KR20060100489A - Yield management system for bare wafer using wafer map and statistical analysis function - Google Patents

Abstract

The present invention relates to a bare wafer yield management system for manufacturing various semiconductors, which is connected to a network of semiconductor bare wafer defect inspection measuring apparatuses (10) to collect data from the equipment, converts it into a KLA RF standard data format, and uploads the data. A loader part 100 serving as a role; Connect to the TCP / IP shared network to receive and analyze the data uploaded from the loader unit 100, summarize the summary information into a DB, and store the data necessary to draw the map in a file, A server part (Server Part) 200 for summarizing in advance a database of accurate wafer information acquired from the MES server and information to be used for statistical analysis (SPC); The client unit 300 is connected to the server unit 200 and outputs a necessary inquiry and analysis result as a map, a chart, a report, and the like.

Bare, Wafer, Yield, Bad, Analysis, Map, SPC, Flatness

Description

Yield Management System for Bare Wafer using wafer map and statistical analysis function}

1 is a schematic diagram of a bare wafer yield management system of the present invention;

2 is an overall configuration diagram of a bare wafer yield management system according to the present invention.

3 is a detailed configuration diagram of the server unit according to the present invention

4 is a detailed configuration diagram of the client unit according to the present invention

5 is a data flow diagram of a data analysis system according to the present invention;

6 is an exemplary map according to the present invention.

7 is an exemplary chart according to the present invention.

8 is an exemplary view of a table and statistical analysis according to the present invention.

9 is an exemplary report according to the present invention.

<Description of the symbols for the main parts of the drawings>

10: measuring equipment 100: loader

110: measuring device interface unit 120: data return unit

130: upload unit 200: server unit

210: DB loader module 211: Raw data file input unit

212: Data Summary Module 213: System Monitoring Module

220: client service module 230: MES interface module

240: message handling module 250: database

300: client unit 310: query module

320: result processing unit 330: user interface module

340: management module 350: result data module

351: display module 361: flat data analysis unit

362: particle data analysis unit 400: MES server unit

410: MES server 420: message server

The present invention relates to a bare wafer yield management system using a bare wafer map (Swafer Map) drawing and the SPC function, by connecting the wafer measuring equipment to a network to show the results of all wafers passing through each process and equipment. The present invention relates to a bare wafer yield management system that analyzes information that can be used to estimate the cause of defects at a glance, and in particular, analyzes wafer map and statistical analysis (SPC) results.

In the manufacture of bare wafers, various measurements and analyzes are performed to check for defects in bare wafers in each process. In the conventional work method, since each measurement and analysis device was configured independently and a separate program was used for each device maker, there was no correlation between data and a clear batch analysis was impossible. Since different equipment manufacturers have different standards, the results from different equipment have to be converted by the user to compare the results with the same standards.

Also, if you do not have a separate analytical program, to review the instrument output data is the only way to check directly on the instrument, and some instruments do not have the ability to recall and display the stored results. There was a problem that was impossible.

In the past, it was inconvenient to copy the data from the equipment one by one and save the results, so that each user needed to integrate them into one report.

In addition, since the measurement equipment can measure only one measurement item, there is an inconvenience in that it is necessary to measure again when one wants to know the result value of another measurement item.

The present invention is to improve the conventional problems as described above, by using the raw data (RAW) generated from a variety of equipment to the data (thickness, flatness, shape, etc.) required for the actual analysis The goal is to provide a bare wafer yield management system that implements algorithms that collect and automatically compute over the network, and comprehensively analyzes and displays the results that users demand through wafer maps and SPC capabilities.

The yield management system according to the present invention collects data generated from various equipment through a network for various defect analysis and quality improvement activities, thereby reducing unnecessary efforts generated in existing work methods in which the user directly intervenes and inherently analyzing activities. It aims to help you focus on quality improvement activities.

It is not only to store the measured results by each device but to manage the data systematically through the integrated database to make the analysis faster and easier, and to enable various high quality analysis desired by the user.

In addition, based on existing statistical analysis tools and statistical theories, algorithms are provided to identify the causes of problem processes, ingots, lots, slots, and equipment using data such as characteristics, dimensions, and yields.

