DE10203997B4 - Method and system for collecting production data in a production plant - Google Patents

Abstract

Method for recording production data in a production plant with several sub-production plants, • in which sub-production data are determined and stored from the sub-production plants, • in which the sub-production data is assigned to groups of the grouping table using at least one grouping table Production data is mapped to compressed partial production data by using a statistical analysis to generate compressed key figures for partial production data grouped according to the assignment of the partial production data to groups in the grouping table, • where in the grouping table • machines of the partial production systems to at least one machine group are grouped, or • machine components of the sub-production systems are grouped into at least one machine component group, or • production recipes are grouped into at least one recipe group, or • production parameters are grouped into at least one parameter ergroup are grouped.

Description

The invention relates to a method and a system for recording production data in a production plant, in particular in a chip production plant.

In the production of highly integrated semiconductor chips, the ever-increasing miniaturization of the structures on the semiconductor chip will result in ever greater demands on the production systems and production processes used for the production of the semiconductor chips. The stability and reproducibility of both the production equipment and the manufacturing processes significantly influence the yield and productivity in semiconductor chip manufacturing. Even small deviations from a desired behavior of a chip manufacturing plant in the context of production can lead to a considerable deterioration of the yield, that is to say to a considerable increase in the error rate in the semiconductor chips produced.

Thus, an essential aspect in the production of semiconductor chips, possible deviations from a desired behavior in a chip manufacturing plant or as part of a manufacturing process to recognize very early and take appropriate countermeasures. Thus, the analysis and monitoring of machines, in particular the chip manufacturing plants and the manufacturing processes to a very high economic importance. Furthermore, the analysis and monitoring of many process steps of the manufacturing process is of considerable importance, since usually a repair of an intermediate product after a process step is rarely possible. In general, a functional test of a manufactured semiconductor chip is provided only at the very end of the manufacturing process, which leads to a very late feedback of the determined results in the manufacturing process.

It is furthermore known to provide inline measurements of intermediate process results, for example the layer thicknesses, the sheet resistance or of line widths, etc. by means of the so-called Statistical Process Control (SPC). However, this leads to additional measuring steps in the context of the entire manufacturing process and is therefore time-consuming and costly.

As part of the so-called Advanced Process Control (APC) data from internal and external sensors of the manufacturing equipment in combination with measuring technology used including in-line measured data on intermediates, or results of measurements on test structures after the wafer has been fully processed, results of functional tests on the semiconductor chips, the yield of defect-free semiconductor chips, etc., analyzed. In this way, both the stability of the production equipment and also the process stability and in this way also the production productivity and the product quality of the manufactured semiconductor chips can be substantially increased.

Despite these considerable advantages, which are offered by the use of automated data acquisition and data analysis, ie the use of, for example, APC processes, the APC processes in semiconductor chip manufacturing have hitherto been used only very occasionally. This is due in particular to the great variety of different production systems and different processes in the production of semiconductor chips and the associated significant number of recorded in different data formats raw data and the significant differences in the raw data due to different manufacturing equipment, sensors and software versions of the control program of the manufacturing equipment, which characterize the production plants or the manufacturing processes in the context of chip production.

If, within the scope of an automated data analysis, at least part of the sensor parameters of a production plant or production process to be monitored and monitored would use individual monitoring by means of an individual specification limit for monitoring the data, then the number of monitored parameters and the monitoring algorithms to be set would be multiplicative the number of recorded sensor parameters, the number of manufacturing equipment used in semiconductor chip manufacturing, the different products to be produced, that is, the different semiconductor chips and the production recipes used for production or even the individual recipe steps.

This shows the considerable number of parameters to be analyzed and monitored, which is no longer manageable in practice.

Also, with such a single monitor, it will no longer be possible to consider influences of individual specification parameters on other parameters. This is an analysis of several Machines no longer possible, in other words, no machine comparison is possible. In particular, no correlations between parameters of the same process, correlations between parameters of different processes, for example, a pre-process and a post-process can be determined.

Furthermore, it is known to calculate key figures for individual sensor quantities determined by a respective sensor, that is to say the respective raw data parameters. This usually takes place by means of a configuration per monitored machine integrated in APC and per recipe or recipe step used during the production of a semiconductor chip.

In the analysis of the parameters and the monitoring of the parameters with the specification limits assigned to the parameters, it may be necessary to additionally differentiate between the respective different manufactured products or product types.

This means a very high configuration effort and maintenance effort, d. H. Maintenance, the respective analysis, monitoring and control tools.

For this reason, APC methods have been used only very occasionally at individual positions in the context of the production process of semiconductor chips and only to a very limited extent.

A systematic comparison between different machines used in a manufacturing process, between recipes or even between different products or product types, ..., is hardly or not at all possible in this way.

Furthermore, a standardized so-called SECS interface (Semiconductor Equipment Communication Standard interface) is known.

From the publication US 5,777,876 For example, a method for controlling a manufacturing process is known in which operating parameters for controlling the manufacturing process are stored using a plurality of table structures. In this case, a grouping of operating parameters can be carried out, for example, in such a way that the same operating parameters are used to control the machines for the production of a product using similar machines.

The publication DE 43 10 332 A1 describes a method for determining optimized parameter values of a casting process, in which known data records with individual parameters are processed in an optimization program in such a way that the individual data sets are automatically traversed without machine standstill. In a subsequent automatic quality control of the sample parts produced, data sets with optimal parameter values for series production are crystallized.

From the publication DE 197 45 386 A1 For example, a method for automated production of wafers in mixed batches is known. In this case, a networked computer controller for an entire production line controls the production devices in a plurality of process sections, wherein at least one production device is classified as a respective process section.

The publication DE 198 07 343 A1 describes a computer aided manufacturing planning system for planning the processing steps in each of a plurality of work processes for manufacturing a particular new product, in which characteristic data representing the characteristics of already planned products and manufacturing data representing the processing steps and work processes for manufacturing these products are related stored to each other in a data storage device.

The invention is based on the problem to capture production data in a production plant, in particular in a chip production plant in a simplified and cost-effective manner compared to the prior art and to provide for analysis.

The problem is solved by the method and the system for detecting production data in a production plant, in particular in a chip production plant, with the features according to the independent patent claims.

In a method for acquiring production data in a production plant with several partial production plants, partial production data are determined by the partial production plants and saved. Using at least one grouping table, the partial production data are grouped into groups of the grouping table to map the partial production data to compressed partial production data by producing compressed key figures by means of a statistical analysis for sub-production data grouped according to the assignment of the partial production data to groups of the grouping table become. In the grouping table, machines of the sub-production plants are grouped into at least one machine group or machine components of the sub-production plants for at least one machine component group. Alternatively, manufacturing recipes, in particular chip production recipes, for the production of a product or an intermediate product are grouped into at least one recipe group or production parameters, in particular chip production parameters, in the context of the production of a product or an intermediate product for at least one parameter group.

• • Maschinen der Teil-Produktionsanlage zu mindestens einer Maschinengruppe gruppiert, oder
• • Maschinenkomponenten der Teil-Produktionsanlagen zu mindestens einer Maschinenkomponentengruppe gruppiert, oder
• • Herstellungsrezepte, insbesondere Chip-Herstellungsrezepte, zur Herstellung eines Produkts oder eines Zwischenprodukts zu mindestens einer Herstellungsrezeptgruppe gruppiert, oder
• • Herstellungsparameter, insbesondere Chip-Herstellungsparameter, im Rahmen der Herstellung eines Produkts oder eines Zwischenprodukts zu mindestens einer Herstellungsparametergruppe gruppiert.

A system for acquiring production data in a production facility has multiple sub-production facilities, and the sub-facilities may generate or generate sub-production data. Furthermore, at least one memory is provided for storing the generated partial production data and an evaluation computer coupled to the memory, which is set up in such a way that the partial production data can be mapped to compressed using at least one grouping table with assignment of the partial production data to groups of the grouping table Partial production data in which compressed key figures are generated by means of a statistical analysis for partial production data grouped according to the assignment of the partial production data to groups of the grouping table. In the grouping table are

• • machines of the sub-production plant grouped into at least one group of machines, or
• • grouped machine components of the sub-production plants into at least one machine component group, or
• • manufacturing recipes, in particular chip manufacturing recipes, grouped to form at least one production recipe group for producing a product or an intermediate product, or
• • Manufacturing parameters, in particular chip manufacturing parameters, grouped into at least one production parameter group as part of the production of a product or intermediate.

In this context, a production facility is to be understood as meaning a system or arrangement in which a product or an intermediate product is produced using different raw materials. It should be noted in this context that the more complex the manufacturing process is constructed, the more efficiently the invention can be used if at least some similar machines, processes, recipes, or parameters are used, but due to slightly different nomenclatures used or different structure different software version of the control program used in each case can not be readily detected and processed together in the context of data acquisition.

The invention is particularly in the context of semiconductor manufacturing, d. H. For example, in the production of monocrystalline or polycrystalline silicon or other semiconductor material, or even the semiconductor chip production, d. H. in the context of the production of semiconductor chips in a chip production plant, can be used advantageously, since in particular in such processes, the above criteria are met very well. However, it should be pointed out in this connection that the invention can be used advantageously in any production process, in particular if many process steps are necessary and information about the production process can be made available automatically by the production plants.

In this context, a chip production facility is to be understood as a system or arrangement in which different raw materials are used, for example using semiconductor materials such as silicon or else other IV main group semiconductor materials (for example germanium) or binary , ternary or even quaternary III-V compound semiconductor materials (for example indium gallium arsenide phosphide, indium gallium arsenide antimonide, etc.), or binary, ternary or even quaternary II-VI compound semiconductor materials semiconductor Components, in particular semiconductor chips are manufactured.

In the context of this description, a semiconductor chip is to be understood as meaning, for example, a memory chip, a microprocessor chip, a communication chip, a chip with an integrated semiconductor laser element, and a chip optimized to an arbitrarily predefinable hardware function a communication chip for decoding received radio signals or a chip for processing video signals.

According to the invention, the chip production plant has several partial production plants, for example different machines with which the physical or chemical process steps necessary for the entire production process of a chip can be carried out.

• • eine Kurzzeittemper-Vorrichtung (Rapid Thermal Processing-Vorrichtung, RTP-Vorrichtung),
• • ein Ofen zum Erhitzen der zu bearbeitenden Wafer,
• • eine Ätzvorrichtung, beispielsweise eine Plasmaätz-Vorrichtung, oder eine Trockenätz-Vorrichtung,
• • eine Lithographie-Vorrichtung,
• • eine Nassbehandlungs-Vorrichtung zum Ätzen, Lackentfernen, Reinigen oder Verändern der Produkt-Oberfläche,
• • eine CMP-Vorrichtung, d. h. eine Vorrichtung zum Durchführen chemisch-mechanischen Polierens,
• • eine Ionen-Implantations-Vorrichtung,
• • eine Abscheide-Vorrichtung zum Abscheiden von Schichten auf den Wafer, beispielsweise eine Abscheide-Vorrichtung zum physikalischen Abscheiden aus der Gasphase (Physical Vapour Deposition, PVD) oder zum chemischen Abscheiden aus der Gasphase (Chemical Vapour Deposition, CVD),
• • eine Messvorrichtung zum Messen vorgegebener Waferparameter oder Prozessparameter,
• • eine Test-Vorrichtung zum Testen hergestellter Wafer.

An example, particularly in the front-end area of chip manufacturing, are devices for carrying out the following process steps:

• A rapid thermal processing device (RTP device),
• A furnace for heating the wafers to be processed,
• An etching device, for example a plasma etching device, or a dry etching device,
• A lithography apparatus,
• A wet treatment device for etching, paint removal, cleaning or changing the product surface,
• A CMP device, ie a device for performing chemical mechanical polishing,
• An ion implantation device,
• A deposition device for depositing layers onto the wafer, for example a physical vapor deposition (PVD) or chemical vapor deposition (CVD) deposition device,
• A measuring device for measuring predetermined wafer parameters or process parameters,
• A test device for testing manufactured wafers.

Depending on the particular product to be produced, for example, depending on the desired configuration of the chip to be manufactured, a multiplicity of different devices, that is to say partial production plants, are provided in the chip production plant and coupled to each other for the realization of the respective overall semiconductor chip manufacturing process.

It should be noted in this context that the sequence of the individual process steps in the respective sub-production plants either "sequentially", d. H. always first in a first plant group A (eg, to be painted) before the second plant group B (eg, to expose) for two production steps to take place one after the other, or "parallel," d. H. in a first plant B1 (eg scanner type XY100) or a second plant B2 (eg stepper type Extra2000) for two different plants and processes that can alternatively carry out one and the same production step.

The machines may be grouped into sufficiently similar properties depending on their respective characteristics to machine groups with machines in terms of a similarity criterion. This clearly corresponds to the parallel sequence of the respective process steps in different machines.

Machines in the field of chemical vapor deposition as well as all other respective machines can be divided into different machine types. For example, a CVD machine typically, but not necessarily, includes a so-called mainframe, that is, a robot or robotic arm and a robotic controller associated with the robot or robot arm for distributing and controlling the respective wafers to the individual process chambers of the CVD machine and at least one process chamber for processing the respective wafers in accordance with the method set in the process chamber. In the case that no mainframe is provided, only the individual process chambers are specific to the process area.

• • einem Mainframe und vier Prozesskammern,
• • einem Mainframe und zwei Prozesskammern, sowie
• • nur einer Prozesskammer und keinem Mainframe.

Thus, such a CVD machine can be of a different type of machine, for example, a CVD machine can be distinguished in machines with:

• • one mainframe and four process chambers,
• • a mainframe and two process chambers, as well
• • only one process chamber and no mainframe.

Generally, however, depending on the requirement, basically any number of process chambers or other machine components can each form a machine type of a machine, in particular a CVD machine.

A machine thus has different machine components, for example the mainframe or a process chamber. For example, a deposition chamber, in which corresponding materials or layers are deposited on the wafer to be processed using the deposition chamber, is understood to mean a process chamber. A process chamber is also to be understood as meaning an etching chamber in which a wafer fed to the etching chamber is used by means of a wafer the etch chamber also structured gas, or plasma or a liquid supplied, that is etched. A process chamber can also be set up in such a way that it epitaxially grows in its material on the respective surface of the wafer to be processed.

Again, the individual machine components may be of different types, hereafter referred to as the respective machine component type.

The types of machine components typically depend on the manufacturer of the machine component who provides a respective machine component.

Partial production data is determined by each sub-manufacturing facility, typically using the sensors associated with the respective machinery or machine components, and stored as partial production data in one or more electronic files. Thus, the partial production data can be stored unambiguously in each file assigned to a machine or a machine component, in particular using the SECS standards (Semiconductor Equipment Communication Standard).

The stored partial production data is now mapped to compressed partial production data using the grouping table, wherein the grouping of the respective machines, machine components, manufacturing processes or chip manufacturing parameters taking into account their respective characteristics and the similarity of their respective properties to other machines, machine components, chip -Herstellungsrezepten or chip production parameters, takes place.

By exploiting the respective similarities of the individual elements to one another, a considerable reduction of partial production data to be taken into account is achieved, whereby for the first time a data volume to be taken into account is achieved which can be achieved with conventional personal computers, ie with conventional monitoring tools, analysis tools, and control systems. Tools can be analyzed, monitored and thus also controlled. Furthermore, an intelligent grouping of the individual elements is provided according to the invention, which improves the statistics and ensures comparability of the partial production data to be considered.

For the first time, a practicable and cost-effective, area-wide and cross-departmental possibility for automated data analysis and process monitoring and control, ie for Advanced Process Control (APC) in the context of semiconductor chip manufacturing is possible.

Preferred developments of the invention will become apparent from the dependent claims. The further embodiments of the invention relate to both the method and the system for detecting production data in a chip production plant.

According to an embodiment of the invention, the partial production data are transmitted from the partial production plants to a central store and stored there in at least one central file, wherein preferably several files are provided for storing the partial production data which are stored in directories, each of which is a machine or process chamber, generally a part-production plant, are stored.

In this way, a centralized and thus easy-to-monitor, easily configurable and therefore cost-maintainable option for semiconductor chip manufacturing process monitoring is provided.

Further, the machines may be grouped into machine groups each having sufficiently similar machine characteristics with respect to a machine similarity criterion.

The machine components may also be grouped into machine component groups each having sufficiently similar machine component characteristics with respect to a machine component similarity criterion.

Further, those semiconductor chip manufacturing recipes may be grouped into recipe groups each having sufficiently similar recipes for a semiconductor chip manufacturing recipe similarity criterion.

Furthermore, those chip fabrication parameters may be grouped into parameter groups that have sufficiently similar chip fabrication parameters with respect to a chip fabrication parameter similarity criterion.

The groupings may alternatively or jointly take place in different or in a common grouping table. The more areas a grouping is provided, the greater the achievable degree of compression of the amount of data to be considered.

The configuration of the similarity criteria depends on the respective device or on the respective parameters or recipes which are used in the (chip) production plant.

In other words, this means that both the definition of the similarity criteria and the grouping of the respective components into groups can vary greatly and are very process-dependent, product-dependent and also machine-dependent or machine-component-dependent.

Furthermore, according to a further embodiment of the invention, the compressed partial production data can be subjected to a subsequent, preferably statistical, data analysis, whereby a further improved and more accurate monitoring of the entire manufacturing process and of the production equipment integrated in the production process is achieved.

Illustratively, the invention can be seen in the provision of a procedure for analyzing and monitoring plant data and process data in the context of APC, in which different machines and processes can be recorded together. With the help of only a few optimally formed grouping tables, which are also referred to as assignment tables or mapping tables, it is possible according to the invention within a corresponding order hierarchy gradually for the monitoring itself irrelevant differences between the individual machines, machine groups, recipes or recipe parameters and compensate to summarize similar elements. A similar configuration within the scope of the APC thus no longer takes place per machine, machine component, recipe or recipe step, as per the prior art, but only per machine group, machine component group, recipe group or recipe parameter group.

This has in particular the considerable advantage of only a small configuration effort and maintenance of the monitoring device and the monitoring programs used.

Thus, from the raw data material, common key figures are generated in a very effective manner, which have very high information content about the properties and the behavior of the respective machines and all production processes. Furthermore, according to the invention, therefore, a long-term comparison of the data of a small subgroup of sub-production plants with the data of the entirety of the group of sub-production plants is possible, allowing a quasi-continuous monitoring, despite discontinuous data of the respective subgroup of the sub-production plants.

Thus, it has become possible for the first time to use automated process monitoring, that is APC nationwide and profitable in semiconductor chip manufacturing. Even within a larger group of production facilities that are not necessarily identical, relevant information such as differences between machines / machine components of a machine group, offsets, drifts, trends, outliers etc. can be detected and analyzed in a large number of used manufacturing processes.

In summary, the invention can thus be seen further in that the temporal signal profile of a sensor can be represented in a compressed manner in characteristic numbers which characterize the respective signal course and the machine or the process.

The configuration of these measures is done together for a large number of different plants and processes using optimized, preferably hierarchical, organized allocation tables, the grouping tables.

Although the invention will be explained in more detail below using the example of a chip production plant, it should be pointed out that the invention is not restricted thereto but can be used in all production plants with partial production plants, for example also in the pharmaceutical industry Manufacture of pharmaceutical products.

An embodiment of the invention is illustrated in the figures and will be explained in more detail below.

Show it

1 a block diagram showing the general organization of a chip production plant is shown;

2 a sketch of a chip production plant according to an embodiment of the invention, wherein the complex material flow, ie the path of a wafer / lot is represented by the chip production plant;

3 a block diagram in which the data flow is shown according to the embodiment of the invention;

4 a block diagram in which the detection, transmission and storage of the raw data detected by the sensors is shown;

5 an extract of a log file of a machine with raw data, which log file has been generated by a sensor;

6 a section of a file according to the PDSF format in which the logistical information is stored;

7 another file according to the PDSF format, in which the sensor parameters are stored per unit time;

8th a representation of a machine with a mainframe and four process chambers;

9a to 9c Sketches of three different machine types of CVD machines, which are grouped into machine groups according to an embodiment of the invention, wherein a first type of machine has a mainframe and four process chambers ( 9a ), a second machine type a mainframe and two process chambers ( 9b ) and a third type of machine a process chamber ( 9c ), wherein the machines of the machine types grouped into a machine group are jointly configurable and interrogatable;

10 a block diagram outlining different grouping options;

11 a representation of a grouping table in which several machines are grouped into component types;

12a to 12g Block diagrams in which for the individual machines, that is, partial production facilities, according to the embodiment, the raw data recording are shown;

13 a table showing the assignment of the parameter sets as part of the raw data recording;

14 a table in which the parameter configuration for machines of a machine type with different mainframes is shown;

15 a table showing the parameter configuration for all the deposition process chambers shown in this embodiment;

16 a table showing the parameter configuration for all the etchback process chambers shown in this embodiment;

17 a block diagram and tables in which all the recipes used in this embodiment of the machine group according to this embodiment are shown as part of a manufacturing process of a semiconductor chip manufacturing;

18 a table in which individual recipes are assigned to recipe groups, also referred to as recipe sets;

19 a table showing the recipe configurations of the deposition process chambers of the machine group;

20 a table showing the recipe configurations of the etch back process chambers of the machine group;

21 a table in which individual machines, component types, parameter sets as well as recipe sets are assigned to different key figure sets;

22 a grouping table, in which the key figure configuration for a mainframe, according to this embodiment for the mainframe Centura and P5000, is shown;

23 a table showing the metric configuration for the deposition process chambers; and

24 a table that shows the key figure configuration for the re-etching process chambers.

In the following, identical reference numerals are used in the figures for the same elements.

1 shows in a block diagram 100 the organization and construction of a semiconductor chip production plant according to an embodiment of the invention.

• • einen ersten Bereich, in den die Front-End-Prozesse der Chip-Fertigung gruppiert werden (Block 102),
• • einen zweiten Bereich des Herstellungsprozesses, in den die Back-End-Prozesse gruppiert werden (Block 103),
• • einen dritten Bereich des Herstellungsprozesses, der den Support betrifft, das heißt die Unterstützung der einzelnen Herstellungsprozesse (Block 104),
• • ein vierter Bereich der die Prozesstechnologie und die Prozessintegration betrifft (Block 105).

The entire manufacturing process, referred to in 1 with a first block 101 , becomes according to this embodiment in four production areas 102 . 103 . 104 . 105 grouped

• • a first area into which the front-end processes of the chip production are grouped (block 102 )
• • a second part of the manufacturing process into which the back-end processes are grouped (block 103 )
• • a third area of the manufacturing process, which concerns the support, that is the support of the individual manufacturing processes (block 104 )
• • A fourth area of process technology and process integration (block 105 ).

• • Ein Ofen zum Erhitzen des jeweiligen zu prozessierenden Wafers,
• • eine Vorrichtung zum Durchführen eines Kurzzeit-Temperverfahrens (Rapid Thermal Processing, RTP),
• • eine Vorrichtung zum Ätzen des Wafers, beispielsweise zum Nassätzen oder zum Trockenätzen,
• • eine Vorrichtung zum Reinigen, beispielsweise Waschen des Wafers,
• • eine Vorrichtung zur Durchführung unterschiedlicher Lithographieschritte,
• • einer Vorrichtung zum chemisch-mechanischen Polieren (Chemical Mechanical Polishing, CMP),
• • eine Einrichtung zur Durchführung einer Ionen-Implantation in vorgegebenen Bereichen des Wafers bzw. des jeweils zu fertigenden Chips,
• • Vorrichtungen zum Aufbringen von Materialien auf dem Wafer, beispielsweise Vorrichtungen zum Abscheiden von Materialien aus der Gasphase, das heißt beispielsweise Vorrichtungen zum Durchführen von Physical Vapour Deposition (PVD) oder Chemical Vapour Deposition (CDV), oder eine Vorrichtung zum epitaktischen Aufwachsen von Material auf einem Substrat,
• • Metrologie-Vorrichtungen, d. h. Messvorrichtungen,
• • Vorrichtungen zum Durchführen von Tests auf den jeweiligen Wafern.

