US20010051836A1 - Fab yield enhancement system - Google Patents
Fab yield enhancement system Download PDFInfo
- Publication number
- US20010051836A1 US20010051836A1 US09/075,254 US7525498A US2001051836A1 US 20010051836 A1 US20010051836 A1 US 20010051836A1 US 7525498 A US7525498 A US 7525498A US 2001051836 A1 US2001051836 A1 US 2001051836A1
- Authority
- US
- United States
- Prior art keywords
- defect
- classification
- data
- global
- wafer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000007547 defect Effects 0.000 claims abstract description 281
- 238000012552 review Methods 0.000 claims abstract description 68
- 238000007689 inspection Methods 0.000 claims abstract description 55
- 238000000034 method Methods 0.000 claims description 83
- 230000008569 process Effects 0.000 claims description 55
- 238000004519 manufacturing process Methods 0.000 claims description 22
- 238000004458 analytical method Methods 0.000 claims description 19
- 230000009471 action Effects 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 9
- 239000004065 semiconductor Substances 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 7
- 230000000977 initiatory effect Effects 0.000 claims 1
- 238000012795 verification Methods 0.000 claims 1
- 235000012431 wafers Nutrition 0.000 description 47
- 239000002245 particle Substances 0.000 description 16
- 230000003287 optical effect Effects 0.000 description 14
- 238000012360 testing method Methods 0.000 description 13
- 238000013439 planning Methods 0.000 description 9
- 239000000463 material Substances 0.000 description 8
- 238000009826 distribution Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 238000013519 translation Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 239000011022 opal Substances 0.000 description 2
- 238000010998 test method Methods 0.000 description 2
- 125000003821 2-(trimethylsilyl)ethoxymethyl group Chemical group [H]C([H])([H])[Si](C([H])([H])[H])(C([H])([H])[H])C([H])([H])C(OC([H])([H])[*])([H])[H] 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000005137 deposition process Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 238000001878 scanning electron micrograph Methods 0.000 description 1
- 238000004626 scanning electron microscopy Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000026676 system process Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
- H01L22/20—Sequence of activities consisting of a plurality of measurements, corrections, marking or sorting steps
Definitions
- the present invention relates to semiconductor fabrication systems and, more specifically, to a system for identifying the sources of wafer defects and using this information to control the processes in a fabrication plant.
- Semiconductor fabrication involves processing a piece of a semiconductor material (referred to as a wafer) to create electronic circuits such as transistors, resistors and inductors on the wafer. These circuits are formed by depositing material on the wafer, doping areas of the semiconductor with other elements and etching the wafer. These processes also create the connections between the circuits.
- a wafer a piece of a semiconductor material
- electronic circuits such as transistors, resistors and inductors on the wafer. These circuits are formed by depositing material on the wafer, doping areas of the semiconductor with other elements and etching the wafer. These processes also create the connections between the circuits.
- the present invention provides an automated yield enhancement system that identifies sources of defects on wafers and uses this information to control the fabrication process.
- the system collects defect information from the tools in the FAB and produces a global set of defect classes. Throughout the fabrication process, the system continually updates and reclassifies the global classification based on the current and prior defect data.
- the system uses the defect classification information and other FAB data to automatically identifies the sources of the defects and refines the analysis procedures used by the FAB inspection and review tools.
- the system may use the defect information and other FAB data to automatically adjust the operation of the FAB process tools to prevent similar defects from being formed on subsequently produced wafers.
- each defect is classified according to its characteristics. Preferably, these characteristics bear on the source or the severity of the defect.
- a system constructed according to the invention provides a single, global classification scheme for defect information that may originate from a variety of tools that may otherwise be unrelated.
- the system stores the classification data and other information related to the defects in a defect database.
- This information may include classification identifiers, defect source candidates and parameters that represent a relative confidence level that the identified candidates are indeed the source of the defect.
- the system may store optical review data, SEM information and wafer defect maps that are associated with the defects.
- the information for the defects may be compared with corresponding information in the defect database. These classifications are continually refined and updated based on the defect information and other inputs from the FAB such as parametric data generated by tests conducted on the wafer.
- the system processes the classification data generated by the yield enhancement system and defect source data provided by FAB tool vendors. Once the potential source of a defect is identified, the system may use this information to modify the inspection and review plans for the optical inspection and review tools, the SEM inspection and review tools or any other inspection, review or metrology tools. In addition, when the source of the defect is associated with a process tool that may be controlled by the yield enhancement system, the system may automatically adjust the process tool to eliminate the defect in subsequently fabricated wafers.
- the system may identify the sources of the defects more accurately than conventional systems. Moreover, by integrating defect information from all of the FAB tools, the system may quickly identify the source of a defect and automatically correct the problem or enable the FAB operator to do so.
- FIG. 1 is a block diagram of one embodiment of a FAB including a yield enhancement system constructed according to the invention
- FIG. 2 is a block diagram of a FAB process flow for one embodiment of a yield enhancement system constructed according to the invention
- FIG. 3 is a flowchart illustrating defect processing operations in one embodiment of a yield enhancement system constructed according to the invention.
- FIG. 4 is a flowchart illustrating a review plan refinement process in one embodiment of a yield enhancement system constructed according to the invention.
- a yield enhancement system 18 in a FAB F processes defect data and other information generated by a FAB host 20 and several FAB tools 22 , 24 and 26 .
- the FAB host generates work-in-progress (“WIP”) data that provides information about the wafer currently being processed.
- the inspection tool 22 automatically generates wafer defect maps and, optionally, automatic defect classification data and other information related to defects on the wafer.
- the automatic defect classification data may include, for example, a classified wafer defect map, defect class data and defect attributes.
- the inspection tool 22 can be, for example, one or more of the “ORBOT WF 7xx” series equipment available from Applied Materials of Santa Clara, Calif.
- the review tool 24 may be used to review the wafer according to the defect map.
- the review tool 24 may generate images of the defects and may provide additional automatic defect classification data and other defect information.
- the review tool 24 can be, for example, an “OPAL 9300” Defect Review SEM available from Applied Materials of Santa Clara, Calif. Irrespective of the defect inspection and review, the test tool 26 performs electrical tests on the wafer. These tests produce information that relates to the actual operating parameters of the wafer circuits.
- the yield enhancement system 18 includes several components that cooperate to identify the sources of the wafer defects and improve the wafer inspection process.
- An integrated automatic defect classification (“IADC”) component 28 integrates the information provided by the tools into a global classification scheme. It then refines this information in an attempt to generate the best defect classifications assignments for the current production process.
- IADC integrated automatic defect classification
- a defect knowledge database 30 accumulates defect data generated by the process tools and the yield enhancement system 18 .
- the defects database 30 provides a reference database of known defects, their characteristics, and their sources.
- An automatic defect source identification system 32 ties together all the information from the IADC and the defect database. The identification system 32 then applies intelligent decision algorithms to this information to identify the sources of the defects. In addition, the identification system 32 may provide defect warnings, suggested test procedures to validate the sources of the defects, and suggested corrective actions to fix the problems.
- a smart sample planner 34 uses the classification data and the accumulated knowledge in the system to optimize the FAB review, the inspection strategy and the overall sampling strategy of the FAB.
- FIG. 2 describes a process flow in a FAB that incorporates a yield enhancement system according to the present invention.
- FIG. 3 describes data flow associated with an IADC.
- FIG. 4 describes data flow for one embodiment of yield control planning and sample planning processes operative according to the invention.
- To reduce the complexity of the drawings only one inspection tool 22 , review tool 24 and test tool 26 may be illustrated in some of the drawings. It should be appreciated, however, that a given FAB may include more than one inspection tool 22 , review tool 24 and/or test tool 26 .
- a wafer 40 is sequentially fed to an automated defect inspection tool 22 and, optionally, to an optical defects review tool 24 A and a SEM defects review tool 24 B as represented by dashed line 41 .
- the tools send defect information to the yield enhancement system 18 .
- the yield enhancement system 18 classifies the defect data, stores it in the defect database 30 and, as necessary, provides updated review procedures to the review tools 24 A and 24 B.
- the inspection tool 22 generates inspection data including defect maps 44 and, optionally, classification attributes and classification results.
- the classification attributes refer to intermediate level data. That is, the results of tool sensing or image processing. Tool sensing may involve, for example, analyzing the wafer with a tool that uses a light source and a light sensor (e.g., a charge-coupled device). These results may be characterized as a collection of attributes such as size, location, shape, color, position on the layer and sensing signal characteristics (e.g., signals from the tool sensing that are proportional to the intensity of the detected light).
- the classification results are class identifiers.
- Each automatic defect classification tool (e.g., 22 , 24 A or 24 B) may define its own set of specific identifiers and associated classes.
- the inspection tool 22 sends its defect information to the IADC 28 .
- the IADC analyzes this information and identifies clusters of defects, repeating defects and overlay defects.
- the IADC calculates a spatial signature analysis for the defects.
- the IADC 28 uses the defect classification data to generate alarms and warnings that are based on control limits for defect levels, defect distributions and other criteria.
- the IADC generates a review plan for an optical review station 24 A, a SEM review station 24 B or any other review tools, as necessary.
- the wafer 40 may be sent to the optical defects review tool 24 A.
- the review tool 24 A generates a classified defects map 46 , optical images and, optionally, optical-based automatic defect classifications (not shown).
- the defect classifications may include classification attributes and classification results.
- the IADC 28 processes this information to obtain global defect classifications. To this end, the IADC may refine the data using data stored in the database 30 . In addition, the IADC reclassifies the incoming defect information to conform to the global classifications that the IADC has generated for all defects in the FAB. Finally, the IADC assigns a confidence level to the defect and stores this information with the defect in the defect database 30 .
- defects are classified in the global classification scheme based on similar spatial signature analysis results and similar on-the-fly automatic defect classification results.
- the images may be associated with the defect data per the defect classification scheme.
- the defects are presented to the FAB operator as a single, uniform set of defect classes.
- the IADC As above, based on the defect classifications, the IADC generates alarms and warnings. In addition, if needed, the IADC generates a SEM review plan.
- the wafer 40 is sent to the SEM defects review tool 24 B.
- the review tool 24 B generates a re-classified defect map (not shown), SEM defect images and, optionally, SEM-based automatic defect classifications and, if needed, composition data such as EDX.
- the IADC 28 processes the SEM information to further refine and globalize the defect information and assign new confidence levels to the results. Finally, the IADC generates alarms and warnings, as necessary, and stores the classified defect map in the database 30 .
- FIG. 3 describes the basic operations of the IADC 28 .
- the IADC 28 receives defect data from the inspection and review tools 22 , 24 A and 24 B.
- the IADC 28 classifies the defects based on the defect data and data from spatial signature analysis 56 of the wafer defects.
- the IADC stores all of the data related to the defects in the database 30 .
- the defect data is readily available for future classification searches 58 and 60 .
- the IADC 28 has “learning” capabilities in that it continually refines the defect classifications and defect source data based on received defect data.
- the defect source identification system 32 uses the defect data to identify the source of the defects and initiate corrective action.
- the IADC 28 aligns and consolidates the defect data from all the local automatic defect classification (“ADC”) tools to produce unified classification.
- Each of ADC tools depicted in FIG. 3 analyzes detected defects and generates defect attribute data.
- the tool 22 inspects the wafer and classifies detected defects.
- the tool 22 performs on-the-fly automatic defect classification.
- Tool 22 then sends the local automatic defect classification attribute data (e.g., intensity of scattered light, etc.) to the IADC 28 as represented by line 66 .