The bare wafer yield management system of the present invention receives a loader part connected through various networks and measuring devices for semiconductor wafer defect inspection and a measurement result collected from the measurement device through the loader, and measuring equipment. It is a server part that matches information with semiconductor wafer information, parses and summarizes data for analysis processing, and stores and manages the result, and a client part that is connected to the server part and processes necessary analysis work. There is a feature that is configured.

The loader part is connected to various measuring devices and networks (using various protocols), collects raw data generated by the device, converts the data into a designated standard format, and then transfers the data to the server part. Configured to transmit.

The server part parses and receives the data transmitted from the loader, receives the correct wafer information and production measurement equipment information from the MES server, and matches the measured information with the given information. In addition, the measurement information is parsed, stored and managed, and the information necessary for statistical analysis (SPC) is summarized into a database and managed.

In the client part, statistical analysis is performed using the statistical analysis (SPC) function or defect analysis is performed by drawing a measurement result of each process / equipment in a wafer map (when the defective part or measurement is performed. Results are displayed on the map). The data necessary to draw the map is made into DB, utilizing the stored reference information and the map data stored in the form of a file, and configured to download from the server to the client when necessary.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic diagram of a bare wafer yield management system of the present invention. As shown in this figure, the loader is connected to various measurement equipment 10 for the semiconductor wafer defect inspection and the network (Private Network) to collect the RAW data from the measurement equipment 10 to convert the data into a specific format upload It is connected to the loader part 100 and the TCP / IP sharing network, receives the measurement results collected from the measuring device 10 through the loader part 100, and MES the wafer basic information and the equipment related data. (manufacturing excutive system) matching the result information retrieved from the server unit 400 with the information of the loader unit 100, parsing (sparing) the data, storing and managing the data summary processing for statistical analysis processing, etc. The server part 200 and a client part connected to the server part 200 by a network and for processing dimension and characteristic data analysis and particle data analysis by a user interface ( 300).

Here, the MES server unit 400 is a system that manages all parts of the production process, and is composed of a DBMS 410 and a message server 420 for storing MES-related data, and the wafer information and information of each device The server unit 200 of the present invention retrieves and automatically loads the server unit 200 to match the RAW data and information input from the loader unit 100 to summarize data for parsing and analysis. Do it.

 2 is an overall configuration diagram of a bare wafer yield management system according to the present invention. As shown in the drawing, a plurality of measuring equipment 10, the loader unit 100, the server unit 200, and the client unit 300 is composed of, the server unit 200 and the measurement-related and wafer information It is configured to receive and retrieve from the MES server unit 400 to use in the analysis process.

The measuring device 10 is a variety of equipment for measuring the characteristics, flatness, thickness value and type and quantity of particles.

The loader part 100 is connected to various measuring devices 10 through a network (at this time, using various protocols) to measure raw (RAW) data generated by the measuring device interface 110. Collected through the data conversion unit 120 is converted to a designated format through the data conversion unit 120 and then transmitted to the server unit 200 through the uploader unit 130.

The server part 200 includes a DB loader module 210, a client service module 220, a MES interface module 230, a message handling module 240, and a database itself (DBMS) 250. After receiving and analyzing the data uploaded from the loader unit 100, the summary information is converted into a DB, and the data necessary to draw a map is stored in a file, and accurate wafer information is provided using the given information. Obtained from DB and stored in its own DB, it is configured to summarize the information necessary for statistical analysis (SPC) to make a DB.

The client part 300 may include a result processing unit 320 including a query module 310, a map module, a report module, a statistical analysis module, and a user interface module 350 including a display. It is configured to include and perform statistical analysis by using the statistical analysis (SPC) function, or to perform the defect analysis work by drawing the measurement results of each process / equipment in a wafer map (when the defective part or measurement result is Displayed on the map). The data required to draw the map is composed of DB and stored standard information and map data stored in the form of a file, and downloaded from the server to the client when necessary.

The MES server unit 400 is composed of a MES database (DBMS) 410 that manages wafer information and statistical process analysis information (SPC), and a message server 420 that manages wafer messages transmitted from ECS. As for the external system, it is the basic equipment of wafer measurement.

3 is a detailed configuration diagram of the server unit according to the present invention. As shown in FIG.

The raw data file input unit 211 which inspects the format of the data file uploaded from the loader unit 100 and parses and inputs the raw data file according to a rule determined for each device, and the input unit 211. DB loader module that stores the data input from the database and obtains detailed information about the wafer from the MES server unit 400 using the message handling module 240 and the MES interface module 230 and stores the data in the DAS database. 210, a client service module 220 for interfacing through the client unit 300 and the input / output interface unit 221 to process data and reference information management requested by the client, and the MES server unit 400. MES interface for acquiring necessary information from server by interfacing wafer information and data for statistical analysis by interfacing with Interface with the message unit 230 and the message server 420 of the MES server unit 400 to receive and process the message sent from the ECS to manage the measurement equipment through the message server 420, the wafer is measured one Message handling module 240 that manages wafer-related information (Lot ID, Slot, Filename etc) that is transmitted every time, and data summary module 212 that summarizes and accumulates data accumulated for a certain period for easy statistical processing. ), A system monitoring module 213 for monitoring the entire system and notifying a user before a shortage occurs, and a database 250 for storing and managing various types of information.

 4 is a detailed block diagram of the client unit according to the present invention.

The client unit 300 allows a user to input various conditions for extracting data to be analyzed, and creates a SQL based on the input conditions to request information from the server unit 200 and the query module 310. A flatness data analyzer 361 analyzing the flatness data among data received by the request of the query module 310, and analyzing particle data among the data received by the request of the query module 310. Particle data analysis unit 362, the user module for each user, various limits (Limit) input, the scale management and management module for managing various settings, and received from the server unit 200 Result data module 350 for storing and managing the result data and the analysis data of each analysis unit, and the result data received from the server unit stored in the result data module 350 and the analyzed result Including a result processing unit 320 for processing the results using a map, chart, statistical analysis (SPC) and reports using the data, and a display module 351 for outputting the processing results of the result processing unit 320 to the screen It is composed.

Here, the MES server unit 400 is a system that manages all parts of the production process, the MES server unit 400 is a wafer and measuring equipment related data generated based on the measurement of the measuring equipment 10 of the present invention The server unit 200 is automatically loaded, and the wafer and measurement equipment information required by the server unit 200 of the present invention is retrieved to receive data. The DAMS database of the MES server unit 400 is a database that stores MES-related data, and the message server 420 is a server that manages a message system that exchanges information between the device and the server, and the server and the server. The message server 420 may exist as a separate server or may be included in the MES server 410. As described above, the present invention receives the information of the wafer and the equipment by using the MES server unit 400 and uses the measurement information for matching and analyzing the information.

In the present invention configured as described above, the data loader 100 collects data from each measurement device 10 and uploads the data to the server 200. The data loader unit 100 is connected to each device and the loader unit 100 so that the measuring device interface unit 110 can receive or import necessary data files, and various OSs such as DOS, Unix, Windows NT, NFS, It is designed to interface with any OS environment by supporting protocols such as FTP and Net BEUI.

5 is a data flowchart of a data analysis system according to the present invention. As shown in the drawing, the measurement device 10 receives a device control signal from an equipment control system (ECS) and receives data related to device control and measurement from an MES server to perform wafer measurement. As the first step, when the measurement device 10 stores the measurement result, the measurement result file (Raw Data File) is read and data conversion is performed as the second step. When data conversion is performed, the conversion result file is stored, and the conversion result is transmitted from the data loader unit as a third step. At this time, the server unit 200 inquires wafer reference information and various production information from the server unit MES server. When the inquiry result is returned, the server unit 200 makes a statistical analysis as a fourth stage while making database of the measurement result data uploaded in the third step and the reference information and the statistical information based on the inquiry result data. Generate data. In addition, the map data is extracted from the data, the result processing unit generates a map data file, and the user performs statistical analysis (SPC) processing.

As described above, when the measuring device interface 110 interfaces with the measuring devices 10 to collect data, the data converting unit 120 converts the file into a file of a standard format.

The data converting unit 120 converts the raw data file from each measuring device 10 into a file of a standard format that is convenient to use (here, the standard format file is analyzed so that the system of the present invention can analyze and see the result processing). Use any standard format for this purpose).

In addition, the data upload unit 130 provides a file automatic deletion function transmitted to transmit the converted file to a desired server, which automatically deletes an old file when a predetermined period of time is reached. In addition, it supports multiple servers so that files can be sent to multiple servers at once. In case of error, resend function is supported to distinguish by server.

When the data loader 100 collects and uploads the data as described above, the server unit 200 analyzes the data and processes the data. The operation of each function of the server unit 200 will be described below.

First, the data is input from the DB loader module 210 through the raw data file input unit 211, and standardizes the input data, analyzes the raw data, adds the analyzed data to the DB, cleans up the loaded raw file, and Storage and so on. The data summarization module 212 processes the entered flatness and particle values of the particle data (Min, Max, Mean, Std, etc.) and calculates the values required by the client in advance. It can improve the speed of experience and calculate statistical values to handle the functions requested by the client. The MES interface (I / F) module 230 is in charge of a function for cooperating with the MES server of the MES server unit 400, and takes the necessary information from the MES server into a DB, and the history of Lot and Wafer. Identify information, identify wafer process by process / equipment, and yield status by process / equipment.

In addition, the system monitoring module 213 grasps the hard disk (HDD) storage capacity and the DB capacity of the server and notifies the manager when a capacity shortage is predicted, and notifies the manager when data abnormality is found. This follows the predetermined criteria. The client service module 220 interfaces the client unit 300 with the input / output interface unit 221 and the input / output interface unit 221. The client service module 220 generates a result value according to the request of the client unit 300, and sets parameters for each function. Set value for option items and deliver the result value of the current monitoring item.

The client unit 300 is used to input a condition in the query module 310 to extract and generate desired information from a database, and each function is processed for this data (Taskset generation), and a desired analysis item is generated. It selects the period based on the present, the desired period, and sets the search conditions for MS, Process, Equip, and Lot. By supporting the grouping function for each item above, select a function to be viewed, and set options for each item such as a map, chart, statistical analysis, report, and table of the result processor 320.

The result data module 350 receives the result data from the server unit 200 and processes the result data. The output data module 350 supports various outputs from the flatness data analyzer 361 and the particle data analyzer 362. Analyze for

FIG. 6 is an exemplary view of a map according to the present invention. In the flatness analysis, the site, surface, and cross-section map functions are provided (one or all can be selected), and each Single, Composite, Gallery Map Display for each map, statistics for each site (cout, Min, Max, Avg, Std Dev, etc.), selection and display, statistical processing Each map display by value, 2D / 3D surface map, cross-section in user's desired direction, data table, chart and basic information can be displayed simultaneously.

Particle Analysis provides Dark Field Wide, Narrow, Composite Particle Map, Haze Map support and single or gallery map functionality. It supports the display of maps (circular arcs, angles, etc.), basic information about the map (particle count, BIN, LPD, LPD_N, etc.) and image display when charts and images are stored.

7 is an exemplary view of the chart according to the present invention, the flatness analysis, provides a trend, histogram, Pareto, Box-Plot chart, etc., each date / equipment / Lot / wafer Yield Trend Chart (Yield: Accepted Wafer / Input Wafer) per MS / MS, various charts based on inspection results for each equipment, single, multi, gallery chart, etc. Process control charts such as upper / lower limit, Cp / Cpk calculation, and significant difference analysis by lot by equipment.

Particle analysis includes yield trend charts for each date / equipment / lot / wafer / MS (Yield = Accept wafers / input wafers), particle size, charts by class, and zonal distribution for circles and angles. , Multiplayer, gallery charts, and more.

8 is an exemplary diagram of a table and statistical analysis (SPC) according to the present invention.

The table may display data according to a search result of the query module in a table form and group various fields in combination. In addition, the table can be output as an Excel file.

Statistical analysis (SPC) displays the flatness / particle statistics according to the search results of the query module, displays the number, minimum, maximum, average, Std Dev, etc., and yields versus equipment or each as a correlation analysis. Correlation and correlation coefficients are extracted and provided for the parameters to be managed.

9 is an exemplary view of a report according to the present invention. The report provides automatic generation of a standardized report (daily, weekly, monthly), report preparation and form registration according to a user's needs, and a report designer. It enables to edit, output screen of each function to excel, and output table result to excel file to use as report data.

By utilizing the present invention as described above, all additional data generated during the process and data for process control can be collected and integrated to provide a basis for managing the entire process at a glance.

In addition, according to the present invention, it is possible to analyze roughness / flatness, defect images, etc. through an analysis function, and roughness and flatness are displayed as 2D / 3D maps and cross-section charts. You can view and compare peak-to-valley data by equipment / process and obtain statistical data for quality control such as real-time fault monitoring and alarms.

In addition, defect maps / charts fmf are displayed by period, size, and equipment, overlay of defect maps by equipment, and comparison of defect trends by location and defect trends by comparison of defect distribution by equipment. Circle / angle, etc.) and real-time defect monitoring / alarm, etc., and statistics for quality control.

The present invention can reduce the effort of the user due to the conventional work method in collecting data generated from various equipment for various failure analysis and yield improvement activities, and to focus on the original analysis activities or quality improvement activities. In addition, it not only stores the measured results of each device, but also enables systematic data management through an integrated database. In addition, it is effective to provide a user-oriented program that implements algorithms for calculating the thickness, flatness, and shape data necessary for actual analysis using RAW data generated from various equipment. have.

Claims (3)

In order to manage and improve the yield of bare wafers in the semiconductor wafer manufacturing process, the system analyzes the cause of defects and provides various statistical analysis. A loader part (100) connected to various measuring devices (10) for semiconductor wafer defect inspection to collect the measured data from the measuring devices, convert the data into a standard data format, and upload the same; Receives the data uploaded from the loader unit 100, connected to the TCP / IP shared network, parses the data, summarizes the information into a DB, and stores the data necessary to draw a map as a file, measuring equipment and measurement related data. And a server part (200) which obtains wafer information from the MES server and stores it in its own DB, and summarizes the information necessary for statistical analysis (SPC) and makes it into a database. Wafer, characterized in that configured to include a client part (300) for processing the necessary analysis work connected to the server unit 200 and outputs the results through the map, statistical analysis and charts and reports, etc. Bare wafer yield management system with map and statistical analysis. The method of claim 1, wherein the server unit 200, A raw data file input unit 211 for inspecting the format of the data file uploaded from the loader unit 100 and parsing the raw data file according to a predetermined rule for each device to receive the raw data file input unit 211; The data input from the input unit 211 is stored in the database, and the detailed information about the wafer is obtained from the MES server unit 400 using the message handling module 240 and the MES interface module 230 and stored in the DAS database. DB loader module 210 and A client service module 220 for interfacing through the client unit 300 and the input / output interface unit 221 to process data and reference information management requested by the client; An MES interface unit 230 for interfacing with the MES server unit 400 to interface with wafer information and data for statistical analysis to obtain necessary information from the server unit;  A message transmitted from the ECS managing the measurement equipment by interfacing with the message server 420 of the MES server unit 400 is received and processed through the message server 420, and the wafer is transmitted every time a single wafer is measured. A message handling module 240 that manages related information (Lot ID, Slot, Filename, etc.); A data summary module 212 that summarizes and accumulates DB accumulated for a certain period for statistical processing; A system monitoring module 213 for monitoring the entire system and notifying a user before a shortage occurs; Bare wafer yield management system providing a wafer map and statistical analysis, characterized in that comprises a database for storing and managing a variety of information. The method of claim 1, wherein the client unit 300, A query module 310 which enables a user to input various conditions for extracting data to be analyzed, creates a SQL query with the input conditions, and requests information from the server unit 200; A flatness data analyzer 361 analyzing the flatness data among data received by the request of the query module 310; A particle data analyzer 362 for analyzing particle data among data received by the request of the query module 310; A management module 340 for managing user rights, inputting various limit values, scaling, and various settings for each user;  A result data module 350 for storing and managing result data received from the server unit 200 and analysis data of each analysis unit; As a result, the result processing unit processing an analysis for estimating a failure cause and managing yield using maps, charts, statistical analysis (SPC) and reports using the result data received from the server unit stored in the data module 350 and the analyzed result data. 320, As a result, the bare wafer yield management system providing a wafer map and statistical analysis, characterized in that it comprises a display module 351 for outputting the processing result of the processing unit 320 to the screen.

Images