In the front-end processes 102 In particular, the following process technologies and the devices designed to carry out the corresponding processes are provided:

• An oven for heating the respective wafer to be processed;
• A device for carrying out a rapid thermal processing (RTP),
• A device for etching the wafer, for example for wet etching or dry etching,
• A device for cleaning, for example washing the wafer,
• A device for carrying out different lithography steps,
• A device for chemical mechanical polishing (CMP),
• A device for carrying out an ion implantation in predetermined regions of the wafer or the respective chip to be produced,
• Devices for applying materials to the wafer, for example devices for depositing materials from the gas phase, that is, for example, devices for performing Physical Vapor Deposition (PVD) or Chemical Vapor Deposition (CDV), or a device for epitaxial growth of material a substrate,
• • Metrology devices, ie measuring devices,
• • Devices for performing tests on the respective wafers.

• • Den Zusammenbau der Chips in Gehäuse,
• • den Endtest des gefertigten und gehäusten Chips,
• • das Einbringen von Informationen, beispielsweise Produktinformation, in oder an das Gehäuse des jeweiligen Chips, sowie
• • allgemein die im Bereich des Back-Ends für gehäuste und ungehäuste Chip eingesetzten Technologien.

The back-end processes mainly concern the following areas:

• • assembling the chips into housings,
• The final test of the finished and packaged chip,
• • The introduction of information, such as product information, in or on the housing of the respective chip, as well
• • In general, the technologies used in the area of the back-end for packaged and unhoused chips.

• • CIM,
• • Prozessüberwachung,
• • ein Transportsystem zur Auslieferung der gefertigten Halbleiterchips,
• • eine Koordination der Produktion,
• • die Unterstützung der jeweiligen Fertigungsstandorte.

The support, ie the process support concerns in particular the following areas:

• • CIM,
• • process monitoring,
• A transport system for delivery of the manufactured semiconductor chips,
• • coordination of production,
• • the support of the respective production sites.

• • die Prozessintegration von Logikbausteinen,
• • die Prozessintegration von Speicherbausteinen,
• • das Produkt-Engineering,
• • das überwachen und Verbessern von Fehlerdichten bei der Herstellung,
• • das überwachen elektrischer Parameter bei den hergestellten Produkten,
• • die Verbesserung der Ausbeute der hergestellten Chips,
• • eine physikalische Fehleranalyse.

The process technology and the process integration concerns in particular

• • the process integration of logic modules,
• • the process integration of memory modules,
• • product engineering,
• • monitoring and improving defect densities in manufacturing,
• • monitoring of electrical parameters of manufactured products,
• The improvement of the yield of the produced chips,
• • a physical error analysis.

2 shows a semiconductor chip production plant, in other words, a semiconductor chip factory 200 according to an embodiment of the invention with a plurality of semiconductor chip part production equipment 201 which are used to process raw materials, such as a silicon wafer or a wafer, from other semiconductor materials (germanium, gallium arsenide, indium phosphide, etc.) to make semiconductor chips from the raw materials.

A common manufacturing process for producing a semiconductor chip comprises hundreds of different process steps, in which lithography steps, etching steps, CMP steps, steps of depositing materials on the respective wafer to be processed, or also doping or implantation steps of doping atoms in the in processing wafers are performed.

Thus arise, as in 2 through lines 202 which shows the path of a wafer or a lot through the semiconductor chip production plant 200 represent. In the semiconductor chip production plant 200 is a variety of sensors, which are the respective part-production equipment 201 are assigned and an even larger amount of raw data, which are respectively detected by the sensors and, as will be explained in more detail, are processed. A respective sensor can be integrated into a respective machine (integrated sensor) or be attached separately to a respective machine (external sensor).

In addition, the partial production plants 201 also as machines 201 designated.

3 shows an example of the data flow for data, which on a machine 201 be detected by means of an integrated or by means of an external sensor 301 , Every sensor 301 , Wherein any number of integrated and / or external sensors can be provided, detects the respectively predetermined for him parameters of the machine 201 For example, physical or chemical states in a process chamber, the position of a robotic arm, etc.

The sensor 301 is via a SECS interface 302 set up for data communication in accordance with the SECS standards with a Local Area Network (LAN) 306 coupled.

According to the SECS standards are used by the sensor 301 and the SECS interface 302 Files generated according to the PDSF format (Process Data Standard Format), also referred to below as PDSF files 303 as well as log files 304 where the PDSF files 303 and the log files 304 as raw data in a memory 307 , are stored according to this embodiment for a limited period of preferably 6 months.

According to this embodiment, the PDSF files contain 303 analogue data of up to 70 channels, ie of up to 70 different internal (ie integrated) and / or external sensors 301 which on a machine 201 can be attached. According to this embodiment, per month and machine 201 about 150 MB of data is generated and stored in memory 307 saved.

In the log files 304 are machine-specific different data stored, such as a time specification, event information about in the machine 201 occurred events, alarms, logistic data. According to this embodiment, about 1 Mbyte of data in a log file per month and machine 304 in the store 307 saved.

The ones in the store 307 stored raw data are, as explained in more detail below, grouped, that is filtered and fed to different algorithms, in 3 symbolized by a block 308 and when a predetermined alarm or message criterion is met, an alarm or message is generated and via an alarm / message transmission unit 309 over the local communication network 306 and the SECS interface 302 to the machine 201 which also directly with the SECS interface 302 is coupled, or to hereafter explained client computer 312 transfer.

Furthermore, process metrics are generated from the raw data using the filter algorithms and stored in an Oracle ^TM database 310 stored, as will be explained in more detail below. According to this embodiment, about 6 MB of data per month and machine become data as key data in the database 310 saved.

Using the code numbers are using different software tools, in 3 symbolized by a block 311 Perform statistical analysis, generate graphs, create reports or provide a model of the manufacturing process to further simulate the manufacturing process.

The corresponding data is in turn transmitted via the local communication network 306 a variety of client machines 312 provided by users, such as machine engineers, process engineers or product engineers, so that they can take appropriate measures to optimize the semiconductor manufacturing process using the analysis data.

According to this embodiment, machines will be referred to hereinafter 201 from the field of front-end processing, however, it should be noted that the invention is by no means limited to the field of front-end processes but, in particular, concerns all areas of semiconductor chip manufacturing.

• • Physikalische Abscheidung aus der Gasphase (Physical Vapour Deposition, PVD),
• • Chemische Abscheidung aus der Gasphase (Chemical Vapour Deposition, CVD),
• • Epitaxie,
• • Lithographie-Verfahren,
• • Ätzverfahren, insbesondere Plasmaätzverfahren, Trockenätzverfahren, Nassätzverfahren, etc.

As explained above, the semiconductor chip production equipment becomes 200 grouped in the area of the front-end processes according to this embodiment into several process areas, in particular in:

• Physical vapor deposition (PVD),
• Chemical vapor deposition (CVD),
• • epitaxy,
• Lithography process,
• Etching process, in particular plasma etching, dry etching, wet etching, etc.

According to the invention, within the process areas of the front-end processes, the grouping of the machines used in each case 201 so done that all machines 201 , which are to be jointly configurable or queried, are assigned to a machine group.

• • eine erste Maschinengruppe zur Abscheidung von TEOS,
• • eine zweite Maschinengruppe zur Abscheidung von Silan_1,
• • eine dritte Maschinengruppe zur Abscheidung von Silan_2,
• • eine vierte Maschinengruppe zum Abscheiden und zum Rückätzen von Wolfram (insbesondere Anlagen des Herstellers Applied Materials^TM).

For example, the process area of chemical vapor deposition is subdivided into four different machine groups that provide different processes in a CVD process, namely:

• A first machine group for the deposition of TEOS,
• A second machine group for depositing Silan_1,
• • a third machine group for the deposition of Silan_2,
• A fourth group of machines for the deposition and re-etching of tungsten (in particular plants of the manufacturer Applied Materials ^TM ).

The second and third machine group differ according to this embodiment, despite the same application area in terms of the logged and stored process parameters and the processes used in each case so strong that a common configurability and interrogation of the machines would not make sense in a same machine group.

The above example shows that the grouping of the machines used 201 is highly dependent on the machines 201 themselves and their respective field of application in the context of the corresponding manufacturing process.

4 shows the raw data record in a block diagram 400 with seven machines 401 . 402 . 403 . 404 . 405 . 406 . 407 , which each have a SECS interface assigned to it 408 . 409 . 410 . 411 . 412 . 413 . 414 each with a personal computer 415 . 416 . 417 . 418 . 419 . 420 . 421 is coupled. About the respective SECS interface 408 . 409 . 410 . 411 . 412 . 413 . 414 transmits the machine 401 . 402 . 403 . 404 . 405 . 406 . 407 or the integrated sensor (s) and / or external sensor (s) of the machine 401 . 402 . 403 . 404 . 405 . 406 . 407 assigned, determined raw data to the respective personal computer 415 . 416 . 417 . 418 . 419 . 420 . 421 which personal computer 415 . 416 . 417 . 418 . 419 . 420 . 421 the received data to a raw data file server 422 transfer. In the raw data file server 422 are those of the personal computer 415 . 416 . 417 . 418 . 419 . 420 . 421 transmitted raw data in each case a plurality of electronic files 423 . 424 . 425 . 426 . 427 . 428 . 429 , each stored in a respective directory, each of a machine or machine component 401 . 402 . 403 . 404 . 405 . 406 . 407 clearly assigned, stored.

In other words, this means that in the raw data file server 422 each in a directory 423 . 424 . 425 . 426 . 427 . 428 . 429 only the raw data of exactly one machine or machine component 401 . 402 . 403 . 404 . 405 . 406 . 407 are stored, so that a unique mapping of in the raw data file server 422 stored files in the directories 423 . 424 . 425 . 426 . 427 . 428 . 429 to each in the directories (folders) 423 . 424 . 425 . 426 . 427 . 428 . 429 stored files generating machine 401 . 402 . 403 . 404 . 405 . 406 . 407 is possible.

In this context, it should be noted that according to alternative embodiments of the invention also not necessarily a 1: 1 assignment of SECS interface 408 . 409 . 410 . 411 . 412 . 413 . 414 to a personal computer 415 . 416 . 417 . 418 . 419 . 420 . 421 is required, but also generally several personal computers for a SECS interface or a personal computer for multiple SECS interfaces can be provided.

5 clearly shows a section of a log file 304 that of the machine and the SECS interface 302 each time is generated and the extraction of key figures from the corresponding log file 304 ,

For example, as in 5 is shown, from the indication "Start Load - 17-without-4c" determines the start of a process sequence of a machine, namely according to the log file 304 the 14.08.1998 at 04:08:25.

The time that a pump is activated in a process chamber becomes two events from the log file 304 determined, according to this embodiment from the event "Load Chamber Pumpdown Started", which has started at a first time, namely 04:09:04 on 14.08.1998 and the event "Load Chamber Pumpdown Done", logged on 14.08.1998 at 04 : 32: 19 generated by subtraction of the two logged times.

Another code, which according to this embodiment of the log file 304 is the indication of a recipe, according to this embodiment of the recipe with the name "C3-MET-1000".

• • in einer ersten Spalte 601 eine Index-Angabe,
• • in einer zweiten Spalte 602 eine Rezeptangabe,
• • in einer dritten Spalte 603 eine Angabe der Kassette, in welcher sich der jeweilige Wafer befindet, auf den sich die protokollierte Angabe bezieht,
• • in einer vierten Spalte 604 eine Angabe des Slots, in welchem sich der jeweilige bezeichnete Wafer befindet,
• • in einer fünften Spalte 605 eine Losangabe, in der das Los angegeben ist, auf welches in den der Wafer enthalten ist, auf den sich die jeweilige Angabe bezieht,
• • in einer sechsten Spalte 606 eine Angabe des jeweiligen Verfahrensschritts angegeben ist, der zu diesem Zeitpunkt gerade durchgeführt wird, sowie
• • in dem Ausschnitt der PDSF-Datei in 7 in einer Zeitspalte 701 eine Zeitangabe, über weitere Spalten 702 und 703 jeweils Verknüpfungsangaben zur Indizierung der jeweiligen Zeile des Ausschnitts aus der PDSF-Datei 303 gemäß 6 angegeben ist, sowie in weiteren Spalten 704, 705, 706, 707 die Angabe eines jeweiligen Prozessschrittes (Steg A, Step B) sowie die Angabe von in dem jeweiligen Schritt in der jeweiligen Prozesskammer herrschenden Drucks (Pressure_A, Pressure_B).

In the 6 and 7 is an example of a section of a PDSF file 303 shown, where

• • in a first column 601 an index,
• • in a second column 602 a recipe,
• • in a third column 603 an indication of the cassette in which the respective wafer is located, to which the recorded indication relates,
• • in a fourth column 604 an indication of the slot in which the particular designated wafer is located,
• • in a fifth column 605 a lot indicating the lot containing the wafer to which the claim relates,
• • in a sixth column 606 an indication of the respective process step is being performed at that time, as well as
• • in the section of the PDSF file in 7 in a time column 701 a time, about more columns 702 and 703 Link information for indexing the respective line of the section from the PDSF file 303 according to 6 specified, and in other columns 704 . 705 . 706 . 707 the indication of a respective process step (web A, step B) as well as the specification of pressure prevailing in the respective step in the respective process chamber (Pressure_A, Pressure_B).

Typically, about 50 different parameters are recorded per machine. In addition to the parameters, lot information is also logged together with logistical information, as explained above.

So every data point, that is every line in a log file can 304 or in a PDSF file 303 time clearly a lot and appropriate logistical information such as the recipe used in each case, the respective technology or the operation performed in each case be performed in a process chamber.

Physically identical parameters often have different names, or because of different sensors or versions of machines or sensors, the same parameters are often recorded in different physical units; For example, a first machine records pressure in Torr, but a second machine in unit mTorr.

8th shows an example of the representation of a machine 800 with a mainframe 801 and four process chambers 802 . 803 . 804 . 805 which can be designed in the same way or in different ways to carry out the same or different processes and recipes.

Referring to the 4 and 9a to 9c are the machines provided according to this embodiment 401 . 402 . 403 . 404 . 405 . 406 . 407 Machines of three different types, one in 9a illustrated first type of machine, a in 9b illustrated second type of machine and a in 9c illustrated third type of machine.

The first machine 401 and the second machine 402 According to the invention are grouped into a second machine group, according to this embodiment of a first type of machine, as in 9b shown.

The machines 401 . 402 of the first type of machine are a Centura ^™ Applied Materials ^™ multi-chamber machine with four process chambers, a Centura ^™ mainframe, and two each loading stations for feeding and discharging wafers to and from the machine.

• • eine erste Ladestation 901 als erste Ladekomponente bezeichnet mit A,
• • eine zweite Ladestation 902 als zweite Ladekomponente bezeichnet mit B,
• • ein Centura^TM-Mainframe 903 als Mainframe-Komponente bezeichnet mit MF (Komponente_1),
• • eine erste Prozesskammer A 904, eingerichtet als Abscheide-Prozesskammer, im Weiteren bezeichnet als Komponente_3,
• • eine zweite Prozesskammer B 905, eingerichtet ebenfalls als Abscheide-Prozesskammer, im Weiteren bezeichnet als Komponente_3,
• • eine dritte Prozesskammer C 906, eingerichtet als Rückätz-Prozesskammer, im Weiteren bezeichnet als Komponente_4,
• • eine vierte Prozesskammer D 907, eingerichtet als Rückätz-Prozesskammer, im Weiteren auch bezeichnet als Komponente_4.

According to this embodiment, the first machine 401 The following machine components are grouped into component types due to the same or similar design:

• • a first charging station 901 as first charging component denoted by A,
• • a second charging station 902 as a second charge component labeled B,
• • a Centura ^TM mainframe 903 as mainframe component with MF (component_1),
• A first process chamber A 904 set up as a deposition process chamber, hereinafter referred to as component_3,
• A second process chamber B 905 also set up as a deposition process chamber, hereinafter referred to as component_3,
• • a third process chamber C 906 set up as an etching back process chamber, hereinafter referred to as component_4,
• • a fourth process chamber D 907 set up as an etching back process chamber, also referred to as component_4 hereinafter.

• • eine erste Ladestation 911 des ersten Komponententyps A,
• • eine zweite Ladestation 912 des zweiten Komponententyps B,
• • ein Centura^TM-Mainframe MF 913 als Komponente_1,
• • eine erste Prozesskammer A 914, eingerichtet als Abscheide-Prozesskammer (Komponente_3),
• • eine zweite Prozesskammer B 915, eingerichtet ebenfalls als Abscheide-Prozesskammer (Maschinenkomponente_3),
• • eine dritte Prozesskammer C 916, eingerichtet als Rückätz-Prozesskammer (Maschinenkomponente_4),
• • eine vierte Prozesskammer D 917, eingerichtet ebenfalls als Rückätz-Prozesskammer (Maschinenkomponente_4).

The second machine 402 has the following machine components as components, which are also grouped into component types in each case on the basis of identical or similar design:

• • a first charging station 911 of the first component type A,
• • a second charging station 912 of the second component type B,
• • a Centura ^TM mainframe MF 913 as component_1,
• A first process chamber A 914 set up as a separation process chamber (component_3),
• A second process chamber B 915 , also set up as a separation process chamber (Maschinenkomponente_3),
• • a third process chamber C 916 set up as an etching back process chamber (machine component_4),
• • a fourth process chamber D 917 , also set up as an etching-back process chamber (machine component_4).

Thus, the first machine 401 and the second machine 402 Due to their very similar design and due to the same existing machine components of the same machine types machine component types to the first machine type, ie a machine types with a multi-chamber system with two charging stations, a mainframe and four process chambers, grouped.

The third machine 403 and the fourth machine 404 According to the invention are grouped into a second machine group, according to this embodiment of a second type of machine, as in 9b shown.

The third machine 403 and the fourth machine 404 are P5000 multi-process plants with two process chambers, a P5000 ^TM mainframe and two charging stations.

• • Eine erste Ladestation 921 des Komponententyps A,
• • eine zweite Ladestation 922 des Komponententyps B,
• • ein P5000-Mainframe MF 923 als Komponente_2, das heißt als eine zweite Maschinentyp-Mainframe-Komponente,
• • eine erste Prozesskammer B 924, eingerichtet als Abscheide-Prozesskammer (Komponente_3), sowie
• • eine zweite Prozesskammer D 925, eingerichtet als Rückätz-Prozesskammer (Komponente_4).

The third machine 403 has the following components of each of the following component types:

• • A first charging station 921 of the component type A,
• • a second charging station 922 of the component type B,
• • a P5000 mainframe MF 923 as component_2, that is, as a second machine type mainframe component,
• A first process chamber B 924 , set up as a separation process chamber (component_3), as well
• • a second process chamber D 925 , set up as an etch-back process chamber (component_4).

• • Eine erste Ladestation 931 des Komponententyps A,
• • eine zweite Ladestation 932 des Komponententyps B,
• • ein P5000^TM-Mainframe MF 933 als Komponente_2,
• • eine erste Prozesskammer A 934, eingerichtet als Abscheide-Prozesskammer als Maschinenkomponente_3,
• • sowie eine zweite Prozesskammer C 935, eingerichtet als Rückätz-Prozesskammer als Maschinenkomponente_4.

The fourth machine 404 has the following components:

• • A first charging station 931 of the component type A,
• • a second charging station 932 of the component type B,
• • a P5000 ^TM mainframe MF 933 as component_2,
• A first process chamber A 934 set up as a separation process chamber as machine component_3,
• • and a second process chamber C 935 set up as an etching back process chamber as machine component_4.

As can be seen from the above, the third machine and the fourth machine are 403 and 404 to a machine group (same machine type), since the mainframe is the same in terms of the required functionality, so that a common consideration of the raw data generated in terms of the desired monitored data and their analysis is readily possible.

The fifth machine 405 , the sixth machine 406 as well as the seventh machine 407 According to the invention are grouped into a third group of machines, according to this embodiment, a third type of machine, as in 9c shown.

The machines of the third machine type are each a single-chamber system, each with a process chamber and only one charging station.

The fifth machine 405 has a charging station 941 as well as a process chamber 942 set up as a deposition process chamber (component_3).

The sixth machine 406 also has a charging station 951 as well as a process chamber 952 , also set up as a deposition process chamber (component_3).

The seventh machine 407 has a charging station 961 as well as a process chamber 962 , also set up as a deposition process chamber (component_3).

It should be noted in this connection that the assignment of the respective machine component to a machine component type (for example, component_2, component_3, component_4) is a physical interpretation of a person skilled in the respective machine, which corresponds to the corresponding specification of the machine 401 . 402 . 403 . 404 . 405 . 406 . 407 Takes into account. It should be noted that the assignment can also be automated.

As in 10 is shown, different grouping options within a machine group can be considered according to the invention.

According to this embodiment, the machines are grouped under the aspect of the machine itself or from the aspect of the component type.

• • Rohdatenaufzeichnung pro Maschine,
• • Maschinenspezifische Probleme,
• • Vor- bzw. Nachprozess auf einer Maschine.

If the machine is grouped under the aspect of the respective machine itself and its properties, the following grouping criteria are taken into account in particular:

• • raw data recording per machine,
• • machine-specific problems,
• • Pre- and post-process on one machine.

• • Die identische Konfiguration für jeweils einen gleichen Komponententyp,
• • der Vergleich gleicher Komponententypen über verschiedene Maschinen hinweg.

If the machines are grouped with regard to the component types contained in the respective machine, in particular the following objectives are considered:

• • The identical configuration for one identical component type,
• • the comparison of identical component types across different machines.

10 illustrates the different grouping options according to this embodiment of the invention.

11 shows in a grouping table 1100 an assignment of machines 1101 , Machine types 1102 , Machine components 1103 as well as machine component types 1104 within a machine group.

It should be noted in this context that the assignment can be automated both using appropriate computer programs that can automatically evaluate the similarities and differences of the machines, the machine components or the data generated by them, as well as manually by a person skilled in the respective machines and machine components or the data generated by them.

11 shows in the grouping table 1100 Furthermore, each of the machines assigned parameter sets 1105 , Recipe sets 1106 as well as key figure sets 1107 , which will be explained in more detail below.

Before a key figure can be defined, differences in parameter names and in physical units used by integrated and / or external sensors are adjusted using an assignment table.

According to this embodiment, one parameter set is used per machine component. In other words, different machine component types are assigned different parameter sets. However, several components can refer to the same parameter set. Depending on how differently the machines name and log the same parameters in each case, there are within the scope of the respective grouping table 1100 correspondingly more or less parameter sets.

• • Eine Zeitangabe;
• • für einen Mainframe, falls dieser vorhanden ist, der in der Transferkammer des Mainframes herrschende Druck,
• • sowie für alle Prozesskammern zusätzlich: • eine Rezeptschrittnummer des gerade in der Kammer ablaufenden Rezepts,
• • für die Abscheidekammern zusätzlich: • eine Angabe über den in der Abscheidekammer herrschenden Gasfluss, sowie • eine Angabe über den in der Abscheidekammer herrschenden Kammerdruck,
• • für die Rückätz-Kammern: • eine Angabe über die in die jeweilige Rückätzkammer eingespeiste Leistung, • eine Angabe über die von der jeweiligen Rückätzkammer reflektierte Leistung.

According to this embodiment, each machine 201 recorded per unit time, preferably per second, the following parameters:

• • a time indication;
• For a mainframe, if present, the pressure prevailing in the transfer chamber of the mainframe,
• • as well as for all process chambers: • a recipe step number of the recipe currently in the chamber,
• • for the separation chambers additionally: • an indication of the gas flow prevailing in the separation chamber, as well as • an indication of the chamber pressure prevailing in the separation chamber,
• • for the re-etching chambers: • an indication of the power fed into the respective re-etching chamber, • an indication of the power reflected by the respective re-etching chamber.

In the 12a to 12g are for the respective machines 401 . 402 . 403 . 404 . 405 . 406 . 407 the respective raw data parameters 1201 , the component types, 1202 on which the raw data parameter specification 1201 as well as the specification of the respective parameter sets 1203 to which the respective raw data parameter specification 1202 refers.

To explain in the 12a to 12g In the following, some basic aspects of the assignments and illustrations are briefly explained.

Thus, all the parameters that are required when configuring cross-departmental key figures have to be used for the first machine 401 for the second machine 402 the same raw data parameter names.

Because of this, the mainframe component points to the first machine 401 and the mainframe component of the second machine 402 to the same parameter set, namely the first parameter set 1204 (see. 12a and 12b ).

All the parameters that are necessary in the configuration of cross-chamber codes have according to this embodiment for the third machine 403 and the fourth machine 404 different raw data parameter names.

Because of this, the mainframe component references the third machine 403 to a second parameter set 1205 and the mainframe component of the fourth machine 404 to a third parameter set 1206 (see. 12c and 12d ).

The first process chamber A 904 the first machine 401 and the first process chamber A 914 the second machine 402 both are set up and use as deposition process chambers, ie log the same raw data parameter names. Thus, the first process chamber A of the first machine refer 401 and the first process chamber A of the second machine 402 to a same parameter set, namely to a fourth parameter set 1207 (see. 12a and 12b ).

The assignment of different raw data parameters to different parameter sets or the same raw data parameters to the same parameter sets and their corresponding renaming into a common parameter is machine-dependent and can be done manually or manually by a person skilled in the art.

The other contexts and the references to the other parameter sets are made in the same manner shown above and are therefore not explained in more detail.

This results in the corresponding assignment of the individual machine components to the individual parameter sets in 13 illustrated grouping table 1300 ,

According to this embodiment, the raw data parameter names for all machine types / machines are renamed as follows. Here, a distinction is made between process chamber-related raw data parameter names and process chamber-comprehensive raw data parameter names.

The process chamber related raw data parameter names are renamed as follows: • Time → TIME • recipe step identification, preferably Recipe step number of just in the process chamber expiring recipe → STEP, • the specification of the person in charge of the process chamber gas flow → gas, • the specification of the person in charge of the process chamber chamber pressure → PRINT, • the indication of the fed into the process chamber power → POWER FW, • the indication of the reflected from the process chamber power → POWER BW.

The process chamber-spanning raw data parameter names are mapped as follows: • Time → TIME, • recipe step identification, preferably Recipe step number of just in the process chamber expiring recipe → STEP_A, STEP_B, STEP_C, STEP_D, • the chamber pressure prevailing in the chamber → DRUCK_A, DRUCK_B, • the pressure prevailing in the transfer chamber → TRANSFER_PRINT.

This picture is in the tables in the 14 . 15 and 16 shown where the raw data parameter names 1401 of a parameter set 1400 to the APC parameter names shown above 1402 be imaged.

Thus put in the 14 . 15 and 16 represented parameter sets in each case the performed mapping of provided in the parameter sets Rohdatenparameternamen on the APC parameter names used, as set forth above, is.

In this way, a significant reduction of the parameter names used is achieved without information content being lost in the analysis of the raw data determined.

The processing of a wafer or simultaneously several wafers in a process chamber of a machine 401 . 402 . 403 . 404 . 405 . 406 . 407 usually runs in one or more process steps.

The process steps required for processing a wafer or simultaneously several wafers are combined into recipes.

Each process step, ie each recipe step, has a recipe step number which is uniquely assigned to the recipe step and characterizes it. Alternatively, the recipe step may also be assigned a description that uniquely identifies the recipe step.

The recipe number or the recipe description is written in the raw data file as parameter "STEP". In a process chamber of a machine 401 . 402 . 403 . 404 . 405 . 406 . 407 Run many different recipes, usually between 5 to 10 in a chip production plant 200 for the production of memory chips and usually 50 to 100 different recipes in a chip production plant 200 for the production of logic chips.

Similarities and differences are also determined for different recipes of a process chamber type and exploited according to the invention in order to achieve a further compression of the raw data.

It should be noted in this context that for different recipes different recipe names can be used.

17 shows in tables seven recipe sets 1701 . 1702 . 1703 . 1704 . 1705 . 1706 . 1707 as well as the assignment of the respective process chambers of the machines 401 . 402 . 403 . 404 . 405 . 406 . 407 to the recipes that are carried out in the respective process chambers.

The similarities of recipes are as generalized recipe steps according to the invention in the in 18 illustrated recipe table 1800 merged. The recipe step numbers or the recipe step descriptions are generally assigned to recipe steps and thus combined into recipe groups, ie a respective recipe set. It should be noted in this context that different recipe sets may have the same generalized recipe names, for example, Recipe_set_1, Recipe_set_2, ..., Recipe_set_7, and may include the generalized recipe steps, for example, "Bar", "Main", "Cool".

19 shows the recipe assignment to the respective recipe sets for all deposition process chambers of the machine group according to an embodiment of the invention.

This assignment can also be made automatically or manually by a person skilled in the art, knowing the particular recipes used and the properties of the machines and process chambers used.

19 shows the recipe configuration for all deposition process chambers of the machine group.

In the following, a part of the assignment and the principle underlying the assignment will be briefly explained for clarification.

For the first process chamber A of the first machine 401 is the "main step" of the recipe "Reci_1" the recipe step number 3, in the recipe "Reci_3" this is the recipe step number 4. In another machine, the main step of all recipes with "MAIN" is called. The assignment of which recipe step is the respective main step of the recipe can also be done automatically using appropriate prior knowledge or by a person skilled in the manual due to a physical interpretation of the respective process step.

About the in 19 shown allocation tables 1901 . 1902 . 1903 . 1904 . 1905 It is now possible to define key figures independent of prescriptions by referring to the generalized recipe steps. Using the recipe tables, that is the allocation tables 1901 . 1902 . 1903 . 1904 . 1905 For each recipe listed there, the generalized recipe step is stored clearly identifiable.

Different machines may have different recipes labeled differently.

Using the recipe tables 1901 . 1902 . 1903 . 1904 . 1905 According to the invention, it becomes possible to rename these recipe names in such a way that identical recipes which are carried out on different machines are designated in the same way; in general this means that descriptive meta-designations are introduced, also referred to as APC recipes.

The mapping and assignment of a machine-specific recipe to an APC recipe is in turn automatically possible or can be done by a physical interpretation of a professional.

The others in the allocation tables 1901 . 1902 . 1903 . 1904 . 1905 The steps illustrated are those of stabilizing the conditions in a process chamber ("STAB") and the phase of cooling the process chamber after the main step ("COOL") has been carried out.

In a similar way, the recipe configuration for in the 20 presented recipes, which are performed in the etching-back process chambers of the machine group.

The process steps performed and the corresponding mapping rules are in the assignment tables 2001 . 2002 shown. In particular, the process steps of starting the process chamber on process conditions ("RAMP"), stabilizing the process conditions in the process chamber ("STAB"), etching as the main step of the recipe running in the process chamber ("ETCH"), and retracting the process State of the process chamber to normal state ("RINSE") shown.

Thus, for etchback process chambers according to this embodiment, three generalized recipe steps are defined (STAB, MAIN, COOL) and four general recipe steps (RAMP, STAB, EDGE, RINSE) are provided for deposition process chambers.

• • Abscheide-Prozesskammern auf drei unterschiedliche APC-Rezeptnamen reduziert (gemäß diesem Ausführungsbeispiel erfolgt eine Reduktion von 8 maschinenspezifischen Rezeptnamen auf drei APC-Rezeptnamen): KOMP_3_REZEPT_1, KOMP_3_REZEPT_2, KOMP_3_REZEPT_3;
• • für Rückätz-Prozesskammern auf vier unterschiedliche APC-Rezeptnamen reduziert, nämlich KOMP_4_REZEPT_1, KOMP_4_REZEPT_2, KOMP_4_REZEPT_3, KOMP_4_REZEPT_4.

Many of the same recipes with different machine-specific recipe names were inventively for the

• • Separates deposition process chambers to three different APC recipe names (according to this embodiment, a reduction of 8 machine-specific recipe names to three APC recipe names takes place): KOMP_3_REZEPT_1, KOMP_3_REZEPT_2, KOMP_3_REZEPT_3;
• • For re-etching process chambers, reduced to four different APC recipe names, namely, KOMP_4_REZEPT_1, KOMP_4_REZEPT_2, KOMP_4_REZEPT_3, KOMP_4_REZEPT_4.

Thus, according to the invention, a considerable reduction in terms of the recipes and recipe names used is made possible. In this way, with a corresponding grouping, the machine components of the same type with the same parameter sets and the same recipe set can be combined to form a same set of code numbers that can be jointly supplied to an analysis designed in the same way.

The assignment of machines, machine types, machine components and machine component types, parameter sets, recipe sets to the respective sets of measures are according to this embodiment of the invention in a corresponding allocation table in 2100 in 21 shown.

22 shows the corresponding key figure configuration for machine component_1 and machine component_2 at two different mainframes.

Like the allocation table 2200 in 22 As can be seen, the respectively determined and grouped data are subjected to a statistical analysis, for example the check as to whether the values to be monitored are within a given admissibility interval of the monitored values or an average value determination of the determined values and the monitoring of the respective determined mean value with respect to the permissibility of the Value compared to given target criteria. In the allocation table 2200 is determined how from the raw data, ie the partial production data (data per time intervals) compressed key figures are generated, for example, the statistical mean of a parameter in a particular recipe step, the statistical dispersion of this parameter in a particular recipe step, etc.

It is also possible to determine a standard deviation for the values to be monitored and to compare the standard deviation with predefined target criteria.

If the determined values deviate from the target criteria in accordance with the algorithm used, then an alarm message can be generated by an additionally provided alarm generation unit and the user at the client computer 312 in which alarm generation unit further alarm generation criteria are stored, which are applied by the alarm generation unit.

Further metric configurations and the statistical analyzes are also for a re-etching process chamber in the configuration table 2300 in 23 and for a deposition process chamber in the configuration table 2400 in 24 shown.

The following is a brief summary of the achievable compression of production data:
According to the embodiment, seven machines are provided in the semiconductor production facility, which are grouped into one machine group. On each machine run according to this embodiment between four and seven different recipes, which in turn each have between three and four essential recipe steps. Between four and fourteen different raw data parameters are recorded and written down per machine.

• • Eine Reduktion der Parameter von 37 auf 13,
• • eine Reduktion der Rezeptnamen von 18 auf 7,
• • eine Reduktion der Rezeptschritte von 53 auf 7,
• • eine Reduktion von Kennzahlentabellen von 19 auf 3.

The different assignment tables of the configuration of key figures described above result in the following reduction:

• • A reduction of the parameters from 37 to 13,
• • a reduction of recipe names from 18 to 7,
• • a reduction of the recipe steps from 53 to 7,
• • a reduction of key figure tables from 19 to 3.

This leads to an overall reduction of 670,662 parameters to be detected and monitored (37 × 18 × 53 × 19) to now only 1,911 parameters to be monitored according to the invention (13 × 7 × 7 × 3).

For a typical semiconductor chip manufacturing plant, assuming that 50 raw data parameters are used in three different software versions, and 50 recipes with 10 to 15 essential recipe steps (for two different machine types, 50 completely different recipes) and 15 machines in a production plant (4 Process chamber machines as well as with two different process chambers) results in a reduction of data to be monitored of about 10,000: 1.

LIST OF REFERENCE NUMBERS

100

block diagram

101

Total manufacturing process

102

Front-end processes

103

Back-end processes

104

Support elements

105

Process technology and process integration

200

Semiconductor chip production installation

201

Semiconductor chip part production system / machine

202

lines

300

block diagram

301

sensor

302

SECS interface

303

PDSF file

304

log file

305

logistics data

306

Local communication network

307

Storage

308

APC filters and algorithms

309

Alarm / message sending unit

310

Database

311

block

312

Client computer

400

block diagram

401

First machine

402

Second machine

403

Third machine

404

Fourth machine

405

Fifth machine

406

Sixth machine

407

Seventh machine

408

SECS interface first machine

409

SECS interface second machine

410

SECS interface third machine

411

SECS interface fourth machine

412

SECS interface fifth machine

413

SECS interface sixth machine

414

SECS interface seventh machine

415

First personal computer

416

Second personal computer

417

Third personal computer

418

Fourth personal computer

419

Fifth personal computer

420

Sixth personal computer

421

Seventh personal computer

422

Raw data file server

423

First file

424

Second file

425

Third file

426

Fourth file

427

Fifth file

428

Sixth file

429

Seventh file

601

First column

602

Second column

603

Third column

604

Fourth column

605

Fifth column

606

Sixth column

701

time specification

702

Index column

703

Further column

704

Further column

705

Further column

706

Further column

707

Further column

800

machine

801

Main Frame

802

process chamber

803

process chamber

804

process chamber

805

process chamber

901

First charging station first machine

902

Second charging station first machine

903

Mainframe first machine

904

First process chamber first machine

905

Second process chamber first machine

906

Third process chamber first machine

907

Fourth process chamber first machine

911

First charging station second machine

912

Second charging station second machine

913

Mainframe second machine

914

First process chamber second machine

915

Second process chamber second machine

916

Third process chamber second machine

917

Fourth process chamber second machine

921

First charging station third machine

922

Second charging station third machine

923

Mainframe third machine

924

First process chamber third machine

925

Second process chamber third machine

931

First charging station fourth machine

932

Second charging station fourth machine

933

Mainframe fourth machine

934

First process chamber fourth machine

935

Second process chamber fourth machine

941

Charging station fifth machine

942

Process chamber fifth machine

951

Charging station sixth machine

952

Process chamber sixth machine

961

Charging station seventh machine

962

Process chamber seventh machine

1100

allocation table

1101

machine

1102

machine type

1103

machine components

1104

Machine components Type

1105

parameter set

1106

recipe set

1107

code set

1201

Raw data parameter names

1202

Machine components Type

1203

parameter set

1204

First parameter set

1205

Second parameter set

1206

Third parameter set

1207

Fourth parameter set

1300

grouping table

1400

parameter set

1401

Raw data Parameter name

1402

APC Parameter name

1701

First recipe set

1702

Second recipe set

1703

Third recipe set

1704

Fourth recipe set

1705

Fifth recipe set

1706

Sixth recipe set

1707

Seventh recipe set

1800

recipe table

1901

allocation table

1902

allocation table

1903

allocation table

1904

allocation table

1905

allocation table

2001

allocation table

2002

allocation table

2100

allocation table

2200

allocation table

2300

configuration table

2400

configuration table

Claims (9)

Method for collecting production data in a production plant with several partial production plants, • in which partial production data are determined and stored by the sub-production plants, In which, based on at least one grouping table under assignment of the partial production data to groups of the grouping table, the partial production data is mapped to compressed partial production data by using a statistical analysis for partial production data grouped according to the assignment of the partial production data to groups of the grouping table compressed key figures are generated, • where in the grouping table • machines of the sub-production plants are grouped into at least one machine group, or • Machine components of the sub-production plants are grouped into at least one machine component group, or • Manufacturing recipes are grouped into at least one recipe group, or • Manufacturing parameters are grouped into at least one parameter group. Method according to claim 1, • where partial production data of partial production facilities of a chip production facility are determined and stored, and • in the grouping table • Chip manufacturing recipes are grouped into at least one recipe group, or • Chip production parameters are grouped into at least one parameter group. Method according to Claim 1 or 2, in which the partial production data is transmitted from the partial production plants to a central store and stored there in at least one central file. Method according to one of claims 1 to 3, wherein the machines are grouped into groups of machines which have sufficiently similar machine characteristics. The method of any of claims 1 to 3, wherein the machine components are grouped into machine component groups having sufficiently similar machine component characteristics. Method according to one of claims 1 to 3, wherein the manufacturing recipes are grouped into recipe groups having sufficiently similar recipes. Method according to one of claims 1 to 3, wherein the production parameters are grouped into parameter groups having sufficiently similar production parameters. Method according to one of Claims 1 to 7, in which the compressed partial production data are subjected to a statistical analysis. System for recording production data in a production plant, comprising: • a plurality of partial production plants which generate partial production data • having at least one memory for storing the generated partial production data, • having an evaluation computer coupled to the memory, which is set up such that Based on at least one grouping table under assignment of the partial production data to groups of Grouping table the partial production data is mapped to compressed partial production data by generating compressed key figures by means of a statistical analysis for partial production data grouped according to the assignment of the partial production data to groups of the grouping table • in the grouping table • machines of the partial production plants are grouped into at least one machine group, or • machine components of the sub-production plants are grouped into at least one machine component group, or • manufacturing recipes are grouped into at least one recipe group, or • production parameters are grouped into at least one parameter group.

Images