This application claims benefit of priority under 35 USC 119 based on Japanese Patent Application No. P2004-257778 filed Sep. 3, 2004, and Japanese Patent Application No. P2005-217030 filed Jul. 27, 2005, the entire contents of which are incorporated by reference herein.
1. Field of the Invention
The present invention relates to a process-state management system configured to manage process-states of a plurality of production machines, which are adapted for a sequence of manufacturing processes configured to manufacture products, a management server and a control server adapted for the system, a method for managing the process-states, a method for manufacturing products using the method for managing the process-states, and a computer program product storing a program being executed on the management server so as to implement the method for managing the process-states.
2. Description of the Related Art
Semiconductor devices are fabricated using many expensive semiconductor production machines through a long sequence of manufacturing processes constituted by a complex combination of manufacturing processes such as, for example, lithography process, etching process, thermal treatment (oxidation, annealing, diffusion) process, ion implantation process, thin-film formation (CVD, sputtering, deposition) process, cleaning (removal of resist, cleaning using solution) process, and inspection process. Recently, large wafers such as a 300-mm diameter wafer have been particularly used. This causes increase in unit cost of wafers and increase in fabrication cost when fault occurs during a long sequence of manufacturing processes.
Therefore, for fabrication of semiconductor devices using many expensive semiconductor production machines, the effective operation rate of production machines inside and outside of a factory has been improved, and the performance thereof has been maintained or improved using an equipment engineering system (EES). “EES” denotes a computer system configured to acquire pieces of equipment information inherent in respective semiconductor production machines, statistically analyze data in the equipment information, and determine whether the respective semiconductor production machines operate normally or abnormally.
More specifically, a highly intelligent computer system has been constructed for managing a precise and correct operation and maintenance of individual production machines using advanced process control (APC) or a fault detection and classification (FDC) system. “APC” denotes a computer system configured to change a semiconductor fabrication process according to how a semiconductor production machine has processed wafers, more specifically, control a processing recipe by controlling a process number as a variable using multivariate model prediction and then providing feedback-control and feedforward-control in a process or between processes. Automated complex processing for semiconductor mass production allows APC technology to reduce production cost, improve production efficiency, consistently maintain high quality, and modify an arbitrary portion in a fabrication process in real time. It is expected that this technology achieves an improvement in profit per wafer and reduction in fabrication cost. In addition, “FDC” denotes a system configured to always monitor operation-states of semiconductor production machines and shut down various complex tools (semiconductor production machines) used for wafer fabrication before deviation in performance of a semiconductor production machine may adversely influence product yields. This achieves reduction in risk to wafers.
As described above, in the semiconductor device fabrication field, various computer technologies for reducing variation in performance of semiconductor production machines with the passage of time, for minimizing difference in film deposition among wafers, and for canceling difference in performance among semiconductor production machines, thereby providing a stable fabrication process have been developed. For example, a system configured to transfer data resulting from inspecting wafers during processing and after processing, record and manage a processed wafer log and a log of semiconductor production machines, carry out self-testing of each semiconductor production machines, and transmit appropriate commands has been proposed in the International Publication WO96/25760 (see line 25 on page 36 through line 2 on page 37.)
Meanwhile, with the earlier FDC systems, when monitoring many semiconductor production machines, data acquisition units (adapters) different for respective machine venders or for respective FDC box manufacturers are used to acquire independent pieces of information of respective semiconductor production machines and then carry out separate fault detection for them using computers (dedicated servers) different for the respective machine benders or for the respective FDC box manufacturers. Therefore, pieces of data regarding the earlier FDC are dispersed for the respective semiconductor production machines.
In other words, according to the earlier FDC architecture, fault detection methods and automatic fault analysis methods are different for the respective machine benders or for the respective FDC box manufacturers, and software programs (applications) used for automatic fault analysis are different for respective semiconductor production machines. Therefore, automatically analyzing applications for analyzing abnormalities of production machines are different from one another for respective semiconductor production machines, requiring additional investment.
An aspect of the present invention inheres in a process-state management system encompassing: a plurality of production machines; a control server configured to collectively control at least part of the production machines; a management server including a data-linking module configured to link operation-management data of the production machines with corresponding management information transmitted from the control server, respectively, the management server analyze the operation-management data linked with the management information with a common analysis application; and a management database configured to store the operation-management data linked with the management information.
Another aspect of the present invention inheres in a management server adapted for a system encompassing a plurality of production machines, a control server configured to control collectively at least part of the production machines, a management server connected to the production machines and the control server through a communication network, and a management database connected to the management server, the management server encompassing: a data-linking module configured to link operation-management data of the production machines with corresponding management information transmitted from the control server, respectively. Here, the management server analyzes the operation-management data linked with the management information with a common analysis application.
Still another aspect of the present invention inheres in a control server adapted for a system encompassing a plurality of production machines, the control server configured to control collectively at least part of the production machines, a management server connected to the production machines and the control server through a communication network, and a management database connected to the management server, the control server encompassing: a data acquisition unit configure to acquire operation-management data from at least one of the production machine; and a data-linking module configured to link the operation-management data of the production machine with corresponding management information.
Yet still another aspect of the present invention inheres in a method for managing process-states by repetition of a sequence of procedures, the sequence of procedures encompassing: transmitting management information for the process executed by a subject production machine to a data-linking module from a control server; linking operation-management data of the subject production machine with the management information; and analyzing the operation-management data linked with the management information with a common analysis application by a management server. Here, the sequence of procedures is applied sequentially to a plurality of production machines so as to manage process-states executed by the production machines, respectively, using the analyzed results by the management server.
Further aspect of the present invention inheres in a method for managing process-states encompassing: transmitting a plurality of pieces of management information for processes executed by a plurality of production machines to a data-linking module from a control server; linking operation-management data of the production machines with corresponding management information, respectively; and analyzing the operation-management data linked with the management information with a common analysis application by a management server. Then process-states executed by the production machines are managed using the analyzed results by the management server.
Further aspect of the present invention inheres in a method for manufacturing a product encompassing: starting a subject process, which is one of the manufacturing processes in a sequence of manufacturing processes configured to manufacture the product, by using a subject production machine, so as to provide a subject intermediate product; linking operation-management data of the subject production machine with a piece of management information of the subject production machine; analyzing the operation-management data of the subject production machine linked with the management information of the subject production machine, and if the analyzed result satisfies a criterion, conveying the subject intermediate product to a next process of the subject process, and further proceeding to an influenced process assigned at a later stage than the subject process in the sequence of manufacturing processes; starting the influenced process by using an influenced production machine, so as to provide a influenced intermediate product; linking operation-management data of the influenced production machine with a piece of management information of the influenced production machine; and analyzing the operation-management data of the influenced production machine linked with the management information of the influenced production machine, and if the analyzed result satisfies a criterion, conveying the influenced intermediate product to another process next to the influenced process in the sequence of manufacturing processes.
Further aspect of the present invention inheres in a computer program product storing a program being executed on a management server in a system encompassing a plurality of production machines, a control server configured to control collectively at least part of the production machines, the management server connected to the production machines and the control server through a communication network, and a management database connected to the management server, the program encompassing a sequence of instructions including, instructions configured to transmit management information for the process executed by a subject production machine to a data-linking module from a control server; instructions configured to link operation-management data of the subject production machine with the management information; and instructions configured to analyze the operation-management data linked with the management information with a common analysis application by a management server. Here, the sequence of instructions is applied sequentially to a plurality of production machines so as to manage process-states executed by the production machines, respectively, using the analyzed results by the management server.
Various embodiments of the present invention will be described with reference to the accompanying drawings. It is to be noted that the same or similar reference numerals are applied to the same or similar parts and elements throughout the drawings, and the description of the same or similar parts and elements will be omitted or simplified. Generally and as it is conventional in the representation of semiconductor devices, it will be appreciated that the various drawings are not drawn to scale from one figure to another nor inside a given figure, and in particular that the layer thicknesses are arbitrarily drawn for facilitating the reading of the drawings.
FIG. 1 is a schematic diagram describing a logical configuration of a process-state management system according to a first embodiment of the present invention;
FIG. 2 is a flowchart describing an outline of a process-state management method according to the first embodiment of the present invention;
FIG. 3A shows exemplary management information for a photolithography process used by a process-state management system according to a second embodiment;
FIG. 3B shows exemplary management information for a CVD (deposition) process used by the process-state management system according to the second embodiment;
FIG. 3C shows exemplary management information for a spike-annealing process used by the process-state management system according to the second embodiment;
FIG. 4 is a flowchart describing a scenario of linking operation-management data (equipment data) with lot management information (product information) according to the process-state management method of the first embodiment of the present invention;
FIG. 5 shows an exemplary data structure of linking the operation-management data (equipment data) with the lot management information (product information) using the process-state management method according to the first embodiment of the present invention;
FIG. 6 is a schematic diagram describing a logical configuration of a process-state management system according to a modification of the first embodiment of the present invention;
FIG. 7 shows a schematic system structure describing a logical configuration of the process-state management system according to the second embodiment of the present invention;
FIG. 8 is a flowchart describing an outline of a process-state management method according to the second embodiment of the present invention;
FIG. 9A is a table of operation-management data (equipment data) transformed into a common format for a photolithography process;
FIG. 9B is a table showing related pretreatment (characteristic value);
FIG. 10A is a table of operation-management data (equipment data) transformed into the common format for a CVD (deposition) process;
FIG. 10B is a table showing pretreatment (characteristic value) related to the table shown in FIG. 10A;
FIG. 11A is a table of operation-management data (equipment data) transformed into the common format for a spike-annealing process;
FIG. 11B is a table showing pretreatment (characteristic value) related to the table shown in FIG. 11A;
FIG. 12 is a schematic diagram showing an outline of a process-state management system according to a first modification of the second embodiment of the present invention;
FIG. 13 is a schematic diagram showing an outline of a process-state management system according to a second modification of the second embodiment of the present invention;
FIG. 14 is a schematic diagram showing an outline of a process-state management system according to a third modification of the second embodiment of the present invention;
FIG. 15 is a schematic diagram showing an outline of a process-state management system according to a third embodiment of the present invention;
FIG. 16 is a flowchart showing a part of a sequence of manufacturing processes according to a semiconductor device fabrication method to which is applied a process-state management method according to the third embodiment of the present invention;
FIG. 17 is a flowchart describing an outline of the process-state management method according to the third embodiment of the present invention;
FIG. 18 is a graph showing an analyzed relationship between a gas flow rate during an amorphous silicon CVD process and an RF power during an amorphous silicon etching process, which results from an exemplary multivariate analysis for the process-state management method according to the third embodiment of the present invention;
FIG. 19 is a flowchart describing an outline of a process-state management method according to a comparative example of the third embodiment of the present invention;
FIG. 20 is a flowchart describing an outline of the process-state management method according to the third embodiment of the present invention;
FIG. 21 is a schematic diagram describing a logical configuration of a process-state management system according to a fourth embodiment of the present invention;
FIG. 22 is a schematic diagram describing a logical configuration of a process-state management system according to a fifth embodiment of the present invention;
FIG. 23A shows a graph describing an example of multivariate analysis or inter-application analysis, for a case of using an algorithm according to Hotelling's T 2 statistics, when there is a linear relationship between a first variable and a second variable, according to the process-state management method of the fifth embodiment of the present invention;
FIG. 23B shows a graph describing an example of multivariate analysis or inter-application analysis, for a case of using k-neighborhood algorithm, when there is nonlinear relationship between the first variable and the second variable, according to the process-state management method of the fifth embodiment of the present invention;
FIG. 23C shows a graph describing an example of multivariate analysis or inter-application analysis, for a case of an OR algorithm, according to the process-state management method of the fifth embodiment of the present invention;
FIG. 23D shows a graph describing an example of multivariate analysis or inter-application analysis, for a case of an AND algorithm, according to the process-state management method of the fifth embodiment of the present invention;
FIG. 24 is a schematic diagram describing a logical configuration of a process-state management system according to a sixth embodiment of the present invention; and
FIG. 25 is a schematic diagram describing a logical configuration of a process-state management system according to another embodiment of the present invention.
In the following description specific details are set forth, such as specific materials, processes and equipment in order to provide a thorough understanding of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known manufacturing materials, processes and equipment are not set forth in detail in order not to unnecessarily obscure the present invention. Although, in the first through sixth embodiments, process-state management systems/methods are explained focusing to semiconductor device fabrication methods, needless to say, the present invention may be applied to miscellaneous fabrication methods for a variety of industrial products, such as liquid crystal displays (LCD), magnetic recording media, optical recording media, thin-film magnetic recording/reading heads, or superconductor devices.
As shown in FIG. 1, a process-state management system according to a first embodiment of the present invention encompasses a plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , required for fabricating products such as semiconductor devices, a control server 11 a configured to collectively control operations of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , a management server 14 a configured to receive operation-management data, or equipment engineering (EE) data including descriptions of operation statuses and machine parameters of respective production machines, so as to monitor corresponding operation states of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time, and a management database 15 configured to store the operation-management data (EE data) of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . Hereafter in this specification, the EE data may be referred to as “equipment data”. The control server 11 a may have the functionality of a manufacturing execution system (MES) server to constitute a group of factory management systems, which link an enterprise resource planning (ERP) package or a head office business system with a group of control systems that control production machines in a factory. Therefore, as shown in FIG. 1, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , the control server 11 a , and the management server 14 a are connected to each other via a communication network 19 such as MES local area network (LAN). A data analyzing personal computer (PC) 18 is further connected to the communication network (MES LAN) 19 .
Analyzed results and determination results provided by the management server 14 a are fed back to the control server 11 a via the communication network (MES LAN) 19 , and individual specific processing commands (job commands) are transmitted to the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . from the control server 11 a , which has the functionality of the MES server. Note that, needless to say, a plurality of control servers may be physically provided via the communication network (MES LAN) 19 instead of the single control server 11 a exemplified in FIG. 1.
The plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . may include, for example, various semiconductor production machines, such as an ion implanter, a diffusion furnace, a thermal oxidation furnace for forming a silicon oxide film (SiO 2 ), a chemical vapor deposition (CVD) furnace for depositing an SiO 2 film, a phosphosilicate glass (PSG) film, a borosilicate glass (BSG) film, a boro-phosphate-silicate glass (BPSG) film, a silicon nitride (Si 3 N 4 ) film, a polysilicon film or related films, an annealing furnace for annealing a PSG film, a BSG film, a BPSG film or related films so as to perform reflow processing (melting), an annealing furnace for densifying a CVD oxide film or related films, an annealing furnace for forming a silicide film or related films, a sputtering equipment or a vacuum evaporator for depositing a metallic interconnect layer, a plating equipment for forming a metallic interconnect layer through plating, a chemical mechanical polishing (CMP) machine for polishing the surface of a semiconductor substrate, a dry/wet etching equipment for etching the surface of a semiconductor substrate, a cleaning equipment for removing a resist film or cleaning the surface of a semiconductor substrate using an aqueous solution, a spin coating machine (spinner) for coating a resist film on the surface of a semiconductor substrate so as to perform photolithography, an exposure tool such as a stepper, a dicing machine for dicing a semiconductor wafer into a plurality of chips, and a bonding machine for connecting each of electrodes of a diced chip-shaped semiconductor device to respective pads on a lead frame. The production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . further include various inspection tools and measuring tools, such as an interference film thickness gauge, an ellipsometer, a contact type film thickness gauge, a microscope, or a resistance measuring tool. Furthermore, miscellaneous facilities, such as an ultrapure water system or a gas purifier may be included. In addition, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . may include both batch type machines and single-wafer type machines. Similarly, all of the production machines, which will be explained in the disclosures of first to sixth embodiments, may include both batch type machines and single-wafer type machines.
As shown in FIG. 1, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are connected to a plurality of data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . , which acquire pieces of operation-management data (equipment data) of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in conformity with their own data collection plans (DCPs) and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). In FIG. 1, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . are connected to the management server 14 a via the communication network (MES LAN) 19 . Alternatively, in addition to the MES LAN 19 , an EES LAN may be established to connect them to each other.
For example, if the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are assumed to be machines each having a chamber, such as vacuum processing equipment for forming a thin film, a diffusion furnace, and a thin-film deposition reactor, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . receive pieces of operation-management data (equipment data), such as temperatures at respective different points in the chamber, temperature of a susceptor disposed in the chamber, temperatures at respective different points on the chamber outer walls, pressure in the chamber, gas flow rate introduced in the chamber, and valve conductance (angle of rotation) for controlling gas flow rate, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). If the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are assumed to be plasma-process related machines having discharge electrodes, such as a dry etching equipment or an ion implanter, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . receive pieces of operation-management data (equipment data), such as information of RF impedance-matching positions, RF voltages (voltages of incident and reflected waves), and information of wafer positions in addition to the aforementioned various parameters for vacuum processing, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). In addition, if the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are a wet etching equipment, a spin coating machine, an exposure tool, and a bonding machine, which perform processing under atmospheric pressure, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . receive pieces of operation-management data (equipment data), such as information of processing times and information of wafer or chip positions, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically).
A central processing unit (CPU) of the management server 14 a includes a data-linking module 142 configured to link the operation-management data (equipment data), which is transmitted from the respective data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . , with the management information transmitted from the control server 11 a . The data-linking module 142 links the operation-management data (equipment data) including unique data regarding product lot information, for example, with the management information including unique data regarding product lot information. The operation-management data (equipment data) of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . linked with the management information are stored in the management database 15 .
“Management information” includes management information and product information of semiconductor substrates. Exemplary management information is shown in FIGS. 3A, 3B and 3 C. FIG. 3A is a table showing exemplary management information of the photolithography process. FIG. 3B is a table showing exemplary management information of the CVD (deposition) process. FIG. 3C is a table showing exemplary management information of the spike-annealing process. Management information includes product name, lot number, wafer number (in the case of a single wafer type machine), sampling frequency of operation-management data (equipment data), corresponding process name, recipe name for the process, name of production machines carrying out the process, job ID number for the production machines, and module name or chamber name of the production machines.
Upon reception of a lot processing command from the control server 11 a , the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . generate a processing command unit (Job) and notify the control server 11 a thereof. A job ID number is attached to the processing command unit by the control server 11 a , and the control server 11 a then carries out lot progress management based on the Job ID notified from the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . (in conformity with semiconductor equipment and materials international (SEMI) standards).
Upon reception of notification that the processing command unit has been generated from the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , the control server 11 a identifies product information, such as a product name, a step name, a lot number, and a wafer number included in the lot, based on the job ID number, and then transmits the product information to the management server 14 a at the same time that the notification is received. High speed SECS message service (HSMS) communication in SEMI equipment communications standard (SECS) defined by SEMI, simple object access protocol (SOAP), file transfer protocol (FMP) or the like may be used as a communication protocol to transmit product information to the management server 14 a , but the communication protocol is not limited thereto. Afterwards, the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . initiate lot processing in order, and then transmit respective pieces of operation-management data (equipment data) to the management server 14 a when a predetermined condition set to the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . is satisfied (e.g., when processing a wafer in a target lot starts.) Each of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . notifies the control server 11 a of information regarding which wafer in the lot having been processed by sending a wafer processing initiated report.
The CPU of the management server 14 a in the process-state management system, according to the first embodiment, further includes a plurality of fault detection and classification (FDC) application execution modules 143 j , 143 j+1 , 143 j+2 , . . . . Each of the FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . is used as a common analysis application to collectively analyze and monitor the operation states of respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , and issue an instruction to stop (shut down) a specific production machine in real time if deviation in performance of the specific production machine may adversely influence product yields, thereby reducing risk to wafers. Alternatively, any one of applications in the plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . may be elected as a common analysis application to collectively analyze and monitor the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . Alternatively, two or more of the applications in the plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . may be combined and used as a common analysis application to collectively analyze and monitor the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . The plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . correspond to respective FDC applications. Alternatively, they may be software programs or dedicated hardware. More specifically, FDC programs to instruct and control the plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . are stored in an application program memory 16 connected to the CPU of the management server 14 a.
As described above, according to the process-state management system of the first embodiment, since a common FDC application may collectively monitor the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time, a unified fault detecting method or a unified automatic fault analyzing method may be used for monitoring many production machines of different machine venders. In other words, even if many production machines of different machine venders constitute a fabrication line in a factory, fault detection or an automatic fault analyzing application is not needed for each production machines, resulting in omission of additional investment.
Furthermore, since the operation-management data (equipment data) for all production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . required for fault detection and automatic analysis is kept (stored) in the management database 15 , the operation-management data (equipment data) can be retrieved from the management database 15 at a high speed as needed. In addition, the FDC application used for the process-state management system, according to the first embodiment, may be freely replaced. Even with replacement with any kind of FDC application, it is used as a common analysis application to collectively analyze and monitor the operation states of respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . However, if deviation in performance of a production machine thereof may adversely influence product yields, an instruction to stop (shut down) the production machine may be issued.
The CPU of the management server 14 a in the process-state management system, according to the first embodiment, further includes a plurality of process status control (PSC) application execution modules 144 k , 144 k+1 , 144 k+2 , . . . , which execute applications for various models such as multivariable model prediction. Each of the PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . is used as a common analysis application to collectively analyze and control processing recipes for the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . and provide feedback-control and feedforward-control for intra-process (within a process) or inter-process (among processes) implementation, resulting in reduction in production cost, improvement of production efficiency, and real-time correction of arbitrary portions in fabrication processes. Alternatively, any one of applications in the plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . may be elected as a common analysis application to collectively analyze and control processing recipes for the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . Alternatively, two or more of the applications in the plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . may be combined and used as a common analysis application to collectively analyze and control processing recipes for the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . The plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . correspond to respective APC applications. They may be software programs or dedicated hardware. More specifically, a plurality of APC programs to instruct and control the plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . are stored in the application program memory 16 connected to the management server 14 a.
Furthermore, since operation-management data (equipment data) for all production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . required for controlling processing recipes are stored in the management database, the operation-management data may be retrieved from the management database at a high speed as needed. In addition, the PSC applications may be freely replaced. Even with replacement with any kind of PSC application, it is used as a common analysis application to collectively analyze and control processing recipes for the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . or the like and then provide feedback-control and feedforward-control for intra-process (within a process) or inter-process (among processes) implementation, resulting in reduction in production cost, improvement of production efficiency, and real-time correction of arbitrary portions in fabrication processes.
Moreover, operation-management data (equipment data) inherent in each production machines stored in the management database 15 may be statistically analyzed using an EES application and may be used to improve the effective operation rate of production machines inside and outside of a factory and maintain or improve the performances of the production machines. Furthermore, since such data may be used in a technology CAD (TCAD) or a yield management system (YMS), the final yield of a semiconductor device may be estimated in intermediate processes prior to completion of the final process.
Needless to say, the management server 14 a includes an input unit used for an operator to enter data and instructions, an output unit for outputting analyzed results, a display unit, and data memory configured to be stored with intermediate data required for analyzing each of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , which are omitted in FIG. 1. The input unit of the management server 14 a may be implemented by a keyboard, a mouse, a light pen, or a flexible disk unit. A process manager (factory manager) may use the input unit to specify input/output data and change an application to be used. In addition, the input unit is provided so as to facilitate entry of a model to be used for analysis and also entry of commands for executing or aborting operation. The output unit and the display unit may be implemented by a printer and a display, respectively. Alternatively, the display unit may display input/output data, analyzed results, abnormal/normal state, and analysis parameters to allow the factory manager to collectively monitor the operation states of the production machines.
A process-state management method, according to the first embodiment of the present invention, is described using flowcharts shown in FIGS. 2 and 4. Note that the process-state management method according to the first embodiment described below is represented by a flowchart regarding the i-th production machine 12 i of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . constituting a fabrication line in a factory. Similarly, flowcharts regarding the (i−1)-th production machine 12 i−1 , the (i+1)-th production machine 12 i+1 , the (i+2)-th production machine 12 i+2 , the (i+3)-th production machine 12 i+3 , . . . may be shown. In addition, the flowchart shown in FIG. 2 is a mere example, and needless to say, various process-state management methods including modifications thereof are available. If we assign the i-th production machine 12 i as a subject production machine, the (i+1)-th production machine 12 i+1 may be assigned as the second production machine, and the (i+2)-th production machine 12 i+2 may be assigned as the second production machine, respectively.
(a) In a step S 101 of FIG. 2, a lot processing command (job command) for the production machine 12 i is first transmitted from the control server 11 a (corresponding to a step S 151 of FIG. 4). In a step S 102 of FIG. 2, the production machine 12 i then generates a job (lot processing information) of the production machine 12 i , and transmits a lot processing information generation report (JobID) to the control server 11 a (corresponding to a step S 152 of FIG. 4).
(b) Once the lot processing information generation report (JobID) of the production machine 12 i is transmitted to the control server 11 a in the step S 102 of FIG. 2, the control server 11 a transmits lot management information (product information) PI for the processing executed by the production machine 12 i to the management server 14 a in a step S 103 of FIG. 2 (corresponding to a step S 153 of FIG. 4).
(c) On the other hand, in a step S 104 of FIG. 2, the production machine 12 i starts corresponding lot processing such as lithography, etching, thermal treatment, ion implantation, CVD, sputtering, deposition, and cleaning in conformity with a predetermined recipe (corresponding to a step S 154 of FIG. 4). The predetermined recipe is managed by the control server 11 a.
(d) As shown in a step S 156 of FIG. 4, once lot processing starts in the step S 104 of FIG. 2, “a first wafer processing start report (JobID, WaferID)” is transmitted to the control server 11 a , which then transmits the report to the management server 14 a . In addition, in a step S 105 of FIG. 2, the data acquisition unit 13 i connected to the production machine 12 i starts collecting operation-management data (equipment data) of the production machine 12 i in conformity with a data collection plan dedicated to the production machine 12 i and the collected data are temporarily stored in a storage unit of the data acquisition unit 13 i .
(e) Afterwards, in a step S 106 of FIG. 2 (corresponding to the step S 157 of FIG. 4), the data acquisition unit 13 i transmits the collected pieces of operation-management data (equipment data) ED (t i ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i to the data-linking module 142 in the CPU of the management server 14 a at predetermined timings (periodically). The data-linking module 142 in the CPU of the management server 14 a then links the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i with the lot management information (product information) PI transmitted from the control server 11 a in a step S 128 of FIG. 2. The data-linking module 142 identifies product information being processed from a job ID number and a wafer ID number in the wafer processing start report, which has been transmitted from the production machine 1 Z to the control server 11 a , and then stores in the management database 15 , corresponding lot management information (product information) PI linked with the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . . In other words, the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i linked with the lot management information (product information) by the data-linking module 142 are stored in the management database 15 as operation-management data. Alternatively, the data may be temporarily stored in main memory (data memory) or cache memory of the management server 14 instead of the management database 15 .
(f) On the other hand, the management server 14 a pre-selects the most appropriate FDC application execution module from a group of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . installed in the CPU of the management server 14 a as a common module for collectively monitoring the plurality of production machine 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time. According to the process-state management method of the first embodiment, it is assumed that the FDC application execution module 143 j is selected. In a step S 109 of FIG. 2, the FDC application execution module 143 j reads out the operation control data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i from the management database 15 , and then determines whether or not the production machine 12 i is operating normally. For example, if the production machine 12 i is a production machine having a chamber, such as a vacuum evaporator or sputtering equipment for depositing thin film, a diffusion furnace, or a thin-film deposition reactor, temperatures in a plurality of regions in the chamber (first machine parameter), temperature of a susceptor disposed in the chamber (second machine parameter), temperatures in a plurality of regions on the chamber outer walls (third machine parameter), pressure in the chamber (fourth machine parameter), gas flow rate flowing into the chamber (fifth machine parameter), and valve conductance for controlling the gas flow rate (sixth machine parameter) are compared with respective predetermined normal values.
(g) In a step S 109 of FIG. 2, if the FDC application execution module 143 j determines that all of the first to sixth machine parameters of the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i fall within the criterions of the respective predetermined values, processing proceeds to a step S 110 of FIG. 2 in which the (i+1)-th processing subsequent to the processing conducted by the i-th production machine 12 i is carried out. In other words, the control server 11 a transmits a lot processing command (job command) for the (i+1)-th processing to the (i+1)-th production machine 12 i+1 .
(h) On the other hand, in the step S 109 of FIG. 2, if the FDC application execution module 143 j determines that any one or more of the first to sixth machine parameters of the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i does not fall within the criterions of the respective predetermined values, processing proceeds to a step S 111 or S 113 of FIG. 2. In general, the transition timing from the step S 109 to the step S 111 differs from the transition timing from the step S 109 to the step S 113 . Although, determination of whether to proceed to the step S 111 or S 113 in FIG. 2 depends on a criterion such as the degree of deviations in the first to sixth machine parameters from corresponding respective reference values or a criterion of whether the first to sixth machine parameters are abnormal, the transition from the step S 109 to the step S 113 is a real-time process so as to suspend (shut down) the operation of the production machine 12 i when the abnormality is detected. On the contrary, the transition timing from the step S 109 to the step S 111 is performed after the timing when the subject lot process is completed, establishing instruction of quality control (QC) or measurement for QC by the instruction. Namely, in a step S 111 of FIG. 2, measurement for QC is instructed and then conducted. “Measurement for QC”, in the case of a CVD furnace, means measurement of thickness of a thin film deposited by the CVD furnace. If the measurement for QC in the step S 111 of FIG. 2 reveals it to fall within a predetermined range, processing proceeds to a step S 110 of FIG. 2 in which a lot processing command is then transmitted to the (i+1)-th production machine 12 i+1 . Otherwise, if the measurement for QC in the step S 111 of FIG. 2 reveals it to be out of the predetermined range, processing proceeds to the step S 113 of FIG. 2 in which the control server 11 a transmits an instruction to suspend (shut down) the operation of the production machine 12 i . In addition, necessary incidental processing such as notifying the abnormality to a manager of the production machine 12 i is carried out at the same time. As described above, the production machine 12 i is always monitored, and is shut down before the deviation of the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machine 12 i from reference values may adversely influence product yields, thereby reducing risk to wafers.
Similarly, the common FDC application execution module 143 j may always collectively monitor the other production machines 12 i+1 , 12 i+2 , 12 i+3 , . . . constituting a fabrication line in a factory, and suspend (shut down) a specific production machine thereof before deviation of the pieces of operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the production machines 12 i+1 , 12 i+2 , 12 i+3 , . . . from corresponding respective reference values may adversely influence product yields, thereby reducing risk to wafers.
FIG. 5 shows an exemplary data structure with the pieces of operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . linked with respective pieces of lot management information (product information) PI. A method of storing data in the management database 15 has been described where the data includes the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . linked with the lot management information (product information) PI. However, the present invention is not limited to that method. Alternatively, a data file may be stored. A user can retrieve the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . linked with the lot management information by accessing the management server 14 a via a data analysis terminal. In the case where lot numbers and recipe names are particularly repetitive among product types in a large-item small-volume production line, data analysis of a limited number of product types may be easily carried out, resulting in improved efficiency and improved accuracy of analysis. In addition, in a case of semiconductor equipment production machines (semiconductor equipment), for example, if the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . conform to communications standard for semiconductor equipment such as SECS defined by SEMI, communication software programs for the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are not required to be modified, resulting in prevention of increase in undesired cost. As such, according to the process-state management method of the first embodiment of the present invention, processing of the steps S 101 through S 113 in FIG. 2 may be applied to the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . one after another, and processes by the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . may be controlled based on results of analyzing using the common analysis application by the management server 14 a . Here, “management” includes operation state changes such as a forced suspension or a temporary suspension of the operation of a specific production machine, among the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , based on the analyzed results. In addition, provision of feedback-control and/or feedforward-control for intra-process (within a process) or inter-process (among processes) implementation is included. Note that with the process-state management system, the management server, the process-state management method, and the process-state management program according to the second to sixth embodiments of the present invention, the term “management” includes feedback-control and feedforward-control.
A program having an algorithm equivalent to the series of procedures according to the process-state management method shown in FIG. 2 may be executed to control the process-state management system in FIG. 1. The program should be stored in both program memories (not shown in the drawing) of the control server 11 a and the management server 14 a constituting the process-state management system, according to the first embodiment of the present invention. In addition, once the program is stored in a computer readable recording medium, and the content of the recording medium is read out to the program memories of the control server 11 a and the management server 14 a , the procedure according to the process-state management method of the first embodiment of the present invention may be carried out. A “computer readable recording medium” means a medium such as an external memory of a computer, semiconductor memory, a magnetic disk, an optical disk, a magnetic optical (MO) disk, or a magnetic tape, which may be stored with programs. More specifically, a “computer readable recording medium” may be a flexible disk, a compact disk (CD)-read-only memory (ROM), an MO disk, a cassette tape, or an open reel tape. For example, the main bodies of the control server 11 a and the management server 14 a may be structured so that a flexible disk drive and an optical disk drive are either incorporated or externally connected. Inserting a flexible disk into the flexible disk drive or a CD-ROM into the optical disk drive and conducting a predetermined read-out operation allows installation of programs stored in those recording media in the program memories of the control server 11 a and the management server 14 a . In addition, by connection of suitable drive units, ROM as a semiconductor memory, or a cassette tape as a magnetic tape unit can be employed. Furthermore, the program may be stored in the program memory via an information-processing network such as the Internet.
For example, a process-state management program required for controlling the management server 14 a includes:
(a) Instructions configured to cause the data-linking module 142 to command the control server 11 a to transmit lot management information (product information) PI for the processing executed by the specified production machine 12 i to the data-linking module 142 in the management server 14 a;
(b) Instructions configured to cause the data-linking module 142 to receive the lot management information (product information) PI for the processing executed by the specified production machine 12 i , which corresponds to the step S 103 of FIG. 2 (after the control server 11 a transmits the lot management information (product information) PI for the processing executed by the specified production machine 12 i to the data-lining module 142 in conformity with the aforementioned command);
(c) Instructions configured to cause the data-linking module 142 in the management server 14 a to receive the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . of the specified production machine 12 i , which corresponds to the step S 106 of FIG. 2, and cause the data-linking module 142 in the management server 14 a to link the operation-management data with the lot management information (product information) PI in the step S 128 of FIG. 2; and
(d) Instructions configured to cause the FDC application execution module 143 j in the management server 14 a to analyze the linked operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . using a common analysis application in the steps S 109 through S 113 of FIG. 2.
The series of instructions are applied one after another to a set of the lot management information (product information) PI and the operation-management data (equipment data) ED (t 1 ), ED (t 2 ), ED (t 3 ), . . . transmitted from the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . Analysis in the steps S 109 through S 113 of FIG. 2 is then conducted. The analyzed results allow the management server 14 a to control the processes carried out by the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . .
With nearly the same configuration as the configuration shown in FIG. 1, a process-state management system according to a modification of the first embodiment of the present invention shown in FIG. 6 encompasses a plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , a control server 11 b configured to collectively control operations of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , a management server 14 a configured to receive operation-management data (equipment data), which includes descriptions of operation statuses and machine parameters of the respective production machines, and monitor corresponding operation states of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time, and a management database 15 storing pieces of operation-management data (equipment data) of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . .
Note that according to the configuration shown in FIG. 1, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are connected to a plurality of data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . , which acquire pieces of operation-management data (equipment data) of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in conformity with their own DCPs and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). However, according to the process-state management system of the modification of the first embodiment of the present invention shown in FIG. 6, the control server 11 b includes a data acquisition unit 13 int , which receives pieces of operation-management data (equipment data) of the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in conformity with their own DCPs and then transmits them to the management server 14 a at predetermined timings (periodically).
With nearly the same configuration as the configuration shown in FIG. 1, the control server 11 b in the process-state management system according to the modification of the first embodiment of the present invention shown in FIG. 6 may have the functionality of a MES server to constitute a group of factory management systems, which link an ERP package or a head office business system with a group of control systems that control machines in a factory. Therefore, as shown in FIG. 6, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , the control server 11 b , and the management server 14 a are connected to each other via a communication network (MES LAN) 19 . A data analyzing personal computer (PC) 18 is further connected to the communication network (MES LAN) 19 .
Analyzed results and determination results provided by the management server 14 a are fed back to the control server 11 b via the communication network (MES LAN) 19 , and individual specific processing commands Gob commands) are transmitted to the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . from the control server 11 b , which has the functionality of the MES server. Note that, needless to say, a plurality of control servers may be physically provided via the communication network (MES LAN) 19 instead of the single control server 11 b exemplified in FIG. 6, as described in FIG. 1.
For example, if the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are assumed to be machines each having a chamber, such as vacuum processing equipment for forming a thin film, a diffusion furnace, and a thin-film deposition reactor, the data acquisition unit 13 int receives pieces of operation-management data (equipment data), such as temperatures at respective different points in the chamber, temperature of a susceptor disposed in the chamber, temperatures at respective different points on the chamber outer walls, pressure in the chamber, gas flow rate introduced in the chamber, and valve conductance (angle of rotation) for controlling gas flow rate, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). If the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are assumed to be plasma-process related machines having discharge electrodes, such as a dry etching equipment or an ion implanter, the data acquisition unit 13 int receives pieces of operation-management data (equipment data), such as information of RF impedance-matching positions, RF voltages (voltages of incident and reflected waves), and information of wafer positions in addition to the aforementioned various parameters for vacuum processing, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). In addition, if the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are a wet etching equipment, a spin coating machine, an exposure tool, and a bonding machine, which perform processing under atmospheric pressure, the data acquisition unit 13 int receives pieces of operation-management data (equipment data), such as information of processing times and information of wafer or chip positions, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically).
A CPU of the management server 14 a includes a data-linking module 142 configured to link the operation-management data (equipment data), which is transmitted from the data acquisition unit 13 int , with the management information transmitted from the control server 11 b , which is nearly the same configuration as the configuration, shown in FIG. 1. The data-linking module 142 links the operation-management data (equipment data) including unique data regarding product lot information, for example, with the management information including unique data regarding product lot information. The operation-management data (equipment data) of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . linked with the management information are stored in the management database 15 .
Other functions, configurations, way of operation are substantially similar to the functions, configurations, way of operation already explained in the first embodiment with FIG. 1, overlapping or redundant description may be omitted.
As shown in FIG. 7, similar to the process-state management system according to the first embodiment, a process-state management system according to a second embodiment of the present invention encompasses a plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , required for fabricating products such as semiconductor devices, a control server 11 a configured to collectively control operations of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , a management server 14 b configured to receive operation-management data (equipment data), which includes descriptions of operation statuses and machine parameters of the respective production machines, and monitor corresponding operation states of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time, and a management database 15 storing pieces of operation-management data (equipment data) of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . The production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are connected to a plurality of data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . , which acquire pieces of operation-management data (equipment data) of the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , in conformity with their own DCPs and transmit the operation-management data to the management server 14 b at predetermined timings (periodically).
However, as shown in FIG. 7, the process-state management system according to the second embodiment is different from the process-state management system according to the first embodiment shown in FIG. 1 in that the former includes a CPU of the management server 14 b , which includes a format transformer 141 configured to transform the operation-management data transmitted from the respective data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . into data represented by a common format (common-format-transformed operation-management data) as shown in FIGS. 9A, 9B, 10 A and 10 B.
As with the process-state management system according to the first embodiment, the CPU of the management server 14 b further includes a data-linking module 142 configured to link the operation-management data transformed into the common format by the format transformer 141 with the management information transmitted from the control server 11 a . The data-linking module 142 links the operation-management data including unique data regarding product lot information, for example, with the management information including unique data regarding product lot information. The operation-management data (equipment data) of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . linked with the management information are stored in the management database 15 with a common format.
Not shown in the drawing, as with FIG. 1, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . are connected to the management server 14 a via a MES LAN. Alternatively, in addition to the MES LAN, an EES LAN may be established to connect them to each other. In addition, the control server 11 a may have the functionality of a MES server to constitute a group of factory managing systems, which link an ERP package or a head office business system with a group of control systems that control machines in a factory, as described in the first embodiment. Therefore, although not shown in the drawing, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are mutually connected to the control server 11 a via the MES LAN. Furthermore, although wiring is not shown in FIG. 7, analyzed results and determination results provided by the management server 14 b are fed back to the control server 11 a , and individual specific processing commands (job commands) are transmitted to the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . from the control server 11 a , which has the functionality of the MES server. Note that a plurality of control servers may be physically provided instead of the single control server 11 a exemplified in FIG. 7, as with the process-state management system according to the first embodiment.
The plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . may include, for example, various semiconductor production machines, such as an ion implanter, a diffusion furnace, a thermal oxidation furnace for forming a silicon oxide film (SiO 2 ), a CVD furnace for depositing an SiO 2 film, a PSG film, a BSG film, a BPSG film, a silicon nitride (Si3N4) film, a polysilicon film or related films, an annealing furnace for depositing a PSG film, a BSG film, a BPSG film or related films through reflow processing (melting), an annealing furnace for densifying a CVD oxide film or related films, an annealing furnace for forming a silicide film or related films, a sputtering equipment or a vacuum evaporator for depositing a metallic interconnect layer, a plating equipment for forming a metallic interconnect layer through plating, a CMP machine for polishing the surface of a semiconductor substrate, a dry/wet etching equipment for etching the surface of a semiconductor substrate, a cleaning equipment for removing resist film or cleaning the surface of a semiconductor substrate using an aqueous solution, a spin coating machine (spinner) for coating a resist film on the surface of a semiconductor substrate so as to implement photolithography, an exposure tool such as a stepper, a dicing machine for dicing a semiconductor into a plurality of semiconductor chips, and a bonding machine for connecting electrodes of a diced chip-shaped semiconductor device to corresponding pads on a lead frame. The production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . further include various inspection tools and measuring tools, such as an interference film thickness gauge, an ellipsometer, a contact type film thickness gauge, a microscope, or a resistance measuring tool. Furthermore, incidental facilities, such as an ultrapure water system or a gas purifier may be included. In addition, the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . may include both batch type machines and single wafer type machine. Similarly, the batch type machine or the single wafer type machine may be applied to all of the production machines in the embodiments described later.
For example, if the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are assumed to be machines each having a chamber, such as vacuum processing equipment for forming a thin film, a diffusion furnace, and a thin-film deposition reactor, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . receive pieces of operation-management data (equipment data), such as temperatures at respective different points in the chamber, temperature of a susceptor disposed in the chamber, temperatures at respective different points on the chamber outer walls, pressure in the chamber, gas flow rate introduced in the chamber, and valve conductance (angle of rotation) for controlling gas flow rate, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). If the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are assumed to be plasma-process related machines having discharge electrodes, such as a dry etching equipment or an ion implanter, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . receive pieces of operation-management data (equipment data), such as information of RF impedance-matching positions, RF voltages (voltages of incident and reflected waves), and information of wafer positions in addition to the aforementioned various parameters for vacuum processing, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically). In addition, if the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . are a wet etching equipment, a spin coating machine, an exposure tool, and a bonding machine, which perform processing under atmospheric pressure, the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . receive pieces of operation-management data (equipment data), such as information of processing times and information of wafer or chip positions, and then transmit the operation-management data to the management server 14 a at predetermined timings (periodically).
FIGS. 9A, 10A, and 11 A are tables representing operation-management data (equipment data) with a common format for three different processes. FIG. 9A is a table representing common format data for a photolithography process. FIG. 10A is a table representing common format data for a CVD (deposition) process. FIG. 11A is a table representing common format data for a spike-annealing process. In FIGS. 9A, 10A, and 11 A, a set of “records”, data for each of records are aligned along horizontal rows, represent respective times when the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . have received data. FIG. 9A is a table representing the operation-management data (equipment data) for the photolithography process with item names such as “autofocus” as a machine parameter A, “monitored exposure value” as machine parameter B, “baking-stage temperature” as machine parameter C, “shot number” as machine parameter D, “recipe step” as machine parameter E, “time in step” as machine parameter F, . . . . Data for each of machine parameters A-F are recorded along vertical columns of FIG. 9A as “field”. Then, each of records represents a collection of information about a separate field, and a collection of records that contain the same set of fields defines the table shown in FIG. 9A.
FIG. 10A is a table representing the operation-management data (equipment data) for the CVD (deposition) process with item names such as “heater temperature” as machine parameter A, “low amount of gas A” as machine parameter B, “chamber pressure” as machine parameter C, “recipe step” as machine parameter D, . . . , which are recorded so as to constitute a set of fields.
Furthermore, FIG. 11A is a table representing the operation-management data (equipment data) for the spike-annealing process with item names such as “pyrometer value” as machine parameter A, “stage rotation” as machine parameter B, “lamp power” as machine parameter C, “flow amount of gas A” as machine parameter D, . . . , which are recorded so as to constitute a set of fields.
For the photolithography process of FIG. 9A, data including a mean (average value) of Shot 1 of the machine parameter C (or the baking-stage temperature) being 65 degrees Centigrade, a deviation of Shot 1 of the machine parameter C being 0.2 degrees Centigrade and the like are provided as pretreatments (characteristic values) as shown in FIG. 9B. For the CVD (deposition) process of FIG. 10A, data including a mean (average value) of recipe 3 of the machine parameter B (or the flow amount of gas A) being 12.5 sccm and a deviation of the recipe 3 of the machine parameter C (or the chamber pressure) being 13 Pa are provided as pretreatments (characteristic value) as shown in FIG. 10B. For the spike-annealing process of FIG. 11A, data including a mean (average value) of the steps 1 and 2 of the machine parameter C (lamp power) being 1100 degrees Centigrade and a deviation within one sec of the step 2 of the machine parameter C being 3 degrees Centigrade are provided as pretreatments (characteristic values) as shown in FIG. 11B.
As shown in FIGS. 9A, 10A, and 11 A, since the operation-management data (equipment data) acquired by the data acquisition units are transformed into a common format as a set of records representing respective acquired times, the pieces of operation-management data (equipment data) of respective production machines used in different processes may be stored in the relational database with the acquired times when the data acquisition units have acquired as a main key. FIGS. 9A, 10A, and 11 A show exemplary table formats. Alternatively, various table formats are available as long as they are common to all production machines.
The CPU of the management server 14 b in the process-state management system, according to the second embodiment, further includes a plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . . Each of the FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . is used as a common analysis application to collectively analyze and monitor the operation states of respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , and issue an instruction to shut down a specific production machine in real time if deviation in performance of the specific production machine may adversely influence product yields, thereby reducing risk to wafers. Alternatively, any one of applications in the plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . may be elected as a common analysis application to collectively analyze and monitor the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . Alternatively, two or more of applications in the plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . may be combined and used as a common analysis application to collectively analyze and monitor the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . The plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . may be implemented by software programs or dedicated hardware such as logic circuits. More specifically, FDC programs to instruct and control the plurality of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . are stored in an application program memory 16 connected to the CPU of the management server 14 b.
As described above, according to the process-state management system of the second embodiment, since a common FDC application may collectively monitor the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time, a unified fault detecting method or a unified automatic fault analyzing method may be used for monitoring many production machines of different machine venders.
In other words, even if many production machines of different machine venders constitute a fabrication line in a factory, a fault detecting or an automatic fault analyzing application is not needed for each production machines, resulting in omission of additional investment.
Furthermore, since the operation-management data for all production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . required for abnormality detection and automatic analysis is kept (stored) in a management database 15 with a common format independent of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . and the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . connected thereto, the operation-management data for all production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . can be retrieved from the management database 15 at a high speed as needed. In addition, the FDC applications used for the process-state management system, according to the first embodiment, may be freely replaced. Even with replacement with any kind of FDC application, it is used as a common analysis application to collectively analyze and monitor the operation states of respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . However, if deviation in performance of a specific production machine thereof may adversely influence product yields, an instruction to stop (shut down) the specific production machine may be issued.
The CPU of the management server 14 b in the process-state management system, according to the first embodiment, further includes a plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . , which execute applications for various models such as multivariable model prediction. Each of the PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . is used as a common analysis application to collectively analyze and control processing recipes for the respective production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . and provide feedback-control and feedforward-control for intra-process (within a process) or inter-process (among processes) implementation, resulting in reduction in production cost, improvement of production efficiency, and real-time correction of arbitrary portions in fabrication processes. Alternatively, an application in the plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . may be elected as a common analysis application to collectively analyze and control processing recipes for the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . Alternatively, two or more of applications in the plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . may be combined and used as a common analysis application to collectively analyze and control the processing recipes for the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . . The PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . may correspond to APC applications. They may be software programs, or dedicated hardware. More specifically, a plurality of APC programs to instruct and control the plurality of PSC application execution modules 144 k , 144 k+1 , 144 k+2 , . . . may be stored in the application program memory 16 connected to the management server 14 b.
Furthermore, since the operation-management data for all production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . required for controlling processing recipes are stored in the management database 15 with a common format independent of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . and the data acquisition units 13 i , 13 i+1 , 13 i+2 , 13 i+3 , . . . connected thereto, the common-format-transformed operation-management data may be retrieved from the management database 15 at a high speed as needed. In addition, the PSC applications may be freely replaced. Even with replacement with any kind of PSC application, it is used as a common analysis application to collectively analyze and control processing recipes for the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . or the like and then provide feedback-control and feedforward-control for intra-process (within a process) or inter-process (among processes) implementation, resulting in reduction in production cost, improvement of production efficiency, and real-time correction of arbitrary portions in fabrication processes.
Moreover, the operation-management data (equipment data) inherent in each production machines stored in the management database 15 with a common format may be statistically analyzed using an EES application and may be used to improve the effective operation rate of production machines inside and outside of a factory and maintain or improve the performances of the production machines. Furthermore, since such data may be used in TCAD or YMS, the final yield of a semiconductor device may be estimated in intermediate processing prior to completion of the final processing. Needless to say, the management server 14 b includes an input unit used for an operator to enter data and instructions, an output unit for outputting analyzed results, a display unit, and data memory configured to be stored with intermediate data required for analyzing each of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . , which are omitted in FIG. 7. The input unit of the management server 14 b may be implemented by a keyboard, a mouse, a light pen, or a flexible disk unit. A process manager (factory manager) may use the input unit to specify input/output data, and change an application to be used. In addition, the input unit allows entry of a model to be used for analysis, and also entry of commands for executing or aborting operation. The output unit and the display unit may be implemented by a printer and a display, respectively. Alternatively, the display unit may display input/output data, analyzed results, abnormal/normal state, and analysis parameters to allow the factory manager to collectively monitor the operation states of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . .
A process-state management method, according to the second embodiment of the present invention, is described using a flowchart shown in FIG. 8. Note that the process-state management method according to the second embodiment described below is represented by a flowchart regarding the i-th production machine 12 u of the production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . constituting a fabrication line in a factory. Similarly, flowcharts regarding the (i−1)-th production machine 12 i−1 , the (i+1)-th production machine 12 i+1 , the (i+2)-th production machine 12 i+2 , the (i+3)-th production machine 12 i+3 , . . . may be shown. In addition, the flowchart shown in FIG. 8 is a mere example, and needless to say, various process-state management methods including modifications thereof are available.
(a) In a step S 101 , a lot processing command (job command) for the production machine 12 i is first transmitted from the control server 11 a . In a step S 102 , the production machine 12 i then generates a job for the production machine 12 i and transmits a job generation report for the production machine 12 i to the control server 11 a.
(b) Once the job generation report for the production machine 12 I is transmitted to the control server 11 a in the step S 102 , the control server 11 a transmits lot management information for the processing executed by the production machine 12 i to the management server 14 b in a step S 103 .
(c) On the other hand, in a step S 104 , the production machine 12 i starts corresponding lot processing, such as lithography, etching, thermal treatment, ion implantation, CVD, sputtering, deposition, and cleaning in conformity with a predetermined recipe. The predetermined recipe is managed by the control server 11 a.
(d) Once lot processing starts in the step S 104 , the data acquisition unit 13 i connected to the production machine 12 i starts collecting the operation-management data (equipment data) of the production machine 12 i in conformity with DCP, which is characteristic to the production machine 12 i , and the corrected data are temporarily stored in a storage unit of the data acquisition unit 13 i in a step S 105 .
(e) Afterwards, the data acquisition unit 13 i transmits the collected operation-management data (equipment data) of the production machine 12 i to the format transformer 141 in the CPU of the management server 14 b at a predetermined tiling (in a step S 106 ). Once the operation-management data (equipment data) of the production machine 12 i is transmitted to the format transformer 141 in the step S 106 , the format transformer 141 transforms the format for the transmitted operation-management data (equipment data) of the production machine 12 i into a common format for all production machines as shown in FIGS. 9A, 9B, 10 A, 10 B, 11 A and 11 B in a step S 107 .
(f) The data-linking module 142 in the CPU of the management server 14 b then links the format-transformed operation-management data (equipment data) of the production machine 12 i with the lot management information transmitted from the control server 11 a in a step S 108 . The operation-management data (equipment data) of the production machine 12 i linked with the lot management information by the data-linking module 142 are stored in the management database 15 as operation-management data with the common format. Alternatively, the data may be temporarily stored in main memory (data memory) or cache memory of the management server 14 b instead of the management database 15 .
(g) On the other hand, the management server 14 b pre-selects the most appropriate FDC application execution module from a group of FDC application execution modules 143 j , 143 j+1 , 143 j+2 , . . . installed in the CPU of the management server 14 b as a common module for collectively monitoring the plurality of production machines 12 i , 12 i+1 , 12 i+2 , 12 i+3 , . . . in real time. According to the process-state management method of the second embodiment, it is assumed that the FDC application execution module 143 j is selected. In a step S 109 , the FDC application execution module 143 j reads out the operation-management data (equipment data) of the production machine 12 i from the management database 15 , and then determines whether or not the production machine 12 i is operating normally. For example, if the production machine 12 i is a production machine having a chamber, such as a vacuum evaporator or sputtering equipment for depositing thin film, a diffusion furnace, or a thin-film deposition reactor, temperatures in a plurality of regions in the chamber (first machine parameter), temperature of a susceptor disposed in the chamber (second machine parameter), temperatures in a plurality of regions on the chamber outer walls (third machine parameter), pressure in the chamber (fourth machine parameter), gas flow rate flowing into the chamber (fifth machine paramete