- the IADC processes the attribute data from the tools (line 68 ) along with spatial signature analysis information (line 70 ) and data from an ADC learning system 64 (via line 65 ) to generate integrated attribute classifications.
- the spatial signature analysis information 70 consists of a conventional set of algorithms that describe a zonal analysis of defects distributed across the wafer.
- the ADC learning system 64 is described in more detail below.
- Each of the ADC tools depicted in FIG. 3 also processes the attribute data to classify the defects into local, tool-specific classes as mentioned above.
- the tool 22 provides local automatic defect classification results (e.g., a classified defect map) to the IADC 28 as represented by line 72 .
- the IADC 28 uses the class data from the tools (line 74 ) and the integrated attributes (line 76 ) to generate globalized classifications 52 .
- the various class identifiers from the FAB are translated into a uniform, global set of classes.
- the IADC 28 stores all of the classification parameters related to the defects in the database 30 . Specifically, the IADC stores the refined classification data (line 78 ) and the integrated attributes classification information (line 80 ) in the defect knowledge database 30 . This includes automatic defect classification attribute data (block 60 ) that may be accessed during knowledge-based searches of the database 30 .
- the IADC 28 also stores defect image and composition data (line 82 ) in the database 30 .
- This information originates from the optical tool 24 A and the SEM tool 24 B in the form of raw images and material identification data (line 84 ).
- the IADC uses this information during knowledge-based searches 58 of the defect database 30 as discussed below.
- the IADC 28 refines and corrects the global classifications (line 86 ) and the integrated attributes classification (line 87 ) to provide enhanced classification results (block 54 ). By combining the data from the different local classifications, the IADC can provide a more accurate classification than the local classifications.
- the classification refinement operation may take several forms.
- the IADC 28 may combine the results of specific defect classifications (e.g., inspection based ADC 22 or optical review ADC 24 A) and spatial defects distribution classification (e.g., spatial signature analysis 56 ) to refine the classification and increase it's accuracy.
- specific defect classifications e.g., inspection based ADC 22 or optical review ADC 24 A
- spatial defects distribution classification e.g., spatial signature analysis 56
- One example for providing the global classification is by using weighting and probability factors. That is, each tool sends classification and level of confidence data to the IADC 28 . If a conflict arises, the IADC 28 can apply the level of confidence to the classification of each tool to arrive at a final determination. Thus, for tools 1 and 2 , classifications X 1 and X 2 , and levels of confidence P 1 and P 2 , the IADC 28 can perform P 1 (X 1 )+P 2 (X 2 ) ⁇ X G , where the classification of X G can be determined from a look-up table or using a threshold.
- the IADC 28 can give preferential treatment to certain tools (e.g., a higher weight may be given to a high sensitivity review tool than to a fast scan inspection tool).
- the weights can be specific to a tool and can be individualized per class of defects. As the system learns the sensitivity of the tools to the specific processing lines (using prior correct/incorrect classifications), the weights can be adjusted accordingly. It should be appreciated that for a certain tool/class, the weight can be set to zero to ignore the classification provided by that tool.
- the IADC 28 may be integrated into the classification chain depicted in FIG. 2.
- the classification is refined after each stage in the chain. That is, the classification is refined after the inspection based ADC 22 . It is refined again after the optical review ADC 24 A and once again after the SEM ADC 24 B.
- the IADC 28 may integrate the above chain into a learning system that combines all past defect results and spatial signature analyses.
- the results of a knowledge-based classification are processed (line 88 ) by the enhanced ADC component 62 .
- the component 62 generates a classification history which may include the original input data (classification attributes), the original classification data, the intermediate refinement data and the final classification data.
- the classification history data is used (line 90 ) by the ADC learning system 64 to assign and refine confidence levels and weights for future ADC results and to optimize the integrated ADC operation for faster and more accurate classifications.
- the learning system 64 uses information from prior defects to continually refine the classifications and the assignment of defects to those classifications.
- the IADC 28 receives defect information from one of the ADC tools.
- the defect may be, for example, a foreign particle on the surface of the wafer where the particle has a particular size and shape.
- the received defect information may initially only include the local class number (e.g., the number 25 ) that the ADC tool assigns to this type of defect.
- the defect information also includes a confidence level determined by the inspection tool.
- the next step in the process may depend on the reported confidence level or the confidence level that the IADC assigned to the local defect classification. For example, when there is a high level of confidence that the ADC tool accurately identifies this class of defect, the IADC may simply translate the local class into a global class. A high level of confidence for a given local class may be indicated by a high confidence value relative to the confidence level of other classes assigned to this defect by other ADC processes (e.g., 90% vs. 10%).
- the translation process may simply involve changing the class number.
- the IADC must fit the class into an existing global class or generate a new global class for this type of defect.
- the IADC 28 may perform additional steps to calculate an appropriate global class. This may be the case, for example, when two ADC tools assign different classes to the defect and each class has approximately the same confidence level.
- the IADC 28 may request the defect attributes (e.g., sensing attributes such as intensity level or defect attributes such as whether it is a particle; its size: e.g., less than 0.1 microns; and its shape: e.g., approximately round) from the ADC tools and assign a class based the corresponding defect attributes 50 and/or on spatial signature analysis information 56 associated with the defect. This may involve, for example, comparing the defect attributes with the attributes associated with the global classes.
- the defect attributes e.g., sensing attributes such as intensity level or defect attributes such as whether it is a particle; its size: e.g., less than 0.1 microns; and its shape: e.g., approximately round
- the IADC 28 will assign an initial confidence level to this class.
- the IADC may assign a confidence level by combining algorithmic approaches with past empirical data. For example, the confidence level may be based on the relative number (e.g., percentage) of matches (e.g., size, shape, density) between the class attributes and the attributes associated with the defect. The class and/or the confidence level may then be refined by subsequent ADC processes.
- ADC learning system 64 may be used to control the ADC tools to improve the inspection and review process in the FAB.
- Some ADC tools have variable defect thresholds or decision criteria.
- an ADC tool may be configured so that only signals above a certain intensity would be reported as defects.
- the ADC learning system 64 may send messages to the appropriate ADC tools to change their threshold settings to result in a proper classification in future inspections.
- the learning system 64 may send a message to a display device 92 that informs the FAB operator that the thresholds of the ADC tools should be changed.
- the IADC 28 may attempt to classify the defect by performing a database search.
- the IADC will compare the attributes, the signature analysis, the defect maps, the image data and other information associated with the defect with the knowledge search information (e.g., 58 and 60 ) stored in the database 30 .
- the selected global class and confidence levels may be based on the relative percentage of matches between the information associated with the defect and the information stored in the database 30 .
- the IADC 28 uses the defect classification data and other FAB data such as overlay, yield and kill ratio information to generate (line 94 ) the best classification possible for a given production line.
- This classification may include a class identifier for each defect and potential alternative classes in case of inconclusive class identification.
- the IADC typically produces a confidence level for each of the classifications.
- the database 30 stores information that is used in the defect classification process, in the defect classification refinement process and in the defect source identification process.
- the database 30 stores all of the defect data including the defect attributes, spatial signatures, the local and global classes, confidence levels, defect densities, descriptive texts of the events, wafer maps and images.
- Database 30 may also store manufacturer-specific data. For example, a manufacturer of an etch chamber may provide specific data related to defects caused by corrosion in the chamber. Such data can help in pointing out the specific chamber causing a specific defect.
- a yield engineer can access this data to determine how a specific classification was assigned. Moreover, the yield engineer may perform manual classifications whereby the engineer manually “teaches” the system to tie specific parameters to a specific class, thereby customizing and improving the classification results for a particular FAB and application. This procedure or a similar procedure may be used at a later date by the system or by the FAB engineer to assign a class to other defects. For example, the yield enhancement system may be programmed to follow similar procedures to perform automatic defect classification.
- the IADC 28 also uses the information in the defects database 30 to validate the algorithmic classifications on an ongoing basis. Thus, over time, the accuracy of the algorithmic classifications will increase and the chance of erroneous or fruitless classifications will decrease. Moreover, it should be appreciated that the database 30 may initially be configured without any defect information. In this case, the yield enhancement system 18 can add defect information to the database as it acquires defect information from the FAB and/or the operator may input information that correlates the defects with potential sources, or with manufacturer-specific data as it is received at the FAB.
- the IADC 28 can correlate yield results to specific defects and defect classes. As a result, the IADC may refine the identification of killer defects. This may, in turn, enable the yield engineer to ignore a significant percentage of nuisance and redundant defects, while concentrating on avoiding killer defects.
- the identification system 32 uses the information from the IADC (line 94 ) and the database 30 to identify potential sources for a given defect or a class.
- the identification system 32 may use the defect source information stored in the database 30 .
- This information typically would be in the form: machine “x” may have problem “y” and this will show up as a defect of class “z.”
- the defect source identification system 32 may also perform more complicated defect root causes analysis.
- the identification system 32 may identify a defect source using the specific defect and defect class identified by the inspection and review tools 22 and 24 and the local and global classification data and associated source information from the IADC 28 and the database 30 .
- a class that covers missing or extra signal path patterns may relate to a problem in the photo-resist process.
- the source of this problem may be the introduction of a foreign particle prior to the etch process.
- an embedded particle may indicate a problem before the deposition process.
- Other inputs to the identification system 32 may include work-in-progress (“WIP”) data and yield data from the FAB host system 20 and the accumulated statistical probability data calculated by the yield enhancement system.
- WIP work-in-progress
- the defect source identification system 32 also generates automatic alarms and warnings associated with the class identifier(s) as configured by the FAB.
- the yield enhancement system 18 can provide class density monitoring rather than conventional total density monitoring. Because prior high defect events may be stored on a per class basis, the yield enhancement system 18 can generate alarms based on a threshold for specific classes. This is in contrast to many conventional systems that generate alarms based on a threshold for the aggregate of the defects. As a result, using a yield enhancement system constructed according to the invention, the yield engineer can rapidly identify developing excursions in a given class that would otherwise be insignificant (and hence relatively difficult to identify) in relation to the aggregate of the defects.
- the defect source identification system 32 may generate suggested test procedures to verify the identification of the source of the defect. For example, the identification system 32 may identify a given sub-process of a process tool (e.g., an incorrect temperature at a given etcher stage) as a possible source of a defect. If that process tool (e.g., the etcher) has a built-in self-test for that sub-process, the identification system 32 may automatically send a message to the process tool (using the WIP data that identifies the etcher number) requesting the process tool to run a self-test. In addition, the identification system 32 also requests the process tool to send back the results of the test to the identification system. This may be done, for example, by sending an appropriate message to the process tool or tools 96 . If that sub-process fails (e.g., the temperature is incorrect), the identification system 32 may increase the level of confidence that that sub-process is the cause of the defect, as appropriate.
- a given sub-process of a process tool e.g., an incorrect temperature at
- the defect source identification system 32 may initiate corrective procedures to fix the problem.
- the etcher may have an interface that allows a remote device to adjust the temperature by sending an appropriate command.
- the identification system 32 may be configured to send this command in the event the self-test failed.
- the identification system may be configured to display a message on the display device 92 that informs the FAB engineer to check parameter “x” (e.g., temperature) on machine “y” (e.g., etcher number 4 ).
- FIG. 4 describes yield control planning and sample planning operations in accordance with one embodiment of the invention.
- a yield modeling and planning component 108 generates a yield control and strategy plan 110 .
- an initial plan is developed when the FAB is set up based on simulation modeling of the performance and interaction of the equipment installed in the FAB.
- a smart sample planner 34 continually refines the yield control plan (e.g., the wafer inspection and review plans) based on in-line information such as WIP data from the FAB equipment, actual wafer yield data and ADC data as discussed above.
- design-based yield modeling incorporates mathematical modeling and simulation algorithms to generate (line 112 ) predictions of yield sensitivities for each layer and for particular areas on the chip. Predictions may also be generated for potential killer defects sizes, distributions, and quantities.
- Process tool-based yield modeling uses historical data and statistical analysis and prediction to generate (line 114 ) predictions of defects for a set of process tools.
- the modeling and planning component 108 From this information and the results from process interaction modeling (line 116 ), the modeling and planning component 108 generates yield predictions for the entire line and the specific production modules (lines 118 and 120 ). In addition, the component 108 generates yield thresholds for each process module and for each process tool in the FAB (line 122 ). Finally, the yield control component 110 generates the yield control plan (line 124 ) for the entire production line. This plan may define which tools are used in the FAB. It may define how to position the tools and how to use them. In addition, the plan may specify which wafers are to be inspected.
- the sample planner 34 uses the in-line defect classification data discussed above and other accumulated knowledge in the system to improve the effectiveness of the yield control plan. By continually refining the yield control plan, the sample planner 34 will enable the yield engineer to more effectively focus on killer defects and defects of interest. As a result, this may reduce the amount of time the yield engineer spends reviewing defects on each wafer and reduce the number of test wafer that must be wasted when setting up a wafer process the FAB.
- the sample planner 34 generates a preliminary inspection and review plan (line 126 ) using the yield control plan and an enhanced set of initial planning rules (line 128 ).
- the initial sample plan is optimized using efficiency estimations (line 130 ) to produce the wafer inspection plan (line 132 , e.g., which wafer to inspect, when to inspect it and how to inspect it), an optical review plan (line 134 ) and a SEM review plan (line 136 ).
- the sample planner uses the above information to generate inspection recipes that are tuned and optimized for the specific layer inspected.
- the sample planner provides optimized ADC, by an inspection tool (re-visited) and/or by a review tool.
- the sample planner also may provide an optimized sampling plan for the review tool.
- the sample planner may be programmed to “learn” which type of defects warrant further review and which type of defects may be skipped during the review process. For example, if a hundred particles are located at a specific location on the wafer, it may only be necessary to review a few (e.g., five) of these particles with the review tool. This may be the case where there is a relatively high level of confidence in the defect information provided by an inspection tool that indicates that all one hundred of the defects are similar particles. This high level of confidence may result from, for example, prior occurrences of similar defects. Thus, once the system is tuned and there is confidence in the inspection tool, this information may be programmed into the sampling plan so that the review tool automatically reviews just five of the one hundred defects.
- the inspection and review information is then fed back into the yield enhancement system 18 (line 138 ).
- the fed back information is used for efficiency evaluations and to generate proposed enhancements to the inspection and review plans and for an efficiency evaluation of the original inspection plan (line 140 ).
- These enhancements are based on the inspections and review direct results (defects of interest) and on the historical correlation of the defects to actual yield (test data).
- the inputs to the efficiency estimator include layer and process tools data acquired from the FAB host (line 142 ). This may include, for example, metrology data such as the state of a process tool.
- the inputs may also include killer defect identification information that is based on past correlation with actual yield data (line 144 ).
- defects classification information (line 148 ) and information about the defects distribution on the wafers that is obtained from spatial signature analysis (line 148 ) may also be fed into the yield enhancement system 18 .
- the sample planner uses the layer and tool information to decrease inspection and review time and improve inspection results. This results in reduced test wafer consumption in the FAB and reduces throughput time for production material, reduces capital expenditures for inspection/review/metrology tools and saves FAB floor space.
- the automatic yield enhancement system shortens the time to identify the sources of yield excursions, shortens the learning time of new defects, and improves the productivity of the wafer inspection and review tools by optimizing their use. These goals are achieved through improved data management; integrating several types of FAB data and utilizing new software algorithms to automate and speed up the decision process to monitor and control FAB yield.
- the system is a learning system, accumulating data in both automatic and manual modes, and improving the accuracy of automatic decisions based on that accumulated data.
- the productivity of the inspections and review tools also improved by the sampling planning module. This module uses the integrated data to provide improved plans for using the tools.
- the system may be deployed across the FAB, as an overall yield enhancement system. Alternatively, it may be deployed in specific areas of the FAB to provide yield control for a specific group of process, inspection, and metrology tools. For example, in a fully automated system (manufacturing cells, short loop control), the validation and corrective action may be communicated automatically to the process tool, thereby prompting it to perform automatic service routines. In this embodiment, the system uses the service routines supplied by the process tool manufacturer and complies with the protocol that activates them automatically.
- defect knowledge database can be used for defect source identification irrespective, or in addition to, defect classification.
- inspection tool 22 (FIG. 1) can send a defect map to the automatic yield enhancement system 18 .
- certain or all of the defects may be reviewed in the optical and/or SEM review tools ( 24 A, 24 B, FIG. 2).
- a material (EDX) analysis can be performed using, for example, a SEM review station such as the Opal 9300 in an EDX mode.
- the defect knowledge database includes defect information, such as manufacturer specific and previous analyses results
- a search for a match can be performed using the optical and/or SEM images and the EDX results.
- a good match can obviate further processing and point directly at the defect source and proposed corrective action.
- particles can come from different sources and different parts of the processing chambers.
- the corrective action is also stored with a link to the particle information. Then, using a material composition match, the appropriate chamber can be identified and the stored corrective action provided as a probable solution for the problem.
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Computer Hardware Design (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Power Engineering (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
Description
- The present invention relates to semiconductor fabrication systems and, more specifically, to a system for identifying the sources of wafer defects and using this information to control the processes in a fabrication plant.
- Semiconductor fabrication involves processing a piece of a semiconductor material (referred to as a wafer) to create electronic circuits such as transistors, resistors and inductors on the wafer. These circuits are formed by depositing material on the wafer, doping areas of the semiconductor with other elements and etching the wafer. These processes also create the connections between the circuits.
- Although a wafer is relatively large (e.g., several inches in diameter), the circuits are quite small. For example, state of the art fabrication processes can produce a circuit having a feature size in the sub-micron range. As a result, the error tolerance of fabrication systems is very small. Consequently, even the slightest error in the fabrication process can cause a significant number of defects in the processed wafer.
- The incentive to reduce the number of defects in the wafer production process is compounded by the cost the fabrication plant (“FAB”). The cost of building a state-of-the-art FAB typically exceeds one billion dollars. As a result, it is imperative that the FAB yield be as high as possible to maximize the return on this investment.
- The need to increase the yield in the FAB has led to the development of a variety of tools that identify defects in a wafer. Some tools enable a FAB operator to visually inspect the wafer (or a representation of the wafer) to locate defects. Typical tools of this type include optical inspection systems and scanning electron microscopes (“SEMs”). Other tools provide defect information in the form of data. Several of these tools group the defect data into classes. Given that similar defects may be caused by similar sources, this technique allows a FAB operator to more easily identify certain sources of defects.
- In sum, conventional FAB tools may provide a plethora of FAB defect information. However, many of these tools do little more than generate the information and leave it to the operator to sort, classify and interpret the data. Thus, a need exists for a FAB yield enhancement system that can interpret this data and provide it to an operator in manner that enables the operator to efficiently locate the sources of the defects.
- The present invention provides an automated yield enhancement system that identifies sources of defects on wafers and uses this information to control the fabrication process. The system collects defect information from the tools in the FAB and produces a global set of defect classes. Throughout the fabrication process, the system continually updates and reclassifies the global classification based on the current and prior defect data. Using the defect classification information and other FAB data, the system automatically identifies the sources of the defects and refines the analysis procedures used by the FAB inspection and review tools. In addition, the system may use the defect information and other FAB data to automatically adjust the operation of the FAB process tools to prevent similar defects from being formed on subsequently produced wafers.
- In the global classification scheme, each defect is classified according to its characteristics. Preferably, these characteristics bear on the source or the severity of the defect. In contrast with conventional systems, a system constructed according to the invention provides a single, global classification scheme for defect information that may originate from a variety of tools that may otherwise be unrelated.
- The system stores the classification data and other information related to the defects in a defect database. This information may include classification identifiers, defect source candidates and parameters that represent a relative confidence level that the identified candidates are indeed the source of the defect. In addition, the system may store optical review data, SEM information and wafer defect maps that are associated with the defects.
- As new defects are processed, the information for the defects may be compared with corresponding information in the defect database. These classifications are continually refined and updated based on the defect information and other inputs from the FAB such as parametric data generated by tests conducted on the wafer.
- To identify the source of a defect, the system processes the classification data generated by the yield enhancement system and defect source data provided by FAB tool vendors. Once the potential source of a defect is identified, the system may use this information to modify the inspection and review plans for the optical inspection and review tools, the SEM inspection and review tools or any other inspection, review or metrology tools. In addition, when the source of the defect is associated with a process tool that may be controlled by the yield enhancement system, the system may automatically adjust the process tool to eliminate the defect in subsequently fabricated wafers.
- By continually updating the defect information and the review tool procedures based on accumulated data, the system may identify the sources of the defects more accurately than conventional systems. Moreover, by integrating defect information from all of the FAB tools, the system may quickly identify the source of a defect and automatically correct the problem or enable the FAB operator to do so.
- These and other features of the invention will become apparent from the following description and claims, when taken with the accompanying drawings, wherein similar references characters refer to similar elements throughout and in which:
- FIG. 1 is a block diagram of one embodiment of a FAB including a yield enhancement system constructed according to the invention;
- FIG. 2 is a block diagram of a FAB process flow for one embodiment of a yield enhancement system constructed according to the invention;
- FIG. 3 is a flowchart illustrating defect processing operations in one embodiment of a yield enhancement system constructed according to the invention; and
- FIG. 4 is a flowchart illustrating a review plan refinement process in one embodiment of a yield enhancement system constructed according to the invention.
- In FIG. 1, a
yield enhancement system 18 in a FAB F processes defect data and other information generated by aFAB host 20 andseveral FAB tools inspection tool 22 automatically generates wafer defect maps and, optionally, automatic defect classification data and other information related to defects on the wafer. The automatic defect classification data may include, for example, a classified wafer defect map, defect class data and defect attributes. Theinspection tool 22 can be, for example, one or more of the “ORBOT WF 7xx” series equipment available from Applied Materials of Santa Clara, Calif. Thereview tool 24 may be used to review the wafer according to the defect map. If used, thereview tool 24 may generate images of the defects and may provide additional automatic defect classification data and other defect information. Thereview tool 24 can be, for example, an “OPAL 9300” Defect Review SEM available from Applied Materials of Santa Clara, Calif. Irrespective of the defect inspection and review, thetest tool 26 performs electrical tests on the wafer. These tests produce information that relates to the actual operating parameters of the wafer circuits. - The
yield enhancement system 18 includes several components that cooperate to identify the sources of the wafer defects and improve the wafer inspection process. An integrated automatic defect classification (“IADC”)component 28 integrates the information provided by the tools into a global classification scheme. It then refines this information in an attempt to generate the best defect classifications assignments for the current production process. - A
defect knowledge database 30 accumulates defect data generated by the process tools and theyield enhancement system 18. Thedefects database 30 provides a reference database of known defects, their characteristics, and their sources. - An automatic defect
source identification system 32 ties together all the information from the IADC and the defect database. Theidentification system 32 then applies intelligent decision algorithms to this information to identify the sources of the defects. In addition, theidentification system 32 may provide defect warnings, suggested test procedures to validate the sources of the defects, and suggested corrective actions to fix the problems. - A
smart sample planner 34 uses the classification data and the accumulated knowledge in the system to optimize the FAB review, the inspection strategy and the overall sampling strategy of the FAB. - With the above overview in mind, the details of the system will now be discussed in more detail in conjunction with FIGS.2-4. FIG. 2 describes a process flow in a FAB that incorporates a yield enhancement system according to the present invention. FIG. 3 describes data flow associated with an IADC. FIG. 4 describes data flow for one embodiment of yield control planning and sample planning processes operative according to the invention. To reduce the complexity of the drawings, only one
inspection tool 22,review tool 24 andtest tool 26 may be illustrated in some of the drawings. It should be appreciated, however, that a given FAB may include more than oneinspection tool 22,review tool 24 and/ortest tool 26. - Referring now to FIG. 2, a
wafer 40 is sequentially fed to an automateddefect inspection tool 22 and, optionally, to an optical defects reviewtool 24A and a SEM defects reviewtool 24B as represented by dashedline 41. At each stage of the inspection and review process, the tools send defect information to theyield enhancement system 18. Theyield enhancement system 18 classifies the defect data, stores it in thedefect database 30 and, as necessary, provides updated review procedures to thereview tools - The
inspection tool 22 generates inspection data including defect maps 44 and, optionally, classification attributes and classification results. The classification attributes refer to intermediate level data. That is, the results of tool sensing or image processing. Tool sensing may involve, for example, analyzing the wafer with a tool that uses a light source and a light sensor (e.g., a charge-coupled device). These results may be characterized as a collection of attributes such as size, location, shape, color, position on the layer and sensing signal characteristics (e.g., signals from the tool sensing that are proportional to the intensity of the detected light). The classification results are class identifiers. Each automatic defect classification tool (e.g., 22, 24A or 24B) may define its own set of specific identifiers and associated classes. - The
inspection tool 22 sends its defect information to theIADC 28. The IADC analyzes this information and identifies clusters of defects, repeating defects and overlay defects. In addition, the IADC calculates a spatial signature analysis for the defects. - Using the defect classification data, the
IADC 28 generates alarms and warnings that are based on control limits for defect levels, defect distributions and other criteria. In addition, the IADC generates a review plan for anoptical review station 24A, aSEM review station 24B or any other review tools, as necessary. - If the
wafer 40 contains defects that need to be studied further, thewafer 40 may be sent to the optical defects reviewtool 24A. Thereview tool 24A generates a classified defects map 46, optical images and, optionally, optical-based automatic defect classifications (not shown). Again, the defect classifications may include classification attributes and classification results. - The
IADC 28 processes this information to obtain global defect classifications. To this end, the IADC may refine the data using data stored in thedatabase 30. In addition, the IADC reclassifies the incoming defect information to conform to the global classifications that the IADC has generated for all defects in the FAB. Finally, the IADC assigns a confidence level to the defect and stores this information with the defect in thedefect database 30. - In one embodiment, defects are classified in the global classification scheme based on similar spatial signature analysis results and similar on-the-fly automatic defect classification results. In addition, the images may be associated with the defect data per the defect classification scheme. In any event, the defects are presented to the FAB operator as a single, uniform set of defect classes.
- As above, based on the defect classifications, the IADC generates alarms and warnings. In addition, if needed, the IADC generates a SEM review plan.
- If SEM review is desired, the
wafer 40 is sent to the SEM defects reviewtool 24B. Thereview tool 24B generates a re-classified defect map (not shown), SEM defect images and, optionally, SEM-based automatic defect classifications and, if needed, composition data such as EDX. - The
IADC 28 processes the SEM information to further refine and globalize the defect information and assign new confidence levels to the results. Finally, the IADC generates alarms and warnings, as necessary, and stores the classified defect map in thedatabase 30. - FIG. 3 describes the basic operations of the
IADC 28. Briefly, theIADC 28 receives defect data from the inspection andreview tools blocks IADC 28 classifies the defects based on the defect data and data fromspatial signature analysis 56 of the wafer defects. The IADC stores all of the data related to the defects in thedatabase 30. Thus, the defect data is readily available for future classification searches 58 and 60. In addition, as represented byblocks IADC 28 has “learning” capabilities in that it continually refines the defect classifications and defect source data based on received defect data. As discussed above, the defectsource identification system 32 uses the defect data to identify the source of the defects and initiate corrective action. - Turning now to the details of FIG. 3, the
IADC 28 aligns and consolidates the defect data from all the local automatic defect classification (“ADC”) tools to produce unified classification. - Each of ADC tools depicted in FIG. 3 analyzes detected defects and generates defect attribute data. For example, referring initially to the inspection based
ADC tool 22, thetool 22 inspects the wafer and classifies detected defects. When available (for example when using the “ORBOT WF 7xx” series equipment), thetool 22 performs on-the-fly automatic defect classification.Tool 22 then sends the local automatic defect classification attribute data (e.g., intensity of scattered light, etc.) to theIADC 28 as represented byline 66. As represented byblock 50, the IADC processes the attribute data from the tools (line 68) along with spatial signature analysis information (line 70) and data from an ADC learning system 64 (via line 65) to generate integrated attribute classifications. The spatialsignature analysis information 70 consists of a conventional set of algorithms that describe a zonal analysis of defects distributed across the wafer. TheADC learning system 64 is described in more detail below. - Each of the ADC tools depicted in FIG. 3 also processes the attribute data to classify the defects into local, tool-specific classes as mentioned above. Referring again to the inspection based
ADC tool 22, thetool 22 provides local automatic defect classification results (e.g., a classified defect map) to theIADC 28 as represented byline 72. TheIADC 28 uses the class data from the tools (line 74) and the integrated attributes (line 76) to generateglobalized classifications 52. At this stage, the various class identifiers from the FAB are translated into a uniform, global set of classes. - The
IADC 28 stores all of the classification parameters related to the defects in thedatabase 30. Specifically, the IADC stores the refined classification data (line 78) and the integrated attributes classification information (line 80) in thedefect knowledge database 30. This includes automatic defect classification attribute data (block 60) that may be accessed during knowledge-based searches of thedatabase 30. - The
IADC 28 also stores defect image and composition data (line 82) in thedatabase 30. This information originates from theoptical tool 24A and theSEM tool 24B in the form of raw images and material identification data (line 84). Typically, the IADC uses this information during knowledge-basedsearches 58 of thedefect database 30 as discussed below. - During the classification process, the
IADC 28 refines and corrects the global classifications (line 86) and the integrated attributes classification (line 87) to provide enhanced classification results (block 54). By combining the data from the different local classifications, the IADC can provide a more accurate classification than the local classifications. - The classification refinement operation may take several forms. For example, the
IADC 28 may combine the results of specific defect classifications (e.g., inspection basedADC 22 oroptical review ADC 24A) and spatial defects distribution classification (e.g., spatial signature analysis 56) to refine the classification and increase it's accuracy. - One example for providing the global classification is by using weighting and probability factors. That is, each tool sends classification and level of confidence data to the
IADC 28. If a conflict arises, theIADC 28 can apply the level of confidence to the classification of each tool to arrive at a final determination. Thus, fortools 1 and 2, classifications X1 and X2, and levels of confidence P1 and P2, theIADC 28 can perform P1(X1)+P2(X2)→XG, where the classification of XG can be determined from a look-up table or using a threshold. - Using weights assigned to each tool, e.g., W1 and W2, the
IADC 28 can give preferential treatment to certain tools (e.g., a higher weight may be given to a high sensitivity review tool than to a fast scan inspection tool). Thus, theIADC 28 will perform W1·P1(X1)+W2·P2(X2)=XG. The weights can be specific to a tool and can be individualized per class of defects. As the system learns the sensitivity of the tools to the specific processing lines (using prior correct/incorrect classifications), the weights can be adjusted accordingly. It should be appreciated that for a certain tool/class, the weight can be set to zero to ignore the classification provided by that tool. - The
IADC 28 may be integrated into the classification chain depicted in FIG. 2. In this case, the classification is refined after each stage in the chain. That is, the classification is refined after the inspection basedADC 22. It is refined again after theoptical review ADC 24A and once again after theSEM ADC 24B. - Furthermore, the
IADC 28 may integrate the above chain into a learning system that combines all past defect results and spatial signature analyses. The results of a knowledge-based classification (one or more class identifiers and related data) are processed (line 88) by theenhanced ADC component 62. From this, thecomponent 62 generates a classification history which may include the original input data (classification attributes), the original classification data, the intermediate refinement data and the final classification data. - The classification history data is used (line90) by the
ADC learning system 64 to assign and refine confidence levels and weights for future ADC results and to optimize the integrated ADC operation for faster and more accurate classifications. Thus, thelearning system 64 uses information from prior defects to continually refine the classifications and the assignment of defects to those classifications. - The classification process may be better understood by way of the following example. Initially, the
IADC 28 receives defect information from one of the ADC tools. The defect may be, for example, a foreign particle on the surface of the wafer where the particle has a particular size and shape. In this example, the received defect information may initially only include the local class number (e.g., the number 25) that the ADC tool assigns to this type of defect. Preferably, the defect information also includes a confidence level determined by the inspection tool. - The next step in the process may depend on the reported confidence level or the confidence level that the IADC assigned to the local defect classification. For example, when there is a high level of confidence that the ADC tool accurately identifies this class of defect, the IADC may simply translate the local class into a global class. A high level of confidence for a given local class may be indicated by a high confidence value relative to the confidence level of other classes assigned to this defect by other ADC processes (e.g., 90% vs. 10%).
- When a translation of this particular local class to a global class has been predefined, the translation process may simply involve changing the class number. When the translation has not been predefined, however, the IADC must fit the class into an existing global class or generate a new global class for this type of defect.
- If the confidence level is not particularly high, the
IADC 28 may perform additional steps to calculate an appropriate global class. This may be the case, for example, when two ADC tools assign different classes to the defect and each class has approximately the same confidence level. - In this case, the
IADC 28 may request the defect attributes (e.g., sensing attributes such as intensity level or defect attributes such as whether it is a particle; its size: e.g., less than 0.1 microns; and its shape: e.g., approximately round) from the ADC tools and assign a class based the corresponding defect attributes 50 and/or on spatialsignature analysis information 56 associated with the defect. This may involve, for example, comparing the defect attributes with the attributes associated with the global classes. - In addition, the
IADC 28, will assign an initial confidence level to this class. In accordance with the invention, the IADC may assign a confidence level by combining algorithmic approaches with past empirical data. For example, the confidence level may be based on the relative number (e.g., percentage) of matches (e.g., size, shape, density) between the class attributes and the attributes associated with the defect. The class and/or the confidence level may then be refined by subsequent ADC processes. - The case where two ADC tools (e.g.,22 or 24) provide conflicting classification data illustrates a situation where the
ADC learning system 64 may be used to control the ADC tools to improve the inspection and review process in the FAB. Some ADC tools have variable defect thresholds or decision criteria. For example, an ADC tool may be configured so that only signals above a certain intensity would be reported as defects. In this case, after the proper classification has been determined, theADC learning system 64 may send messages to the appropriate ADC tools to change their threshold settings to result in a proper classification in future inspections. Alternatively, thelearning system 64 may send a message to adisplay device 92 that informs the FAB operator that the thresholds of the ADC tools should be changed. - If the confidence level is still unacceptably low after the step previously discussed, the
IADC 28 may attempt to classify the defect by performing a database search. In this case, the IADC will compare the attributes, the signature analysis, the defect maps, the image data and other information associated with the defect with the knowledge search information (e.g., 58 and 60) stored in thedatabase 30. As above, the selected global class and confidence levels may be based on the relative percentage of matches between the information associated with the defect and the information stored in thedatabase 30. - Using the defect classification data and other FAB data such as overlay, yield and kill ratio information, the
IADC 28 attempts to generate (line 94) the best classification possible for a given production line. This classification may include a class identifier for each defect and potential alternative classes in case of inconclusive class identification. In addition, the IADC typically produces a confidence level for each of the classifications. - It should be appreciated that the use of the
database 30 as described herein provides an advantageous method of coordinating the various processes of theyield management system 18. Thedatabase 30 stores information that is used in the defect classification process, in the defect classification refinement process and in the defect source identification process. - Briefly, the
database 30 stores all of the defect data including the defect attributes, spatial signatures, the local and global classes, confidence levels, defect densities, descriptive texts of the events, wafer maps and images.Database 30 may also store manufacturer-specific data. For example, a manufacturer of an etch chamber may provide specific data related to defects caused by corrosion in the chamber. Such data can help in pointing out the specific chamber causing a specific defect. - By storing all classification data in the
database 30, a yield engineer can access this data to determine how a specific classification was assigned. Moreover, the yield engineer may perform manual classifications whereby the engineer manually “teaches” the system to tie specific parameters to a specific class, thereby customizing and improving the classification results for a particular FAB and application. This procedure or a similar procedure may be used at a later date by the system or by the FAB engineer to assign a class to other defects. For example, the yield enhancement system may be programmed to follow similar procedures to perform automatic defect classification. - The
IADC 28 also uses the information in thedefects database 30 to validate the algorithmic classifications on an ongoing basis. Thus, over time, the accuracy of the algorithmic classifications will increase and the chance of erroneous or fruitless classifications will decrease. Moreover, it should be appreciated that thedatabase 30 may initially be configured without any defect information. In this case, theyield enhancement system 18 can add defect information to the database as it acquires defect information from the FAB and/or the operator may input information that correlates the defects with potential sources, or with manufacturer-specific data as it is received at the FAB. - Given FAB yield data as an input, the
IADC 28 can correlate yield results to specific defects and defect classes. As a result, the IADC may refine the identification of killer defects. This may, in turn, enable the yield engineer to ignore a significant percentage of nuisance and redundant defects, while concentrating on avoiding killer defects. - Referring now to the defect
source identification system 32, theidentification system 32 uses the information from the IADC (line 94) and thedatabase 30 to identify potential sources for a given defect or a class. In particular, theidentification system 32 may use the defect source information stored in thedatabase 30. This information typically would be in the form: machine “x” may have problem “y” and this will show up as a defect of class “z.” - The defect
source identification system 32 may also perform more complicated defect root causes analysis. For example, theidentification system 32 may identify a defect source using the specific defect and defect class identified by the inspection andreview tools IADC 28 and thedatabase 30. For example, a class that covers missing or extra signal path patterns may relate to a problem in the photo-resist process. Thus, the source of this problem may be the introduction of a foreign particle prior to the etch process. Alteratively, an embedded particle may indicate a problem before the deposition process. These problems may, in turn, be traced to specific equipment in the FAB, especially when manufacturer-specific data is available. - Other inputs to the
identification system 32 may include work-in-progress (“WIP”) data and yield data from theFAB host system 20 and the accumulated statistical probability data calculated by the yield enhancement system. - The defect
source identification system 32 also generates automatic alarms and warnings associated with the class identifier(s) as configured by the FAB. In particular, theyield enhancement system 18 can provide class density monitoring rather than conventional total density monitoring. Because prior high defect events may be stored on a per class basis, theyield enhancement system 18 can generate alarms based on a threshold for specific classes. This is in contrast to many conventional systems that generate alarms based on a threshold for the aggregate of the defects. As a result, using a yield enhancement system constructed according to the invention, the yield engineer can rapidly identify developing excursions in a given class that would otherwise be insignificant (and hence relatively difficult to identify) in relation to the aggregate of the defects. - The defect
source identification system 32 may generate suggested test procedures to verify the identification of the source of the defect. For example, theidentification system 32 may identify a given sub-process of a process tool (e.g., an incorrect temperature at a given etcher stage) as a possible source of a defect. If that process tool (e.g., the etcher) has a built-in self-test for that sub-process, theidentification system 32 may automatically send a message to the process tool (using the WIP data that identifies the etcher number) requesting the process tool to run a self-test. In addition, theidentification system 32 also requests the process tool to send back the results of the test to the identification system. This may be done, for example, by sending an appropriate message to the process tool ortools 96. If that sub-process fails (e.g., the temperature is incorrect), theidentification system 32 may increase the level of confidence that that sub-process is the cause of the defect, as appropriate. - The defect
source identification system 32 may initiate corrective procedures to fix the problem. In the previous example, the etcher may have an interface that allows a remote device to adjust the temperature by sending an appropriate command. In this case, theidentification system 32 may be configured to send this command in the event the self-test failed. Alternatively, the identification system may be configured to display a message on thedisplay device 92 that informs the FAB engineer to check parameter “x” (e.g., temperature) on machine “y” (e.g., etcher number 4). - FIG. 4 describes yield control planning and sample planning operations in accordance with one embodiment of the invention. A yield modeling and
planning component 108 generates a yield control andstrategy plan 110. Typically, an initial plan is developed when the FAB is set up based on simulation modeling of the performance and interaction of the equipment installed in the FAB. In accordance with the invention, asmart sample planner 34 continually refines the yield control plan (e.g., the wafer inspection and review plans) based on in-line information such as WIP data from the FAB equipment, actual wafer yield data and ADC data as discussed above. - In the yield
control planning component 108, design-based yield modeling incorporates mathematical modeling and simulation algorithms to generate (line 112) predictions of yield sensitivities for each layer and for particular areas on the chip. Predictions may also be generated for potential killer defects sizes, distributions, and quantities. Process tool-based yield modeling uses historical data and statistical analysis and prediction to generate (line 114) predictions of defects for a set of process tools. - From this information and the results from process interaction modeling (line116), the modeling and
planning component 108 generates yield predictions for the entire line and the specific production modules (lines 118 and 120). In addition, thecomponent 108 generates yield thresholds for each process module and for each process tool in the FAB (line 122). Finally, theyield control component 110 generates the yield control plan (line 124) for the entire production line. This plan may define which tools are used in the FAB. It may define how to position the tools and how to use them. In addition, the plan may specify which wafers are to be inspected. - The
sample planner 34 uses the in-line defect classification data discussed above and other accumulated knowledge in the system to improve the effectiveness of the yield control plan. By continually refining the yield control plan, thesample planner 34 will enable the yield engineer to more effectively focus on killer defects and defects of interest. As a result, this may reduce the amount of time the yield engineer spends reviewing defects on each wafer and reduce the number of test wafer that must be wasted when setting up a wafer process the FAB. - The
sample planner 34 generates a preliminary inspection and review plan (line 126) using the yield control plan and an enhanced set of initial planning rules (line 128). The initial sample plan is optimized using efficiency estimations (line 130) to produce the wafer inspection plan (line 132, e.g., which wafer to inspect, when to inspect it and how to inspect it), an optical review plan (line 134) and a SEM review plan (line 136). - Using the above information, the sample planner generates inspection recipes that are tuned and optimized for the specific layer inspected. The sample planner provides optimized ADC, by an inspection tool (re-visited) and/or by a review tool. The sample planner also may provide an optimized sampling plan for the review tool.
- The sample planner may be programmed to “learn” which type of defects warrant further review and which type of defects may be skipped during the review process. For example, if a hundred particles are located at a specific location on the wafer, it may only be necessary to review a few (e.g., five) of these particles with the review tool. This may be the case where there is a relatively high level of confidence in the defect information provided by an inspection tool that indicates that all one hundred of the defects are similar particles. This high level of confidence may result from, for example, prior occurrences of similar defects. Thus, once the system is tuned and there is confidence in the inspection tool, this information may be programmed into the sampling plan so that the review tool automatically reviews just five of the one hundred defects.
- As discussed above in conjunction with FIGS.1-3, the inspection and review information is then fed back into the yield enhancement system 18 (line 138). The fed back information is used for efficiency evaluations and to generate proposed enhancements to the inspection and review plans and for an efficiency evaluation of the original inspection plan (line 140). These enhancements are based on the inspections and review direct results (defects of interest) and on the historical correlation of the defects to actual yield (test data).
- The inputs to the efficiency estimator include layer and process tools data acquired from the FAB host (line142). This may include, for example, metrology data such as the state of a process tool. The inputs may also include killer defect identification information that is based on past correlation with actual yield data (line 144). In addition, defects classification information (line 148) and information about the defects distribution on the wafers that is obtained from spatial signature analysis (line 148) may also be fed into the
yield enhancement system 18. - From the above, it may be seen that the sample planner uses the layer and tool information to decrease inspection and review time and improve inspection results. This results in reduced test wafer consumption in the FAB and reduces throughput time for production material, reduces capital expenditures for inspection/review/metrology tools and saves FAB floor space.
- The automatic yield enhancement system shortens the time to identify the sources of yield excursions, shortens the learning time of new defects, and improves the productivity of the wafer inspection and review tools by optimizing their use. These goals are achieved through improved data management; integrating several types of FAB data and utilizing new software algorithms to automate and speed up the decision process to monitor and control FAB yield. The system is a learning system, accumulating data in both automatic and manual modes, and improving the accuracy of automatic decisions based on that accumulated data. The productivity of the inspections and review tools also improved by the sampling planning module. This module uses the integrated data to provide improved plans for using the tools.
- The system may be deployed across the FAB, as an overall yield enhancement system. Alternatively, it may be deployed in specific areas of the FAB to provide yield control for a specific group of process, inspection, and metrology tools. For example, in a fully automated system (manufacturing cells, short loop control), the validation and corrective action may be communicated automatically to the process tool, thereby prompting it to perform automatic service routines. In this embodiment, the system uses the service routines supplied by the process tool manufacturer and complies with the protocol that activates them automatically.
- It should be appreciated that the defect knowledge database can be used for defect source identification irrespective, or in addition to, defect classification. Thus, for example, inspection tool22 (FIG. 1) can send a defect map to the automatic
yield enhancement system 18. Then, certain or all of the defects (depending, for example, of the output of the smart sampling plan 34) may be reviewed in the optical and/or SEM review tools (24A, 24B, FIG. 2). Further, if particles are found, a material (EDX) analysis can be performed using, for example, a SEM review station such as the Opal 9300 in an EDX mode. - Since the defect knowledge database includes defect information, such as manufacturer specific and previous analyses results, a search for a match can be performed using the optical and/or SEM images and the EDX results. In many cases, a good match can obviate further processing and point directly at the defect source and proposed corrective action. For example, particles can come from different sources and different parts of the processing chambers. When particles of particular composition are identified, they are stored in the defect knowledge database. When the source of the particles has been eliminated using a particular corrective action, the corrective action is also stored with a link to the particle information. Then, using a material composition match, the appropriate chamber can be identified and the stored corrective action provided as a probable solution for the problem.
- Thus, for example, when it is discovered that particles of particular composition have been introduced from a dome of an etcher, and that a plasma clear has solved the problem, such data can be stored in the defect knowledge database. Then, during an investigation of particles problem, when the EDX reveals that the particles are of the same composition as stored in the database, the system may suggest that a plasma clean be performed on the etcher dome. Thus, the suggested corrective action can be reached at very fast, while avoiding having to perform much processing. Of course, at the same time the system may perform the classification and other processing as discussed above.
- From the above, it may be seen that a system constructed according to the invention provides improved semiconductor fabrication. While certain specific embodiments of the invention are disclosed as typical, the invention is not limited to these particular forms, but rather is applicable broadly to all such variations as fall within the scope of the appended claims. To those skilled in the art to which the invention pertains many modifications and adaptations will occur. For example, various method classification schemes and methods may be used in implementing the teachings of the invention. In addition, a variety of methods may be sued to identify the source of defects. Moreover, the learning aspects of the invention may be implemented in many ways. Thus, the specific structures and methods discussed in detail above are merely illustrative of a few specific embodiments of the invention.
Claims (17)
Priority Applications (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/075,254 US6408219B2 (en) | 1998-05-11 | 1998-05-11 | FAB yield enhancement system |
KR1020007012589A KR20010043501A (en) | 1998-05-11 | 1999-05-04 | Fab yield enhancement system |
EP99920382A EP1080493A1 (en) | 1998-05-11 | 1999-05-04 | Fab yield enhancement system |
PCT/US1999/009933 WO1999059200A1 (en) | 1998-05-11 | 1999-05-04 | Fab yield enhancement system |
JP2000548916A JP2002515650A (en) | 1998-05-11 | 1999-05-04 | Manufacturing equipment yield improvement system |
TW088107554A TW509962B (en) | 1998-05-11 | 1999-06-11 | FAB yield enhancement system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/075,254 US6408219B2 (en) | 1998-05-11 | 1998-05-11 | FAB yield enhancement system |
Publications (2)
Publication Number | Publication Date |
---|---|
US20010051836A1 true US20010051836A1 (en) | 2001-12-13 |
US6408219B2 US6408219B2 (en) | 2002-06-18 |
Family
ID=22124518
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/075,254 Expired - Lifetime US6408219B2 (en) | 1998-05-11 | 1998-05-11 | FAB yield enhancement system |
Country Status (6)
Country | Link |
---|---|
US (1) | US6408219B2 (en) |
EP (1) | EP1080493A1 (en) |
JP (1) | JP2002515650A (en) |
KR (1) | KR20010043501A (en) |
TW (1) | TW509962B (en) |
WO (1) | WO1999059200A1 (en) |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030182252A1 (en) * | 2002-03-21 | 2003-09-25 | Applied Materials, Inc. | Correlation of end-of-line data mining with process tool data mining |
US6658361B1 (en) * | 2001-10-10 | 2003-12-02 | Lsi Logic Corporation | Heaviest only fail potential |
US6701477B1 (en) * | 2000-06-09 | 2004-03-02 | Hueristics Physics Laboratories | Method for identifying the cause of yield loss in integrated circuit manufacture |
US20040080742A1 (en) * | 2002-10-24 | 2004-04-29 | Mitsubishi Denki Kabushiki Kaisha | Device inspecting for defect on semiconductor wafer surface |
US20060100730A1 (en) * | 2002-07-12 | 2006-05-11 | Parkes Alan S | Method for detection and relocation of wafer defects |
US20060235560A1 (en) * | 2005-04-05 | 2006-10-19 | Akira Ogawa | Quality control system, quality control method, and method of lot-to-lot wafer processing |
US20070004051A1 (en) * | 2002-12-17 | 2007-01-04 | Yasuhiro Okumoto | Processing method and device |
US20070135956A1 (en) * | 2005-12-13 | 2007-06-14 | Taiwan Semiconductor Manufacturing Co., Ltd. | Data location systems and methods |
US20070282544A1 (en) * | 2006-06-02 | 2007-12-06 | Zhang Guohai | Data analysis method for integrated circuit process and semiconductor process |
US20080221831A1 (en) * | 2004-06-14 | 2008-09-11 | Semiconductor Manufacturing Intenational (Shanghai) Corporation | Method and system for yield similarity of semiconductor devices |
US7650199B1 (en) * | 2005-08-03 | 2010-01-19 | Daniel Kadosh | End of line performance prediction |
US20100067781A1 (en) * | 2003-11-19 | 2010-03-18 | Kla-Tencor Corporation | Process Excursion Detection |
US20120136470A1 (en) * | 2009-05-22 | 2012-05-31 | Aurora Control Technologies, Inc. | Process for improving the production of photovoltaic products |
US20150039118A1 (en) * | 2013-08-01 | 2015-02-05 | Applied Materials, Inc. | Event processing based system for manufacturing yield improvement |
US8989478B2 (en) | 2011-07-29 | 2015-03-24 | Kla-Tencor Corporation | Method and system for visualization of semiconductor wafer inspection data acquired in a photovoltaic cell production process |
EP2663998A4 (en) * | 2011-01-11 | 2016-06-15 | Kla Tencor Corp | Methods and systems of object based metrology for advanced wafer surface nanotopography |
CN106653639A (en) * | 2016-12-26 | 2017-05-10 | 中航(重庆)微电子有限公司 | Sampling method of semiconductor measurement site |
US10955832B2 (en) * | 2018-10-09 | 2021-03-23 | Applied Materials, Inc. | Adaptive chamber matching in advanced semiconductor process control |
US11177111B2 (en) | 2019-04-19 | 2021-11-16 | Hitachi High-Tech Corporation | Defect observation device |
US11346882B2 (en) * | 2017-11-03 | 2022-05-31 | Tokyo Electron Limited | Enhancement of yield of functional microelectronic devices |
CN117878025A (en) * | 2024-03-11 | 2024-04-12 | 江苏芯德半导体科技有限公司 | Method for automatically clamping and controlling defective chips on wafer map |
Families Citing this family (109)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020158197A1 (en) * | 1999-01-12 | 2002-10-31 | Applied Materials, Inc | AFM-based lithography metrology tool |
KR100303321B1 (en) * | 1999-05-20 | 2001-09-26 | 박종섭 | Appratus for controlling abnormal lot in automization system to produce semiconductor and control method using the same |
US6922482B1 (en) * | 1999-06-15 | 2005-07-26 | Applied Materials, Inc. | Hybrid invariant adaptive automatic defect classification |
US7796801B2 (en) | 1999-08-26 | 2010-09-14 | Nanogeometry Research Inc. | Pattern inspection apparatus and method |
US6636818B1 (en) * | 1999-09-15 | 2003-10-21 | Becton, Dickinson And Company | Systems, methods and computer program products for constructing sampling plans for items that are manufactured |
US6999614B1 (en) | 1999-11-29 | 2006-02-14 | Kla-Tencor Corporation | Power assisted automatic supervised classifier creation tool for semiconductor defects |
JP2001168160A (en) * | 1999-12-07 | 2001-06-22 | Sony Corp | System for inspecting semiconductor wafer |
AU2001248851A1 (en) | 2000-04-25 | 2001-11-07 | Tokyo Electron Limited | Semiconductor device inspection system |
US6725115B2 (en) * | 2000-05-08 | 2004-04-20 | Teradyne, Inc. | Dynamic testing of electronic assemblies |
US6963813B1 (en) * | 2000-09-13 | 2005-11-08 | Dieter Rathei | Method and apparatus for fast automated failure classification for semiconductor wafers |
US20020065900A1 (en) * | 2000-10-02 | 2002-05-30 | Applied Materials, Inc. | Method and apparatus for communicating images, data, or other information in a defect source identifier |
TWI256468B (en) * | 2000-10-02 | 2006-06-11 | Applied Materials Inc | Defect source identifier |
US6701259B2 (en) * | 2000-10-02 | 2004-03-02 | Applied Materials, Inc. | Defect source identifier |
US6744266B2 (en) * | 2000-10-02 | 2004-06-01 | Applied Materials, Inc. | Defect knowledge library |
US20020046001A1 (en) * | 2000-10-16 | 2002-04-18 | Applied Materials, Inc. | Method, computer readable medium and apparatus for accessing a defect knowledge library of a defect source identification system |
US6873720B2 (en) | 2001-03-20 | 2005-03-29 | Synopsys, Inc. | System and method of providing mask defect printability analysis |
US6925202B2 (en) | 2001-03-20 | 2005-08-02 | Synopsys, Inc. | System and method of providing mask quality control |
US6970757B1 (en) * | 2001-04-19 | 2005-11-29 | Advanced Micro Devices, Inc. | Method and apparatus for updating control state variables of a process control model based on rework data |
US7698012B2 (en) * | 2001-06-19 | 2010-04-13 | Applied Materials, Inc. | Dynamic metrology schemes and sampling schemes for advanced process control in semiconductor processing |
US7160739B2 (en) | 2001-06-19 | 2007-01-09 | Applied Materials, Inc. | Feedback control of a chemical mechanical polishing device providing manipulation of removal rate profiles |
US6721928B2 (en) | 2001-07-26 | 2004-04-13 | Numerical Technologies, Inc. | Verification utilizing instance-based hierarchy management |
US6560766B2 (en) | 2001-07-26 | 2003-05-06 | Numerical Technologies, Inc. | Method and apparatus for analyzing a layout using an instance-based representation |
US7014955B2 (en) | 2001-08-28 | 2006-03-21 | Synopsys, Inc. | System and method for indentifying dummy features on a mask layer |
US6976240B2 (en) | 2001-11-14 | 2005-12-13 | Synopsys Inc. | Simulation using design geometry information |
US6650955B1 (en) * | 2001-12-18 | 2003-11-18 | Advanced Micro Devices, Inc. | Method and apparatus for determining a sampling plan based on process and equipment fingerprinting |
US6821792B1 (en) * | 2001-12-18 | 2004-11-23 | Advanced Micro Devices, Inc. | Method and apparatus for determining a sampling plan based on process and equipment state information |
US20030135295A1 (en) * | 2002-01-14 | 2003-07-17 | Applied Materials, Inc. | Defect source identifier with static manufacturing execution system |
US20030223639A1 (en) * | 2002-03-05 | 2003-12-04 | Vladimir Shlain | Calibration and recognition of materials in technical images using specific and non-specific features |
US6687561B1 (en) * | 2002-04-03 | 2004-02-03 | Advanced Micro Devices, Inc. | Method and apparatus for determining a sampling plan based on defectivity |
TW535207B (en) * | 2002-05-08 | 2003-06-01 | Powerchip Semiconductor Corp | Method for automatically controlling defect-specification in manufacturing process of semiconductors |
JP4118703B2 (en) * | 2002-05-23 | 2008-07-16 | 株式会社日立ハイテクノロジーズ | Defect classification apparatus, automatic defect classification method, defect inspection method, and processing apparatus |
US6785623B2 (en) * | 2002-09-11 | 2004-08-31 | The United States Of America As Represented By The Secretary Of The Navy | Business to business electronic test monitoring information system |
US7043071B2 (en) | 2002-09-13 | 2006-05-09 | Synopsys, Inc. | Soft defect printability simulation and analysis for masks |
US7968354B1 (en) | 2002-10-04 | 2011-06-28 | Kla-Tencor Technologies Corp. | Methods for correlating backside and frontside defects detected on a specimen and classification of backside defects |
US7379175B1 (en) | 2002-10-15 | 2008-05-27 | Kla-Tencor Technologies Corp. | Methods and systems for reticle inspection and defect review using aerial imaging |
US7123356B1 (en) | 2002-10-15 | 2006-10-17 | Kla-Tencor Technologies Corp. | Methods and systems for inspecting reticles using aerial imaging and die-to-database detection |
US7027143B1 (en) | 2002-10-15 | 2006-04-11 | Kla-Tencor Technologies Corp. | Methods and systems for inspecting reticles using aerial imaging at off-stepper wavelengths |
US7133119B1 (en) | 2002-12-17 | 2006-11-07 | Kla-Tencor Technologies Corp. | Systems for simulating high NA and polarization effects in aerial images |
US7075318B1 (en) | 2003-01-16 | 2006-07-11 | Kla-Tencor Technologies Corp. | Methods for imperfect insulating film electrical thickness/capacitance measurement |
US6996790B2 (en) * | 2003-01-30 | 2006-02-07 | Synopsys, Inc. | System and method for generating a two-dimensional yield map for a full layout |
US7359544B2 (en) * | 2003-02-12 | 2008-04-15 | Kla-Tencor Technologies Corporation | Automatic supervised classifier setup tool for semiconductor defects |
US7602962B2 (en) * | 2003-02-25 | 2009-10-13 | Hitachi High-Technologies Corporation | Method of classifying defects using multiple inspection machines |
US6952653B2 (en) * | 2003-04-29 | 2005-10-04 | Kla-Tencor Technologies Corporation | Single tool defect classification solution |
US6954711B2 (en) * | 2003-05-19 | 2005-10-11 | Applied Materials, Inc. | Test substrate reclamation method and apparatus |
US9002497B2 (en) | 2003-07-03 | 2015-04-07 | Kla-Tencor Technologies Corp. | Methods and systems for inspection of wafers and reticles using designer intent data |
WO2005006002A2 (en) * | 2003-07-12 | 2005-01-20 | Leica Microsystems Semiconductor Gmbh | Method of learning a knowledge-based database used in automatic defect classification |
US6968252B1 (en) * | 2003-07-15 | 2005-11-22 | Advanced Micro Devices, Inc. | Method and apparatus for dispatching based on metrology tool performance |
JP4568216B2 (en) * | 2003-09-08 | 2010-10-27 | 株式会社東芝 | Semiconductor device manufacturing system |
JP4008899B2 (en) * | 2003-09-08 | 2007-11-14 | 株式会社東芝 | Semiconductor device manufacturing system and semiconductor device manufacturing method |
US7227628B1 (en) | 2003-10-10 | 2007-06-05 | Kla-Tencor Technologies Corp. | Wafer inspection systems and methods for analyzing inspection data |
TW200516686A (en) | 2003-11-10 | 2005-05-16 | Renesas Tech Corp | Manufacturing method for semiconductor integrated circuit device |
JP4771714B2 (en) * | 2004-02-23 | 2011-09-14 | 株式会社Ngr | Pattern inspection apparatus and method |
US7558999B2 (en) * | 2004-05-21 | 2009-07-07 | International Business Machines Corporation | Learning based logic diagnosis |
JP4758427B2 (en) * | 2004-07-21 | 2011-08-31 | ケーエルエー−テンカー コーポレイション | Computer-implemented method for generating input for simulation programs or for generating simulated images of reticles |
US7069098B2 (en) * | 2004-08-02 | 2006-06-27 | Advanced Micro Devices, Inc. | Method and system for prioritizing material to clear exception conditions |
JP4317805B2 (en) * | 2004-09-29 | 2009-08-19 | 株式会社日立ハイテクノロジーズ | Defect automatic classification method and apparatus |
US7296103B1 (en) | 2004-10-05 | 2007-11-13 | Advanced Micro Devices, Inc. | Method and system for dynamically selecting wafer lots for metrology processing |
US7076321B2 (en) * | 2004-10-05 | 2006-07-11 | Advanced Micro Devices, Inc. | Method and system for dynamically adjusting metrology sampling based upon available metrology capacity |
JP5134188B2 (en) * | 2004-10-15 | 2013-01-30 | ケーエルエー−テンカー コーポレイション | Equipment for analyzing defects on specimens |
US7263451B1 (en) * | 2004-10-25 | 2007-08-28 | Advanced Micro Devices, Inc. | Method and apparatus for correlating semiconductor process data with known prior process data |
JP4787673B2 (en) * | 2005-05-19 | 2011-10-05 | 株式会社Ngr | Pattern inspection apparatus and method |
JP4681356B2 (en) * | 2005-06-13 | 2011-05-11 | 株式会社日立ハイテクノロジーズ | Defect automatic observation classification system, apparatus selection method, program, and observation apparatus in defect automatic observation classification system |
US7657390B2 (en) | 2005-11-02 | 2010-02-02 | Applied Materials, Inc. | Reclaiming substrates having defects and contaminants |
US7570796B2 (en) | 2005-11-18 | 2009-08-04 | Kla-Tencor Technologies Corp. | Methods and systems for utilizing design data in combination with inspection data |
US7305325B2 (en) * | 2006-01-12 | 2007-12-04 | International Business Machines Corporation | Method to improve requirements, design manufacturing, and transportation in mass manufacturing industries through analysis of defect data |
KR100801075B1 (en) * | 2006-01-27 | 2008-02-11 | 삼성전자주식회사 | Method for enhancing yield of semiconductor integrate circuit device and system for the same using a hole's systematic fault rate |
KR100696276B1 (en) * | 2006-01-31 | 2007-03-19 | (주)미래로시스템 | Automatic defect classification system based on the measurement data from wafer defect inspection equipments |
US7954072B2 (en) * | 2006-05-15 | 2011-05-31 | Taiwan Semiconductor Manufacturing Company, Ltd. | Model import for electronic design automation |
US8136067B2 (en) * | 2006-05-15 | 2012-03-13 | Taiwan Semiconductor Manufacturing Company, Ltd. | Method of design for manufacturing |
US8336002B2 (en) * | 2006-05-15 | 2012-12-18 | Taiwan Semiconductor Manufacturing Company, Ltd. | IC design flow enhancement with CMP simulation |
US7725861B2 (en) * | 2006-05-15 | 2010-05-25 | Taiwan Semiconductor Manufacturing Company, Ltd. | Method, apparatus, and system for LPC hot spot fix |
US8136168B2 (en) * | 2006-05-15 | 2012-03-13 | Taiwan Semiconductor Manufacturing Company, Ltd. | System and method for design-for-manufacturability data encryption |
US7685558B2 (en) * | 2006-05-15 | 2010-03-23 | Taiwan Semiconductor Manufacturing Company, Ltd. | Method for detection and scoring of hot spots in a design layout |
US7805692B2 (en) * | 2006-05-15 | 2010-09-28 | Taiwan Semiconductor Manufacturing Company, Ltd. | Method for local hot spot fixing |
US20070266360A1 (en) * | 2006-05-15 | 2007-11-15 | Taiwan Semiconductor Manufacturing Company, Ltd. | Metal Thickness Simulation for Improving RC Extraction Accuracy |
US7904844B2 (en) * | 2007-03-30 | 2011-03-08 | Taiwan Semiconductor Manufacturing Company, Ltd. | System, method, and computer program product for matching cell layout of an integrated circuit design |
US7788612B2 (en) * | 2007-03-30 | 2010-08-31 | Taiwan Semiconductor Manufacturing Company, Ltd. | System, method, and computer program product for matching cell layout of an integrated circuit design |
US8682466B2 (en) * | 2007-05-04 | 2014-03-25 | Taiwan Semiconductor Manufacturing Company, Ltd. | Automatic virtual metrology for semiconductor wafer result prediction |
US8145337B2 (en) * | 2007-05-04 | 2012-03-27 | Taiwan Semiconductor Manufacturing Company, Ltd. | Methodology to enable wafer result prediction of semiconductor wafer batch processing equipment |
US7783999B2 (en) * | 2008-01-18 | 2010-08-24 | Taiwan Semiconductor Manufacturing Company, Ltd. | Electrical parameter extraction for integrated circuit design |
US8037575B2 (en) * | 2008-02-28 | 2011-10-18 | Taiwan Semiconductor Manufacturing Company, Ltd. | Method for shape and timing equivalent dimension extraction |
KR101841897B1 (en) | 2008-07-28 | 2018-03-23 | 케이엘에이-텐코어 코오포레이션 | Computer-implemented methods, computer-readable media, and systems for classifying defects detected in a memory device area on a wafer |
US8001494B2 (en) * | 2008-10-13 | 2011-08-16 | Taiwan Semiconductor Manufacturing Company, Ltd. | Table-based DFM for accurate post-layout analysis |
JP2010112833A (en) * | 2008-11-06 | 2010-05-20 | Hitachi High-Technologies Corp | Electron beam type substrate inspecting apparatus |
JP5429869B2 (en) | 2008-12-22 | 2014-02-26 | 株式会社 Ngr | Pattern inspection apparatus and method |
US8150140B2 (en) | 2008-12-22 | 2012-04-03 | Ngr Inc. | System and method for a semiconductor lithographic process control using statistical information in defect identification |
KR20120093820A (en) * | 2009-07-01 | 2012-08-23 | 케이엘에이-텐코 코포레이션 | Monitoring of time-varying defect classification performance |
JP5573058B2 (en) * | 2009-09-14 | 2014-08-20 | 凸版印刷株式会社 | Cleaning instruction system for substrate transfer roller |
US8806386B2 (en) * | 2009-11-25 | 2014-08-12 | Taiwan Semiconductor Manufacturing Company, Ltd. | Customized patterning modulation and optimization |
US8745554B2 (en) * | 2009-12-28 | 2014-06-03 | Taiwan Semiconductor Manufacturing Company, Ltd. | Practical approach to layout migration |
US9170211B2 (en) | 2011-03-25 | 2015-10-27 | Kla-Tencor Corp. | Design-based inspection using repeating structures |
US9087367B2 (en) | 2011-09-13 | 2015-07-21 | Kla-Tencor Corp. | Determining design coordinates for wafer defects |
CN102569118B (en) * | 2011-11-28 | 2014-06-04 | 上海华力微电子有限公司 | Yield increasing system of excursion management in semiconductor manufacturing process |
CN102522350B (en) * | 2011-11-29 | 2014-05-28 | 上海华力微电子有限公司 | Method and device for detecting failed production machine platforms |
CN102412168B (en) * | 2011-11-30 | 2013-10-30 | 上海华力微电子有限公司 | Wafer defect defection method and system |
US9189844B2 (en) | 2012-10-15 | 2015-11-17 | Kla-Tencor Corp. | Detecting defects on a wafer using defect-specific information |
US9053527B2 (en) | 2013-01-02 | 2015-06-09 | Kla-Tencor Corp. | Detecting defects on a wafer |
US9134254B2 (en) | 2013-01-07 | 2015-09-15 | Kla-Tencor Corp. | Determining a position of inspection system output in design data space |
US9311698B2 (en) | 2013-01-09 | 2016-04-12 | Kla-Tencor Corp. | Detecting defects on a wafer using template image matching |
US9092846B2 (en) | 2013-02-01 | 2015-07-28 | Kla-Tencor Corp. | Detecting defects on a wafer using defect-specific and multi-channel information |
US9865512B2 (en) | 2013-04-08 | 2018-01-09 | Kla-Tencor Corp. | Dynamic design attributes for wafer inspection |
US9310320B2 (en) | 2013-04-15 | 2016-04-12 | Kla-Tencor Corp. | Based sampling and binning for yield critical defects |
US10114368B2 (en) * | 2013-07-22 | 2018-10-30 | Applied Materials Israel Ltd. | Closed-loop automatic defect inspection and classification |
KR101978995B1 (en) * | 2015-06-04 | 2019-05-16 | 가부시키가이샤 히다치 하이테크놀로지즈 | Defect image classification apparatus and defect image classification method |
US10234401B2 (en) * | 2016-02-22 | 2019-03-19 | Qoniac Gmbh | Method of manufacturing semiconductor devices by using sampling plans |
US10598617B2 (en) | 2017-05-05 | 2020-03-24 | Kla-Tencor Corporation | Metrology guided inspection sample shaping of optical inspection results |
JP6908654B2 (en) * | 2018-08-31 | 2021-07-28 | ファナック株式会社 | Knowledge information service system |
US11244873B2 (en) * | 2018-10-31 | 2022-02-08 | Tokyo Electron Limited | Systems and methods for manufacturing microelectronic devices |
US11263737B2 (en) * | 2019-01-10 | 2022-03-01 | Lam Research Corporation | Defect classification and source analysis for semiconductor equipment |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6009545A (en) * | 1995-04-25 | 1999-12-28 | Mitsubishi Denki Kabushiki Kaisha | System for analyzing a failure in a semiconductor wafer by calculating correlation coefficient between collated data of defects per prescribed unit and failures per prescribed unit |
US5991699A (en) * | 1995-05-04 | 1999-11-23 | Kla Instruments Corporation | Detecting groups of defects in semiconductor feature space |
US5539752A (en) * | 1995-06-30 | 1996-07-23 | Advanced Micro Devices, Inc. | Method and system for automated analysis of semiconductor defect data |
US5777901A (en) | 1995-09-29 | 1998-07-07 | Advanced Micro Devices, Inc. | Method and system for automated die yield prediction in semiconductor manufacturing |
US5761064A (en) * | 1995-10-06 | 1998-06-02 | Advanced Micro Devices, Inc. | Defect management system for productivity and yield improvement |
US5943437A (en) * | 1995-10-09 | 1999-08-24 | Kabushiki Kaisha Kobe Seiko Sho | Method and apparatus for classifying a defect on a semiconductor wafer |
EP0783170B1 (en) | 1996-01-04 | 2001-03-28 | Infineon Technologies AG | Apparatus and method for the acquisition and analysis of a three-dimensional distribution of discrete points |
JPH09199551A (en) * | 1996-01-12 | 1997-07-31 | Mitsubishi Electric Corp | Inspection data analyzing/processing system for inline inspection |
DE69738979D1 (en) * | 1996-03-19 | 2008-10-23 | Hitachi Ltd | PROCESS CONTROL SYSTEM |
US6246787B1 (en) * | 1996-05-31 | 2001-06-12 | Texas Instruments Incorporated | System and method for knowledgebase generation and management |
US6092059A (en) * | 1996-12-27 | 2000-07-18 | Cognex Corporation | Automatic classifier for real time inspection and classification |
TW331650B (en) | 1997-05-26 | 1998-05-11 | Taiwan Semiconductor Mfg Co Ltd | Integrated defect yield management system for semiconductor manufacturing |
US6148099A (en) * | 1997-07-03 | 2000-11-14 | Neopath, Inc. | Method and apparatus for incremental concurrent learning in automatic semiconductor wafer and liquid crystal display defect classification |
US5966459A (en) * | 1997-07-17 | 1999-10-12 | Advanced Micro Devices, Inc. | Automatic defect classification (ADC) reclassification engine |
US5862055A (en) * | 1997-07-18 | 1999-01-19 | Advanced Micro Devices, Inc. | Automatic defect classification individual defect predicate value retention |
US5946213A (en) * | 1997-10-22 | 1999-08-31 | Advanced Micro Devices, Inc. | Intelligent adc reclassification of previously classified propagator defects |
US6185511B1 (en) * | 1997-11-28 | 2001-02-06 | Advanced Micro Devices, Inc. | Method to accurately determine classification codes for defects during semiconductor manufacturing |
US5999003A (en) * | 1997-12-12 | 1999-12-07 | Advanced Micro Devices, Inc. | Intelligent usage of first pass defect data for improved statistical accuracy of wafer level classification |
US6200823B1 (en) * | 1999-02-09 | 2001-03-13 | Advanced Micro Devices, Inc. | Method for isolation of optical defect images |
-
1998
- 1998-05-11 US US09/075,254 patent/US6408219B2/en not_active Expired - Lifetime
-
1999
- 1999-05-04 EP EP99920382A patent/EP1080493A1/en not_active Withdrawn
- 1999-05-04 WO PCT/US1999/009933 patent/WO1999059200A1/en not_active Application Discontinuation
- 1999-05-04 JP JP2000548916A patent/JP2002515650A/en active Pending
- 1999-05-04 KR KR1020007012589A patent/KR20010043501A/en not_active Application Discontinuation
- 1999-06-11 TW TW088107554A patent/TW509962B/en not_active IP Right Cessation
Cited By (32)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6701477B1 (en) * | 2000-06-09 | 2004-03-02 | Hueristics Physics Laboratories | Method for identifying the cause of yield loss in integrated circuit manufacture |
US6658361B1 (en) * | 2001-10-10 | 2003-12-02 | Lsi Logic Corporation | Heaviest only fail potential |
KR100956959B1 (en) | 2002-03-21 | 2010-05-11 | 어플라이드 머티어리얼즈 인코포레이티드 | Correlation of end-of-line data mining with process tool data mining |
WO2003081663A2 (en) * | 2002-03-21 | 2003-10-02 | Applied Materials, Inc. | Correlation of end-of-line data mining with process tool data mining |
WO2003081663A3 (en) * | 2002-03-21 | 2004-01-22 | Applied Materials Inc | Correlation of end-of-line data mining with process tool data mining |
US7401066B2 (en) | 2002-03-21 | 2008-07-15 | Applied Materials, Inc. | Correlation of end-of-line data mining with process tool data mining |
US20030182252A1 (en) * | 2002-03-21 | 2003-09-25 | Applied Materials, Inc. | Correlation of end-of-line data mining with process tool data mining |
CN1324515C (en) * | 2002-03-21 | 2007-07-04 | 应用材料有限公司 | Correlation of end-of-line data mining with process tool data mining |
US20060100730A1 (en) * | 2002-07-12 | 2006-05-11 | Parkes Alan S | Method for detection and relocation of wafer defects |
US7023541B2 (en) * | 2002-10-24 | 2006-04-04 | Renesas Technology Corp. | Device inspecting for defect on semiconductor wafer surface |
US20040080742A1 (en) * | 2002-10-24 | 2004-04-29 | Mitsubishi Denki Kabushiki Kaisha | Device inspecting for defect on semiconductor wafer surface |
US20070004051A1 (en) * | 2002-12-17 | 2007-01-04 | Yasuhiro Okumoto | Processing method and device |
US20100067781A1 (en) * | 2003-11-19 | 2010-03-18 | Kla-Tencor Corporation | Process Excursion Detection |
US7831409B2 (en) * | 2004-06-14 | 2010-11-09 | Semiconductor Manufacturing International (Shanghai) Corporation | Method and system for yield similarity of semiconductor devices |
US20080221831A1 (en) * | 2004-06-14 | 2008-09-11 | Semiconductor Manufacturing Intenational (Shanghai) Corporation | Method and system for yield similarity of semiconductor devices |
US7463941B2 (en) * | 2005-04-05 | 2008-12-09 | Kabushiki Kaisha Toshiba | Quality control system, quality control method, and method of lot-to-lot wafer processing |
US20060235560A1 (en) * | 2005-04-05 | 2006-10-19 | Akira Ogawa | Quality control system, quality control method, and method of lot-to-lot wafer processing |
US7650199B1 (en) * | 2005-08-03 | 2010-01-19 | Daniel Kadosh | End of line performance prediction |
US20070135956A1 (en) * | 2005-12-13 | 2007-06-14 | Taiwan Semiconductor Manufacturing Co., Ltd. | Data location systems and methods |
US7386420B2 (en) * | 2006-06-02 | 2008-06-10 | United Microelectronics Corp. | Data analysis method for integrated circuit process and semiconductor process |
US20070282544A1 (en) * | 2006-06-02 | 2007-12-06 | Zhang Guohai | Data analysis method for integrated circuit process and semiconductor process |
US20120136470A1 (en) * | 2009-05-22 | 2012-05-31 | Aurora Control Technologies, Inc. | Process for improving the production of photovoltaic products |
EP2663998A4 (en) * | 2011-01-11 | 2016-06-15 | Kla Tencor Corp | Methods and systems of object based metrology for advanced wafer surface nanotopography |
US8989478B2 (en) | 2011-07-29 | 2015-03-24 | Kla-Tencor Corporation | Method and system for visualization of semiconductor wafer inspection data acquired in a photovoltaic cell production process |
US20150039118A1 (en) * | 2013-08-01 | 2015-02-05 | Applied Materials, Inc. | Event processing based system for manufacturing yield improvement |
US10120372B2 (en) * | 2013-08-01 | 2018-11-06 | Applied Materials, Inc. | Event processing based system for manufacturing yield improvement |
CN106653639A (en) * | 2016-12-26 | 2017-05-10 | 中航(重庆)微电子有限公司 | Sampling method of semiconductor measurement site |
US11346882B2 (en) * | 2017-11-03 | 2022-05-31 | Tokyo Electron Limited | Enhancement of yield of functional microelectronic devices |
US11435393B2 (en) | 2017-11-03 | 2022-09-06 | Tokyo Electron Limited | Enhancement of yield of functional microelectronic devices |
US10955832B2 (en) * | 2018-10-09 | 2021-03-23 | Applied Materials, Inc. | Adaptive chamber matching in advanced semiconductor process control |
US11177111B2 (en) | 2019-04-19 | 2021-11-16 | Hitachi High-Tech Corporation | Defect observation device |
CN117878025A (en) * | 2024-03-11 | 2024-04-12 | 江苏芯德半导体科技有限公司 | Method for automatically clamping and controlling defective chips on wafer map |
Also Published As
Publication number | Publication date |
---|---|
JP2002515650A (en) | 2002-05-28 |
WO1999059200A1 (en) | 1999-11-18 |
KR20010043501A (en) | 2001-05-25 |
TW509962B (en) | 2002-11-11 |
EP1080493A1 (en) | 2001-03-07 |
US6408219B2 (en) | 2002-06-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6408219B2 (en) | FAB yield enhancement system | |
US5801965A (en) | Method and system for manufacturing semiconductor devices, and method and system for inspecting semiconductor devices | |
US6546308B2 (en) | Method and system for manufacturing semiconductor devices, and method and system for inspecting semiconductor devices | |
US6542830B1 (en) | Process control system | |
US6496958B1 (en) | Yield prediction and statistical process control using predicted defect related yield loss | |
US10234401B2 (en) | Method of manufacturing semiconductor devices by using sampling plans | |
US7346470B2 (en) | System for identification of defects on circuits or other arrayed products | |
US20210193537A1 (en) | Method and machine for examining wafers | |
US6556955B2 (en) | Method of determining lethality of defects in circuit pattern inspection, method of selecting defects to be reviewed, and inspection system of circuit patterns involved with the methods | |
KR20010029984A (en) | Real time defect source identification | |
CN107407648B (en) | Adaptively create disturbances to filter | |
US10970834B2 (en) | Defect discovery using electron beam inspection and deep learning with real-time intelligence to reduce nuisance | |
US6430572B1 (en) | Recipe management database system | |
JP3641604B2 (en) | Method for adjusting lithography tool | |
US20200132610A1 (en) | Algorithm selector based on image frames | |
JP2007538383A (en) | Defect detection and control method for ion implantation process and execution system thereof | |
JP2004198436A (en) | Defect inspection method and its device | |
US6828776B2 (en) | Method for analyzing defect inspection parameters | |
US11295432B2 (en) | Broad band plasma inspection based on a nuisance map | |
US6272393B1 (en) | