image.
  • FIG. 1 is a schematic diagram of an example flow for enhancing automatic defect classification by lithography simulation
  • FIG. 2 is a schematic diagram of an example flow for enhancing automatic defect classification by lithography simulation
  • FIG. 3 is a schematic diagram of another example flow for enhancing automatic defect classification by lithography simulation
  • FIG. 4 is a schematic diagram of an exemplary enhanced automatic defect classification system
  • FIG. 5 is a schematic diagram of an exemplary enhanced automatic defect classification system that can implement aspects of the present invention.
  • ADC automatic defect classification
  • defect classification after analyzing the characteristics of a defect-related defect image, a label can be assigned to the defect. By assigning the correct label to the inspected defect, predetermined measures can be quickly implemented to process defects, wafers or reticle on the production line.
  • classification-based defect analysis can be used to guide remediation measures to improve process and yield development.
  • defects identified by the automatic defect classification process can be classified into different types, such as system defects and random defects.
  • System defects, such as distortion are often caused by resolution enhancement techniques, which can be harmless and have no lethal effects on the manufacturing process.
  • system defects can also be fatal, for example, if they are caused by geometric design and the manufacturing process cannot be resolved.
  • Random defects on the other hand, are usually caused by particles (such as dust on the wafer), inhomogeneities or irregularities. Random defects may be harmless (for example, chips that are affected by random defects are not of any electronic importance) or they may be fatal (for example, random defects cause open or short circuits).
  • other types of defects can also be classified.
  • an automated defect classification process enhanced with lithographic simulation is used to supplement the automatic defect classification process with information on system defects and to direct the automatic defect classification process to assign the correct classification to defects.
  • inspection of the wafer or reticle can generate a defect record.
  • the location of the defect can be extracted from the defect record, and a database clip including the corresponding relevant design data can be extracted from the design database for the region around the defect.
  • a lithography simulation can be performed based on the database segment to generate a background region that will be used to compare with the defect image to identify system defects. In addition to system defects, other types of defects can be identified accordingly.
  • FIG. 1 is a schematic diagram of an example process 100 for enhancing automatic defect classification by lithography simulation.
  • the enhanced automatic defect classification process 100 illustrated in FIG. 1 may be performed by a software module (such as instructions or code) running on a processor of a computer system, a hardware module of a computer system, or a combination of both.
  • a software module such as instructions or code
  • One or more of the steps described herein can be integrated into, for example, a product that performs inspections on a wafer or reticle and used by semiconductor manufacturers.
  • a defect record is received based on the inspection of the target sample.
  • the target sample may include, for example, a wafer or a reticle.
  • Inspection may include, for example, optical or electron beam inspection.
  • the defect record can be part of the inspection report and can include one or more defects.
  • Each defect can include information such as location and image.
  • the location can include data about the location of the defect, such as layer information and coordinates. Defective images and/or locations can be extracted from the defect records.
  • an area around the location of a defect in the defect record, associated with the associated design data is extracted from the design database.
  • the design database can include at least one Design database files corresponding to target samples, such as wafers or reticle, may also include, for example, chip blueprints and polygon descriptions.
  • the relevant design data can be extracted in the form of a database segment that contains relevant data for the region surrounding the location of the defect. Relevant design data may include any data related to the design of the wafer or reticle, such as physical data (eg, size), geometric data (eg, shape or layout), technical data, layer data, or any combination of the above.
  • a lithography simulation is performed based on the associated design data to determine a background region around the location of the defect. For example, a lithography simulation can be run based on the database segment to generate a background region around the defect location.
  • the background area may include a simulated image that is substantially a database image that is simulated based on a calibration model of the lithography process.
  • the background area may also include data such as circuits, traces, and other data generated or rendered using a lithography process.
  • the background area is compared with a defect image of the defect in the defect record to determine whether the background area matches the defect image.
  • the comparison between the background area and the defect image can be done by, for example, aligning the two images and subtracting one from the other. This comparison can also utilize a defect inspection algorithm, which can be determined based on the specific requirements of the application.
  • the inspection can be an optical inspection or an electron beam inspection.
  • the background area can be compared to an optical image of the defect (from an optical inspection), which is similar to a die-to-database inspection.
  • the contour of the background area can be compared to the contour of the defective Scanning Electron Microscope (SEM) image (from an electron beam inspection). In some embodiments, the contours may be extracted prior to comparison.
  • SEM Scanning Electron Microscope
  • the simulated image (of the background area) matches the defective image, it indicates that the defective image can be recreated from the design database through the simulation process. This means that the defect is more likely to be caused by system factors (such as OPC or process changes, etc.) rather than random factors (such as dust).
  • the defect is defined as a system defect based on a determination that the background region matches the defect image.
  • the defect can be classified as a system defect or a non-system defect. If the background area and the defect image are judged to be mismatched, the defect is not classified as a system defect.
  • the defect may be classified as a non-system defect (such as a random defect or another type of defect), or may be assigned based on the background and defect information based on other aspects of the automatic defect classification decision Specific label.
  • system defect information (or non-system defect information), along with other related information (such as background information), can be used to enhance its flow in the automatic defect classification decision process.
  • the information contained in the background area can also be used in other ways by the enhanced automatic defect classification process.
  • certain defects may be pre-categorized or eliminated using predetermined rules.
  • system defects can be automatically categorized using predetermined rules in an enhanced automatic defect classification process.
  • the classifier in the automatic defect classification process will know the location of the defect and the type of the circuit in the circuit, so as to intelligently judge the influence of the defect on the circuit, and then accurately return Class defect. Furthermore, the enhanced automatic defect classification process can be applied to situations other than the above-described embodiments, and can be used to classify all kinds of defects (such as shape and hue) based on information such as the defect surrounding environment provided by lithography simulation.
  • FIG. 2 is a schematic diagram of an example process 200 for enhancing automatic defect classification by lithography simulation.
  • the process 200 illustrated in Figure 2 may be performed by a software module (such as instructions or code) running on a processor of a computer system, a hardware module of a computer system, or a combination of both.
  • the enhanced automatic defect classification process 200 includes extracting relevant design data (database segment 206) associated with the region around the location of the defect 202 from the design database 204.
  • the defect 202 can be obtained from a defect record as an inspection result (for example, an optical or electron beam inspection) on a target sample such as a wafer or a reticle.
  • the lithography simulation can be performed based on the associated design data (database segment 206) associated with the region to determine the background region 208.
  • the enhanced automatic defect classification 210 can then compare the image of the defect 202 ("defect image") to the background region 208 to determine if the background region and the defect image match, and/or classify the defect as a system defect or Non-system defects, and other matters. In this way, the original automatic defect classification is enhanced, and information on system defects can be taken into account when processing defects.
  • a design database for a wafer or reticle may be very large, and may include hundreds of gigabytes of data, while a database segment containing associated design data associated with a region around the location of the defect 202 is relatively small, including, for example, One or several MB of data.
  • the size of the region (or database segment 206) may vary, such as based on the size of the shape or defect. In one example, the width and height of the area can be 32 x 32 or 64 x 64.
  • FIG. 3 is a schematic diagram of another example process 300 for enhancing automatic defect classification with lithography simulation.
  • the process 300 shown in FIG. 3 may be performed by a software module (such as instructions or code) running on a processor of a computer system, a hardware module of a computer system, or a combination of both.
  • a software module such as instructions or code
  • defects can be grouped.
  • the enhanced automatic defect classification process 300 differs from the enhanced automatic defect classification process 200 in that defects from defect records can be classified into groups. For example, all defects in one defect group (such as defect group 302) may have the same or similar images.
  • Each defect group may have one representative defect 304; it may be a randomly selected defect or a defect specified based on one or more predetermined rules. For example, a new defect can be compared to an image of representative defect 304 of each defect group ("representing a defect image") to determine the defect group to which the new defect belongs.
  • a representative defect image for a group can be compared to a background region 208 corresponding to the representative set of defects (as determined at operation 106 of Figure 1) (as in Figure 1) Operation 108). Comparing the defective image with the corresponding background area in the comparison group and the corresponding background area can greatly reduce the calculation work compared to comparing each defective image with the background area. For example, the comparison result can be used for enhanced automatic defect classification 210, which will process the same information for each defect within the same group. For example, when a representative defect image of a defect group is determined to match the background region generated for the defect group, all defects within the group can be classified as system defects without repeating all operations.
  • the enhanced automatic defect classification process 300 includes extracting from the design database 204 relevant design data (database segment 206) associated with the region around the defect 304 location of the defect group 302. As described above, the lithography simulation can be performed based on the associated design data (database segment 206) associated with the region to determine the background region 208. Next, the enhanced automatic defect classification 210 can compare the representative defect image of the defect group 302 with the background region 208 to determine if the background region and the representative defect image match, and/or classify all defects of the defect group as a system. Defective or non-systematic defects, among other things.
  • the enhanced automatic defect classification system 400 can include a device, such as a computing device, that can be implemented as any configuration of one or more computers, such as a microcomputer, a mainframe computer, a supercomputer, a general purpose computer, a dedicated/special purpose meter A computer, integrated computer, database computer, remote server computer, personal computer, or computing service provided by a computing service provider, such as a network host or cloud service provider.
  • the computing device can be implemented in the form of multiple sets of computers in different geographic locations.
  • the enhanced automatic defect classification system 400 can include a defect inspection module 402, a design database 404, a lithography analysis module 406, and an automatic defect classification module 408.
  • the defect inspection module 402 can be configured to perform an inspection (eg, an optical or electron beam inspection) on a target sample (eg, a wafer or reticle) and generate an inspection report including the defect record.
  • the defect record can include a list of defects, and each defect can include image and location information.
  • defects can be classified into groups, and representative defects in the group can be used for comparison rather than comparing each individual defect, as discussed in FIG.
  • Design database 404 can be any database that holds wafer or reticle design data. As previously mentioned, relevant data for the area surrounding the defect location ("database fragment" herein) can be selected and retrieved from the design database 404.
  • the lithography analysis module 406 can perform lithographic simulation based on the database segments to generate a background region.
  • the background area can include various background information around the location of the defect, such as a simulated database image as a result of the lithography process.
  • the automatic defect classification module 408 can use the background area (generated by the lithography analysis module 406) and the defect information (generated by the defect inspection module 402) for its decision. For example, the automatic defect classification module 408 can compare the defect image with the corresponding background area: if matched, the defect is identified as a system defect. In some embodiments, a set of representative defect images can be compared to corresponding background regions; if matched, the group (and all defects within the group) are identified as system defects. Information included in the background area may also be used by the automatic defect classification module 408 in other ways, for example, certain defects may be pre-categorized or excluded using predetermined rules.
  • FIG. 5 is a schematic illustration of an exemplary enhanced automatic defect classification system 500 that can implement aspects of the present invention.
  • the enhanced automatic defect classification system 500 can include a device, such as a computing device.
  • the computing device can be implemented as any configuration of one or more computers, such as a microcomputer, mainframe computer, supercomputer, general purpose computer, special purpose/special purpose computer, integrated computer, database computer, remote server computer, personal A computer, or computing service provided by a computing service provider, such as a network host or cloud server.
  • the computing device may be implemented in a plurality of groups of computers in different geographical locations that do not communicate with each other, such as through a network. Some operations may be shared by multiple computers, but in some embodiments, different computers may be assigned different operations.
  • the computing device can have an intrinsic hardware configuration including a processor 502 and a memory 504.
  • Processor 502 can include at least one processing unit, such as a central processing unit (CPU) or any other type of device, or multiple devices that can manipulate or process information that is present or later developed. Although the embodiments herein can be implemented with a single processor as shown, speed and efficiency advantages can be achieved with multiple processors. For example, processor 502 can be distributed across multiple machines or devices (each machine or device having one or more processors) that can be directly connected or connected through a local area network or other network.
  • Memory 504 can be a random access memory device (RAM), a read only memory device (ROM), an optical disk, a magnetic disk, or any other suitable type of storage device.
  • memory 504 can be distributed across multiple machines or devices, such as network-based memory or memory that performs operations on multiple machines - for ease of illustration, it can be described herein as using a single computer or multiple Computing device execution.
  • memory 504 can store code and data that processor 502 can access with a bus.
  • memory 504 can include data 5042 that processor 502 can access via bus 512.
  • the memory 504 can also include an operating system 5046 and an installed application 5044.
  • Application 5044 includes programs to allow processor 502 to implement instructions to generate control signals, implement functions for the enhanced automatic defect classification system 500 described herein.
  • the instructions may also include processing defect information that is not part of the automated defect classification system for classification, such as generating context information that may be used by the automated defect classification system to generate system defect tags.
  • the enhanced automatic defect classification system 500 can also include auxiliary, additional or external memory 506, such as a memory card, flash drive, external hard drive, optical drive, or any other form of computer readable media.
  • the application 5044 can be stored, in whole or in part, in the memory 506 and loaded into the memory 504 as needed by the processing.
  • the enhanced automatic defect classification system 500 can include one or more output devices, such as an output 508.
  • Output 508 can be implemented in a variety of ways - for example, it can be a display connected to enhanced automatic defect classification system 500 configured to display rendering of video data.
  • Output 508 can be any device that transmits a visual, audible, or tactile signal to a user, such as a display, a touch sensitive device (such as a touch screen), a speaker, a headset, a light emitting diode (LED) indicator, or a vibration motor.
  • the output 508 is a display that can be a liquid crystal display (LCD), a cathode ray tube (CRT), or any other output device capable of providing a visible output to an individual.
  • the output device can also function as an input device - for example, a touch screen display configured to receive touch-based input.
  • Output 508 may alternatively or additionally constitute a communication device for transmitting signals and/or data.
  • output 508 can include wired means for transmitting signals or data from enhanced automatic defect classification system 500 to other devices.
  • output 508 can include a wireless transmitter that uses a wireless receiver compatible protocol to transmit signals from enhanced automatic defect classification system 500 to other devices.
  • the enhanced automatic defect classification system 500 can include one or more input devices, such as input 510.
  • Input 510 can be implemented in a variety of manners, such as a keyboard, a numeric keypad, a mouse, a microphone, a touch sensitive device (such as a touch screen), a sensor, or a gesture input device. Any other type of input device is possible, including input devices that do not require user intervention.
  • input 510 can be a communication device, such as a wireless receiver that receives signal operations in accordance with any wireless protocol.
  • Input 510 can output, as along bus 512, an input signal or data to the enhanced automatic defect classification system 500.
  • the enhanced automatic defect classification system 500 can communicate with other devices over a network (e.g., network 516) using a communication device, such as communication device 514.
  • Network 516 can be any combination of one or more communication networks of any suitable type including, but not limited to, using Bluetooth, infrared, near field connection (NFC), wireless networks, wired networks, local area networks (LANs), wide area networks (WANs), A network of virtual private networks (VPNs), cellular data networks, and Internet communications.
  • Communication device 514 can be implemented in various manners such as a transponder/transceiver, modem, router, gateway, circuit, chip, wired network adapter, wireless network adapter, Bluetooth adapter, infrared adapter, NFC adapter, cellular network chip, or any Any combination of suitable types of devices that utilize bus 512 to connect to enhanced automatic defect classification system 500 to provide communication via network 516.
  • the enhanced automatic defect classification system 500 can communicate with a wafer or reticle inspection device.
  • the enhanced automatic defect classification system 500 can be coupled to one or more wafer or reticle inspection devices configured to generate wafer or reticle inspection results, such as defect records or reports.
  • the enhanced automatic defect classification system 500 may be implemented in hardware, including, for example, an intellectual property (IP) core, an application specific integrated circuit (ASIC), Programming logic arrays, optical processors, programmable logic controllers, microcode, firmware, microcontrollers, servers, microprocessors, digital signal processors, or any other suitable circuitry.
  • IP intellectual property
  • ASIC application specific integrated circuit
  • Programming logic arrays optical processors
  • programmable logic controllers microcode, firmware, microcontrollers
  • servers microprocessors, digital signal processors, or any other suitable circuitry.
  • signal processors digital signal processors
  • the enhanced automatic defect classification system 500 can be implemented with a general purpose computer/processor with a computing program; when executed, implements any of the methods, algorithms, and/or instructions described herein.
  • a special purpose computer/processor may be additionally or alternatively used, which may comprise dedicated hardware for performing any of the methods, algorithms or instructions described herein.

Landscapes

  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Databases & Information Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Analytical Chemistry (AREA)
  • Chemical & Material Sciences (AREA)
  • Biochemistry (AREA)
  • Testing Or Measuring Of Semiconductors Or The Like (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • General Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Analysing Materials By The Use Of Radiation (AREA)

Abstract

本发明公开了一种用于在自动缺陷分类流程中管理缺陷的方法及系统。该方法包括：基于对目标试样的检查，接收一个缺陷记录；从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据；通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域的仿真图像；通过所述处理器比较所述背景区域的仿真图像与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。

Description

在自动缺陷分类流程中管理缺陷的方法及系统
本申请要求2016年7月13日递交的发明名称为“在自动缺陷分类流程中管理缺陷的方法及系统”的申请号201610551876.4的在先申请优先权，上述在先申请的内容以引入的方式并入本文中。
技术领域
本发明涉及在自动缺陷分类(automatic defect classification，ADC)流程中，利用仿真技术——如光刻仿真——来增强自动缺陷分类的功能。
背景技术
随着半导体制造技术发展推进到更细的分辨率——如小于20纳米，缺陷数目的增加可因各种系统条件所引发，例如工艺变化及光学临近修正(optical proximity correction,OPC)技术。不断增加的系统性缺陷可能会导致性能下降。
目前，自动缺陷分类(ADC)已被广泛用于半导体制造。
发明内容
本文公开了一种以光刻仿真增强自动缺陷分类流程的方法。该方法包括：
基于对目标试样的检查，接收一个缺陷记录；
从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域的仿真图像；
通过所述处理器比较所述背景区域的仿真图像与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；
基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
本文还公开了一种用于在自动缺陷分类流程中管理缺陷的系统，所述系统包括：
缺陷检查模块，用于基于对目标试样的检查，接收一个缺陷记录；
设计数据库，用于存储晶片或掩模版设计数据；
光刻分析模块，用于从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据，并且，通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域的仿真图像；
自动缺陷分类模块，用于通过所述处理器比较所述背景区域的仿真图像与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配，并且，基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
本文还公开了一种非易失性的存储有一套指令集的计算机可读介质，所述指令集被计算机系统的处理器执行时，可被所述处理器操作以在自动缺陷分类流程中管理缺陷，所述非易失性的计算机可读介质包括指令，以：
基于对目标试样的检查，接收一个缺陷记录；
从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域；
通过所述处理器比较所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；
基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
本文还公开了一种用于在自动缺陷分类流程中管理缺陷的系统，所述系统包括：
一个处理器；
一个连接到所述处理器的存储器，所述存储器经配置以存储一套指令集，所述指令集被所述处理器执行时，可被所述处理器操作，以：
基于对目标试样的检查，接收一个缺陷记录；
从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提 取与其关联的相关设计数据；
通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域；
通过所述处理器比较所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；
基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
上述实施方式的细节、上述实施方式的修改，以及更多的实施方式，将通过下文详细描述。
附图说明
通过以下的详细描述，并结合附图，可以更好地理解本发明。需要强调的是，根据通常的作法，附图的各种特征不一定成比例。相反，为解释清楚，所述各种特征的尺寸可能被任意地放大或缩小。本发明的公开内容提到所附的附图时，类似的附图标记在所有附图中表示类似的部分。在附图中：
图1是一个以光刻仿真增强自动缺陷分类的示例流程的示意图；
图2是一个以光刻仿真增强自动缺陷分类的示例流程的示意图；
图3是另一个以光刻仿真增强自动缺陷分类的示例流程的示意图；
图4是一个示例性的增强的自动缺陷分类系统的示意图；
图5是一个示例性的、可以实现本发明各方面的增强的自动缺陷分类系统的示意图。
具体实施方式
在半导体制造过程中，随着技术推进到小于20纳米的设计分辨率，缺陷的数目越来越多。缺陷分类，包括自动缺陷分类(ADC)，已被广泛用于半导体制造。在缺陷分类中，在对一个缺陷相关的缺陷图像的特性进行分析之后，可给该缺陷分配一个标签。通过为检查到的缺陷分配正确的标签，可以快速实施预定的措施以处理生产线上的缺陷、晶片或掩模版。此外，基于分类的缺陷分析可用于指导补救措施，以提高工艺和产量的制定。
这些缺陷往往是由系统条件造成的，如工艺变化、设计和工艺间的交互，以及先进的光刻技术(如OPC)。因此，由自动缺陷分类流程辨别的缺陷可被分门别类，如系统缺陷和随机缺陷。系统缺陷(如失真)往往由分辨率增强技术引起，它们可能是无害的且对制造流程不具致命影响。然而，系统缺陷也可能是致命的，例如，如果它们是由几何设计所引起且制造过程无法解决。另一方面，随机缺陷通常是由颗粒(例如晶片上的尘埃)、不均匀性或不规则性引起的。随机缺陷可能是无害的(例如，受随机缺陷影响的芯片局部不具任何电子学重要性)，也可能是致命的(例如，随机缺陷导致电路开路或短路)。除了系统缺陷和随机缺陷，其他类型的缺陷也可以被分类。
根据本文公开的实施例，一种用光刻仿真增强的自动缺陷分类工艺被用于向所述自动缺陷分类流程补充系统缺陷的信息，并引导自动缺陷分类流程给缺陷分配正确的分类。如下文详述，对晶片或掩模版(“目标试样”)执行检查，可以生成缺陷记录。缺陷的位置可从缺陷记录中提取，而从设计数据库中可以针对缺陷周围位置的区域提取包括对应的相关设计数据的数据库片段(database clip)。光刻仿真可以基于所述数据库片段来执行，以生成一个背景区域，而背景区域将用于与缺陷图像进行比较，以识别系统缺陷。除了系统缺陷，其他类型的缺陷也可以相应地识别。
图1是一个以光刻仿真增强自动缺陷分类的示例流程100的示意图。图1所示的增强的自动缺陷分类流程100可由运行于计算机系统的处理器的软件模块(如指令或代码)、计算机系统的硬件模块或两者的组合来执行。本文所描述的一个或多个步骤可以被集成进——例如——对晶片或掩模版执行检查的产品，并由半导体制造商所使用。
在步骤102中，基于对目标试样的检查，一个缺陷记录被接收。目标试样可以包括诸如晶片或掩模版。检查可以包括诸如光学或电子束检查。缺陷记录可以是检查报告的一部分，并且可以包括一个或多个缺陷。每个缺陷可以包括诸如位置和图像信息。位置可以包括有关缺陷位置的数据，如层信息和坐标。缺陷图像和/或位置可从缺陷记录中提取。
在步骤104中，对所述缺陷记录中一个缺陷的位置周围的一个区域，与其关联的相关设计数据被提取出设计数据库。设计数据库可以包括诸如至少一个 对应于目标样本——如晶片或掩模版——的设计数据库文件，还可以包括诸如芯片蓝图和多边形描述等。相关设计数据可以数据库片段的形式被提取，而其包含围绕缺陷位置的区域的相关数据。相关设计数据可包括与晶片或掩模版的设计有关的任何数据，诸如物理数据(如大小)、几何数据(如形状或布局)、技术数据、层数据，或上述的任意组合。
在步骤106中，光刻仿真基于所述相关设计数据被执行，以确定所述缺陷的位置周围的背景区域。例如，光刻仿真可以基于所述数据库片段运行，以产生围绕该缺陷位置的背景区域。背景区域可以包括一个仿真图像，而其基本上是基于所述光刻工艺的校准模型模拟的数据库图像。背景区域还可以包括诸如电路、走线的数据，和其他用光刻工艺生成或渲染的数据。
在步骤108中，所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像进行比较，以判断所述背景区域与所述缺陷图像是否匹配。背景区域与缺陷图像之间的比较可以通过——例如——将两个图像对齐后从一个中减去另一个。该比较也可以利用缺陷检查算法，而其可以根据应用的具体要求来确定。
如上所讨论，所述检查可以是光学检查或电子束检查。如果是光学检查，背景区域可以与缺陷的光学图像(来自光学检查)进行比较，而这类似于芯片对数据库(die-to-database)检查。如果是电子束检查，背景区域的轮廓可以与缺陷的扫描电子显微镜(Scanning Electron Microscope,SEM)图像的轮廓(来自电子束检查)进行比较。在一些实施例中，所述的各轮廓可以先被提取，再进行比较。
如果(背景区域的)仿真图像与缺陷图像相匹配，则表明缺陷图像可以通过仿真过程从设计数据库被重新创建。这意味着该缺陷更可能是由系统因素(如OPC或工艺变化等)所引起的，而非随机因素(如尘埃)。
在步骤110中，基于所述背景区域与所述缺陷图像相匹配的判断，所述缺陷被定义为系统缺陷。根据匹配的结果，所述缺陷可以被归类为系统缺陷或非系统缺陷。如果背景区域与缺陷图像被判定为不匹配，所述缺陷不会被归类为系统缺陷。在一些实施例中，所述缺陷可被归类为非系统缺陷(如随机缺陷或另一种类型的缺陷)，或根据自动缺陷分类决策的其他方面、基于所述背景和缺陷信息被分配一个特定的标签。
利用上述操作，原有的自动缺陷分类工艺得以增强，在处理缺陷时可以考虑系统缺陷的信息。如上所述，系统缺陷信息(或非系统缺陷信息)与其他相关信息(例如背景信息)一起，可被用于在自动缺陷分类决策过程中增强其流程。例如，包含在背景区域的信息，也可以被增强的自动缺陷分类工艺以其他方式使用。在一个实例中，利用预定的规则，某些缺陷可以被预先归类或消除。在另一实例中，系统缺陷在增强的自动缺陷分类工艺中可以利用预定的规则被自动归类。自动缺陷分类工艺通过光刻仿真增强(或辅助)时，自动缺陷分类流程中的分类器将得知诸如电路中缺陷的位置和电路的类型，从而得以智能判断缺陷对电路的影响，继而准确归类缺陷。此外，增强的自动缺陷分类工艺可以应用到上述实施例以外的情况，并可基于诸如由光刻仿真提供的缺陷周边环境的信息，用于归类所有种类的缺陷(如形状和色调)。
图2是一个以光刻仿真增强自动缺陷分类的示例流程200的示意图。在图1所示的过程200。图2所示的流程200可由运行于计算机系统的处理器的软件模块(如指令或代码)、计算机系统的硬件模块或两者的组合来执行。
根据图2所示的例子，增强的自动缺陷分类流程200包括从设计数据库204提取缺陷202位置周围的区域所关联的相关设计数据(数据库片段206)。例如，缺陷202可以从作为对目标试样(如晶片或掩模版)的检查结果(例如光学或电子束检查)的缺陷记录中获得。如上所述，光刻仿真可以基于所述区域所关联的相关设计数据(数据库片段206)来执行，以确定背景区域208。接着，增强的自动缺陷分类210可以将缺陷202的图像(“缺陷图像”)与背景区域208进行比较，以确定背景区域和缺陷图像是否匹配，和/或将所述缺陷归类为系统缺陷或非系统缺陷，以及其他事项。以这种方式，原有自动缺陷分类得以增强，在处理缺陷时可以考虑到系统缺陷的信息。
例如，晶片或掩模版的设计数据库可能是非常大的，可以包括成百上千GB的数据，而包含有缺陷202位置四周区域所关联的相关设计数据的数据库片段却相对较小，包含如大约一个或几个MB的数据。区域(或数据库片段206)的尺寸可以有所变化，例如基于形状或缺陷的大小。在一个例子中，区域的宽度和高度可以是32×32或64×64。
图3是另一个以光刻仿真增强自动缺陷分类的示例流程300的示意图。 图3所示的流程300可由运行于计算机系统的处理器的软件模块(如指令或代码)、计算机系统的硬件模块或两者的组合来执行。
基于图像相似性，缺陷可以进行分组。增强的自动缺陷分类流程300与增强的自动缺陷分类流程200有所不同，不同在于来自缺陷记录的缺陷可被分类成组。例如，所有在一个缺陷组中的缺陷(如缺陷组302)可以具有相同或相似的图像。每个缺陷组可以有一个代表缺陷304；其可为随机选择的缺陷，或者基于一个或多个预定规则所指定的缺陷。例如，一个新缺陷可以与每个缺陷组的代表缺陷304的图像(“代表缺陷图像”)进行比较，以确定该新缺陷所属的缺陷组。
一旦所有来自缺陷记录的缺陷被处理和分类成组，某组的代表缺陷图像可以与对应于该组的代表缺陷的背景区域208(如在图1的操作106确定)进行比较(如在图1的操作108)。相较于比较每个缺陷图像与背景区域，比较组里的代表缺陷图像与相应的背景区域可以大大减少计算工作。例如，该比较结果可被用于增强的自动缺陷分类210，而其对同组内的每个缺陷将处理相同的信息。例如，当某个缺陷组的代表缺陷图像被判定为与针对该缺陷组生成的背景区域相匹配，该组内的所有缺陷都可被归类为系统缺陷而无需重复所有操作。
在图示的例子中，增强的自动缺陷分类流程300包括从设计数据库204提取缺陷组302的代表缺陷304位置周围的区域所关联的相关设计数据(数据库片段206)。如上所述，光刻仿真可以基于所述区域所关联的相关设计数据(数据库片段206)来执行，以确定背景区域208。接着，增强的自动缺陷分类210可以将缺陷组302的代表缺陷图像与背景区域208进行比较，以确定背景区域和代表缺陷图像是否匹配，和/或将所述缺陷组的所有缺陷归类为系统缺陷或非系统缺陷，以及其他事项。
图4是是一个示例性的增强的自动缺陷分类系统400的示意图。本文公开的内容，例如流程100、200、300的操作，可被实现为增强的自动缺陷分类系统400内的软件和/或硬件模块。例如，增强的自动缺陷分类系统400可以包括一个装置，诸如计算设备，其可以实现为任意配置的一个或多个计算机，诸如微型计算机、大型计算机、超级计算机、通用计算机、专用/特殊目的计 算机、集成计算机、数据库计算机、远程服务器计算机、个人计算机，或由计算服务提供商——如网络主机或云服务提供商——提供的计算服务。在一些实施例中，计算设备可以实现为处于不同地理位置的多组计算机的形式。
在图示的例子中，增强的自动缺陷分类系统400可以包括缺陷检查模块402、设计数据库404、光刻分析模块406，以及自动缺陷分类模块408。
缺陷检查模块402可被配置用于对一个目标试样(如晶片或掩模版)执行检查(如光学或电子束检查)，并生成包括缺陷记录的检查报告。如上所述，缺陷记录可以包括缺陷的列表，且每个缺陷可以包括图像和位置信息。在一些实施例中，缺陷可以被分类成组，组里的代表缺陷可用于比较，而不是比较每个单独的缺陷，如图3中所讨论。
设计数据库404可以是存有晶片或掩模版设计数据的任何数据库。如前所述，缺陷位置周围的区域的相关数据(本文中“数据库片段”)可从设计数据库404中被选取和取回。
光刻分析模块406可以基于数据库片段进行光刻仿真，以生成背景区域。如上所述，背景区域可以包括缺陷位置周围的各种背景信息，例如作为光刻流程结果的仿真数据库图像。
自动缺陷分类模块408可以使用背景区域(由光刻分析模块406产生)和缺陷信息(由缺陷检查模块402生成)为其决策。例如，自动缺陷分类模块408可以比较缺陷图像和对应的背景区域：如果相匹配，该缺陷被识别为系统缺陷。在一些实施例中，某组的代表缺陷图像可以与对应的背景区域进行比较；如果相匹配，该组(以及该组内的所有缺陷)被识别为系统缺陷。包括在背景区域的信息也可以被自动缺陷分类模块408以其他方式使用，例如，某些缺陷可以利用预定的规则进行预先归类或排除。
图5是一个示例性的、可以实现本发明各方面的增强的自动缺陷分类系统500的示意图。例如，增强的自动缺陷分类系统500可包括一个装置，如一个计算设备。在一些实施例中，计算设备可以实现为任意配置的一个或多个计算机，诸如微型计算机、大型计算机、超级计算机、通用计算机、专用/特殊目的计算机、集成计算机、数据库计算机、远程服务器计算机、个人计算机，或由计算服务提供商——如网络主机或云服务器——提供的计算服务。在一些 实施例中，计算设备实现的形式可以为处于不同地理位置的、互相或互不通信的多组计算机——如通过一个网络进行通信。某些操作可以由多个计算机共同分担，但在一些实施例中，不同的计算机可以被分配不同的操作。
所述计算设备可以具有内在的硬件配置，包括处理器502和存储器504。处理器502可以包括至少一个处理单元，诸如中央处理单元(CPU)或任何其它类型的设备、或多个设备，其可操纵或处理目前存在的或以后开发的信息。虽然本文的实施例可以如图所示用单一的处理器来实施，但利用多个处理器可以达成速度和效率上的优势。例如，处理器502可以分布在多台机器或设备上(每个机器或设备具有一个或多个处理器)，而这些机器或设备可以直接连接，或通过局域网或其他网络进行连接。存储器504可以是随机存取存储器设备(RAM)、只读存储器设备(ROM)、光盘、磁盘，或任何其它适当类型的存储装置。在一些实施例中，存储器504可分布在多台机器或设备上，诸如基于网络的存储器或在多个机器上执行操作的存储器——为了便于说明，可以在此描述为使用单个计算机或多个计算设备执行。在一些实施例中，存储器504可存储处理器502可用总线来访问的代码和数据。例如，存储器504可以包括处理器502可用总线512访问的数据5042。
存储器504还可以包括操作系统5046和安装的应用5044。应用5044包括程序，以允许处理器502实现指令来为本文所述的增强的自动缺陷分类系统500产生控制信号、实现功能。所述指令还可以包括为分类而处理不属于自动缺陷分类系统的缺陷信息，诸如产生可被自动缺陷分类系统用于生成系统缺陷标签的背景信息。增强的自动缺陷分类系统500还可以包括辅助、额外或外部存储器506，例如存储卡、闪存驱动器、外部硬盘驱动器、光盘驱动器，或任何其它形式的计算机可读介质。在一些实施例中，应用5044可以整体或部分存储在存储器506中，并在处理需要时加载到存储器504。
增强的自动缺陷分类系统500可包括一个或多个输出设备，诸如输出508。输出508可以实现为各种方式——例如，它可以是连接到增强的自动缺陷分类系统500的显示器，被配置用于显示视频数据的渲染。输出508可以是任何向用户传输视觉、听觉或触觉信号的设备，诸如显示器、触敏设备(如触摸屏)、扬声器、耳机、发光二极管(LED)指示器，或振动马达。例如，如果输出 508是显示器，它可以是一个液晶显示器(LCD)、阴极射线管(CRT)、或任何其他能够向个人提供可见输出的输出设备。在某些情况下，输出设备也可作为输入设备——例如，配置成接收基于触摸的输入的触摸屏显示器。
输出508可以替换地或额外地构成用于传输信号和/或数据的通信设备。例如，输出508可以包括用于从增强的自动缺陷分类系统500向其他设备传输信号或数据的有线手段。又例如，输出508可包括使用兼容于无线接收器的协议的无线发射器，用以从增强的自动缺陷分类系统500向其他设备传输信号。
增强的自动缺陷分类系统500可包括一个或多个输入设备，例如输入510。输入510可以实现为各种方式，诸如键盘、数字键盘、鼠标、麦克风、触敏设备(如触摸屏)、传感器，或手势输入装置。任何其他类型的输入设备都是可能的，包括不要求用户干预的输入装置。例如，输入510可以是通信设备，诸如根据任何无线协议接收信号操作的无线接收器。输入510可以输出——如沿总线512——指示输入的信号或数据到增强的自动缺陷分类系统500中。
可选地，增强的自动缺陷分类系统500可以用通信设备(如通信设备514)通过网络(如网络516)与其他设备通信。网络516可以是任何适当类型的一个或多个通信网络的任意组合，包括但不限于使用蓝牙、红外线、近场连接(NFC)、无线网络、有线网络、局域网(LAN)、广域网(WAN)、虚拟专用网(VPN)、蜂窝数据网及因特网通信的网络。通信设备514可以实现为各种方式，诸如应答器/收发器、调制解调器、路由器、网关、电路、芯片、有线网络适配器、无线网络适配器、蓝牙适配器、红外线适配器、NFC适配器、蜂窝网络芯片，或任何适当类型的装置的任意组合，其利用总线512连接到增强的自动缺陷分类系统500以通过网络516提供通信的功能。
增强的自动缺陷分类系统500可以与晶片或掩模版的检查设备进行通信。例如，增强的自动缺陷分类系统500可连接到一个或多个经配置用于生成晶片或掩模版检查结果(如缺陷记录或报告)的晶片或掩模版检查设备。
增强的自动缺陷分类系统500(及存储其上和/或由其执行的算法、方法、指令等)可以在硬件上实现，包括例如知识产权(IP)核心、应用专用集成电路(ASIC)、可编程逻辑阵列、光学处理器、可编程逻辑控制器、微代码、固件、微控制器、服务器、微处理器、数字信号处理器，或任何其它合适的电路。 在权利要求中，术语“处理器”应被理解为包括单独或组合使用的任意前述装置。术语“信号”和“数据”可互换使用。此外，增强的自动缺陷分类系统500的各部分不必以相同的方式来实现。
在一些实施例中，增强的自动缺陷分类系统500可用带计算程序的通用计算机/处理器来实现；其在执行时，实现任意在此描述的各方法、算法和/或指令。例如，可以额外地、或可替代地使用专用计算机/处理器，其可包含专用硬件用于执行任意本文描述的方法、算法或指令。
本文已经结合特定实施例和实现方式进行了描述，但是应当理解，本发明并不限于所公开的实施例；相反，其意在覆盖包括在所附权利要求的所述范围内的各种修改和等同布置，所述范围应被赋予最宽的解释，以包含所有依法允许的类似修改及等同结构。

Claims (18)

  1. 一种用于在自动缺陷分类流程中管理缺陷的方法，包括：
    基于对目标试样的检查，接收一个缺陷记录；
    从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域的仿真图像；
    通过所述处理器比较所述背景区域的仿真图像与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；
    基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
  2. 根据权利要求1所述的方法，进一步包括：
    基于所述背景区域与所述缺陷图像不匹配的判断，定义所述缺陷为非系统缺陷；
    将所述缺陷分类到数个缺陷组中的一个缺陷组，基于所述缺陷图像与所述缺陷组的代表缺陷图像的相似度。
  3. 根据权利要求2所述的方法，其特征在于，
    从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据包括：
    从所述设计数据库中，对所述缺陷记录中所述缺陷组的所述代表缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    所述通过所述处理器比较所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配包括：
    比较所述缺陷组的所述代表缺陷图像与所述背景区域。
  4. 根据权利要求3所述的方法，其特征在于，所述基于所述背景区域与 所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷包括：
    基于所述背景区域与所述缺陷组的所述代表图像匹配的判断，定义所述缺陷组中的每一个缺陷为系统缺陷。
  5. 根据权利要求1所述的方法，其特征在于，所述自动缺陷分类流程是针对所述缺陷记录而执行，并在决策中考虑所述系统缺陷；所述系统缺陷通过在所述自动缺陷分类流程中使用预定规则而被自动分类。
  6. 根据权利要求1所述的方法，其特征在于，所述目标试样包括晶片或掩模版；
    所述检查包括光学检查或电子束检查,并且所述通过所述处理器比较所述背景区域与所述缺陷图像以判断所述背景区域与所述缺陷图像是否匹配包括：
    如果所述检查包括电子束检查，比较所述背景区域的轮廓与所述缺陷的扫描电子显微镜(SEM)图像的轮廓；
    如果所述检查包括光学检查，比较所述背景区域与所述缺陷的光学图像。
  7. 一种用于在自动缺陷分类流程中管理缺陷的系统，所述系统包括：
    缺陷检查模块，用于基于对目标试样的检查，接收一个缺陷记录；
    设计数据库，用于存储晶片或掩模版设计数据；
    光刻分析模块，用于从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据，并且，通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域的仿真图像；
    自动缺陷分类模块，用于通过所述处理器比较所述背景区域的仿真图像与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配，并且，基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
  8. 根据权利要求7所述的系统，其特征在于，
    所述自动缺陷分类模块，还用于基于所述背景区域与所述缺陷图像不匹配的判断，定义所述缺陷为非系统缺陷；还用于将所述缺陷分类到数个缺陷组中的一个缺陷组，基于所述缺陷图像与所述缺陷组的代表缺陷图像的相似度。
  9. 根据权利要求8所述的系统，其特征在于，
    从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据包括：
    从所述设计数据库中，对所述缺陷记录中所述缺陷组的所述代表缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    所述通过所述处理器比较所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配包括：
    比较所述缺陷组的所述代表缺陷图像与所述背景区域。
  10. 根据权利要求9所述的系统，其特征在于，所述基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷包括：
    基于所述背景区域与所述缺陷组的所述代表图像匹配的判断，定义所述缺陷组中的每一个缺陷为系统缺陷。
  11. 根据权利要求7所述的系统，其特征在于，所述自动缺陷分类流程是针对所述缺陷记录而执行，并在决策中考虑所述系统缺陷；所述系统缺陷通过在所述自动缺陷分类流程中使用预定规则而被自动分类。
  12. 根据权利要求7所述的系统，其特征在于，所述目标试样包括晶片或掩模版；
    所述检查包括光学检查或电子束检查,并且所述通过所述处理器比较所述背景区域与所述缺陷图像以判断所述背景区域与所述缺陷图像是否匹配包括：
    如果所述检查包括电子束检查，比较所述背景区域的轮廓与所述缺陷的扫 描电子显微镜(SEM)图像的轮廓；
    如果所述检查包括光学检查，比较所述背景区域与所述缺陷的光学图像。
  13. 一种用于在自动缺陷分类流程中管理缺陷的系统，所述系统包括：
    一个处理器；
    一个连接到所述处理器的存储器，所述存储器经配置以存储一套指令集，所述指令集被所述处理器执行时，可被所述处理器操作，以：
    基于对目标试样的检查，接收一个缺陷记录；
    从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域；
    通过所述处理器比较所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；
    基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
  14. 根据权利要求13所述的系统，其特征在于，所述存储器经进一步配置以存储一套指令集，所述指令集被所述处理器执行时，可被所述处理器操作，以：
    将所述缺陷分类到数个缺陷组中的一个缺陷组，基于所述缺陷图像与所述缺陷组的代表缺陷图像的相似度；
    从所述设计数据库中，对所述缺陷记录中所述缺陷组的所述代表缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    比较所述缺陷组的所述代表缺陷图像与所述背景区域；
    基于所述背景区域与所述缺陷组的所述代表图像匹配的判断，定义所述缺陷组中的每一个缺陷为系统缺陷。
  15. 根据权利要求13所述的系统，其特征在于，所述目标试样包括晶片 或掩模版；
    所述检查包括光学检查或电子束检查,并且所述通过所述处理器比较所述背景区域与所述缺陷图像以判断所述背景区域与所述缺陷图像是否匹配的指令包括以下指令，以：
    如果所述检查包括电子束检查，比较所述背景区域的轮廓与所述缺陷的扫描电子显微镜(SEM)图像的轮廓；
    如果所述检查包括光学检查，比较所述背景区域与所述缺陷的光学图像。
  16. 一种非瞬时的储存有一套指令集的计算机可读介质，所述指令集被计算机系统的处理器执行时，可被所述处理器操作以在增强自动缺陷分类流程中管理缺陷，所述非瞬时的计算机可读介质包括指令，以：
    基于对目标试样的检查，接收一个缺陷记录；
    从设计数据库中，对所述缺陷记录中一个缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    通过处理器，基于所述区域所关联的所述相关设计数据执行光刻仿真，以确定一个背景区域；
    通过所述处理器比较所述背景区域与所述缺陷记录中所述缺陷的一个缺陷图像以判断所述背景区域与所述缺陷图像是否匹配；
    基于所述背景区域与所述缺陷图像相匹配的判断，定义所述缺陷为系统缺陷。
  17. 根据权利要求16所述的非瞬时的计算机可读介质，进一步包括指令，以：
    将所述缺陷分类到数个缺陷组中的一个缺陷组，基于所述缺陷图像与所述缺陷组的代表缺陷图像的相似度；
    从所述设计数据库中，对所述缺陷记录中所述缺陷组的所述代表缺陷的位置周围的一个区域，提取与其关联的相关设计数据；
    比较所述缺陷组的所述代表缺陷图像与所述背景区域；
    基于所述背景区域与所述缺陷组的所述代表图像匹配的判断，定义所述缺 陷组中的每一个缺陷为系统缺陷。
  18. 根据权利要求16所述的非瞬时的计算机可读介质，其特征在于，所述目标试样包括晶片或掩模版；
    所述检查包括光学检查或电子束检查,并且所述通过所述处理器比较所述背景区域与所述缺陷图像以判断所述背景区域与所述缺陷图像是否匹配的指令包括以下指令，以:
    如果所述检查包括电子束检查，比较所述背景区域的轮廓与所述缺陷的扫描电子显微镜(SEM)图像的轮廓；
    如果所述检查包括光学检查，比较所述背景区域与所述缺陷的光学图像。
PCT/CN2016/097168 2016-07-13 2016-08-29 在自动缺陷分类流程中管理缺陷的方法及系统 WO2018010272A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/285,955 US9928446B2 (en) 2016-07-13 2016-10-05 Augmented automatic defect classification

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610551876.4 2016-07-13
CN201610551876.4A CN106204598B (zh) 2016-07-13 2016-07-13 在自动缺陷分类流程中管理缺陷的方法及系统

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/285,955 Continuation US9928446B2 (en) 2016-07-13 2016-10-05 Augmented automatic defect classification

Publications (1)

Publication Number Publication Date
WO2018010272A1 true WO2018010272A1 (zh) 2018-01-18

Family

ID=57477480

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/097168 WO2018010272A1 (zh) 2016-07-13 2016-08-29 在自动缺陷分类流程中管理缺陷的方法及系统

Country Status (2)

Country Link
CN (1) CN106204598B (zh)
WO (1) WO2018010272A1 (zh)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10713776B2 (en) * 2017-04-13 2020-07-14 Instrumental, Inc. Method for predicting defects in assembly units
CN107590511A (zh) * 2017-08-30 2018-01-16 武汉华星光电技术有限公司 一种缺陷识别方法以及用于自动检查机的缺陷识别系统
CN108037739A (zh) * 2017-11-09 2018-05-15 广州兴森快捷电路科技有限公司 Pcb产品的过程管控方法和系统
CN111542855A (zh) * 2017-12-31 2020-08-14 Asml荷兰有限公司 用于缺陷检查和复验的方法和系统
CN108257166B (zh) * 2018-01-11 2022-03-04 上海华虹宏力半导体制造有限公司 版图的仿真图像和硅片sem图像自动匹配的方法
US10605745B2 (en) * 2018-06-28 2020-03-31 Applied Materials Israel Ltd. Guided inspection of a semiconductor wafer based on systematic defects
CN115575411A (zh) * 2022-09-28 2023-01-06 东方晶源微电子科技（北京）有限公司 晶片的缺陷检测方法、装置、设备及计算机可读存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8204297B1 (en) * 2009-02-27 2012-06-19 Kla-Tencor Corp. Methods and systems for classifying defects detected on a reticle
CN103748670A (zh) * 2011-09-07 2014-04-23 株式会社日立高新技术 区域决定装置、观察装置或检查装置、区域决定方法以及使用了区域决定方法的观察方法或检查方法
CN104568979A (zh) * 2013-10-23 2015-04-29 旺宏电子股份有限公司 芯片对数据库的影像检测方法
TWI494558B (zh) * 2008-12-23 2015-08-01 Hermes Microvision Inc 識別晶圓缺陷的方法與系統
US20150221076A1 (en) * 2014-02-06 2015-08-06 Kla-Tencor Corporation Defect detection and classification based on attributes determined from a standard reference image

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003263841A (ja) * 2002-03-08 2003-09-19 Sony Corp 信号処理装置及び信号処理方法、並びに情報再生装置及び情報再生方法
US7788629B2 (en) * 2004-07-21 2010-08-31 Kla-Tencor Technologies Corp. Systems configured to perform a non-contact method for determining a property of a specimen
SG163442A1 (en) * 2009-01-13 2010-08-30 Semiconductor Technologies & Instruments System and method for inspecting a wafer
CN102142355B (zh) * 2010-02-02 2013-07-17 吕一云 物体制造缺陷的应用方法
JP5957357B2 (ja) * 2012-10-15 2016-07-27 株式会社日立ハイテクノロジーズ パターン検査・計測装置及びプログラム
CN103439339B (zh) * 2013-09-02 2015-11-25 深圳市华星光电技术有限公司 贴附有偏振片的液晶面板的缺陷检测装置及缺陷检测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI494558B (zh) * 2008-12-23 2015-08-01 Hermes Microvision Inc 識別晶圓缺陷的方法與系統
US8204297B1 (en) * 2009-02-27 2012-06-19 Kla-Tencor Corp. Methods and systems for classifying defects detected on a reticle
CN103748670A (zh) * 2011-09-07 2014-04-23 株式会社日立高新技术 区域决定装置、观察装置或检查装置、区域决定方法以及使用了区域决定方法的观察方法或检查方法
CN104568979A (zh) * 2013-10-23 2015-04-29 旺宏电子股份有限公司 芯片对数据库的影像检测方法
US20150221076A1 (en) * 2014-02-06 2015-08-06 Kla-Tencor Corporation Defect detection and classification based on attributes determined from a standard reference image

Also Published As

Publication number Publication date
CN106204598A (zh) 2016-12-07
CN106204598B (zh) 2019-02-05

Similar Documents

Publication Publication Date Title
WO2018010272A1 (zh) 在自动缺陷分类流程中管理缺陷的方法及系统
JP6905954B2 (ja) 自動欠陥分類のための未知欠陥除去の最適化
US11935722B2 (en) Machine learning on wafer defect review
TWI755613B (zh) 基於機器學習之圖案分組方法
US20190139212A1 (en) Inspection apparatus, data generation apparatus, data generation method, and data generation program
WO2018035878A1 (zh) 缺陷分类方法和缺陷检查系统
TWI709105B (zh) 用於影像分析之系統、方法及非暫時性電腦可讀儲存媒體
TW202013314A (zh) 確定晶圓檢查參數方法
US20150110384A1 (en) Image inspection method of die to database
CN111584397A (zh) 半导体元件的制造过程中进行自动缺陷筛选的系统
US10146036B2 (en) Semiconductor wafer inspection using care area group-specific threshold settings for detecting defects
US20180218492A1 (en) Method And System For Identifying Defects Of Integrated Circuits
US10719655B2 (en) Method and system for quickly diagnosing, classifying, and sampling in-line defects based on CAA pre-diagnosis database
CN105917455B (zh) 通过利用交替失效模式的标称特性的图案失效发现
US10140400B2 (en) Method and system for defect prediction of integrated circuits
TW201445347A (zh) 智慧型弱點圖形診斷方法、系統與電腦可讀取記憶媒體
US20200334807A1 (en) System and Method for Generation of Wafer Inspection Critical Areas
TWI663667B (zh) 用於在檢查物件時自我調整取樣的方法和其系統
CN108780051B (zh) 用于界定设计数据的重复结构中的关注区域的系统及方法
US10191112B2 (en) Early development of a database of fail signatures for systematic defects in integrated circuit (IC) chips
US9928446B2 (en) Augmented automatic defect classification
TWI402928B (zh) 智慧型缺陷篩選及取樣方法
TWI732803B (zh) 用於儲存動態層內容於一設計檔案中之方法、系統及非暫時性電腦可讀媒體
JP2004087820A (ja) 検査条件データ管理方法及びシステム並びにプログラム、検査装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16908597

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16908597

Country of ref document: EP

Kind code of ref document: A1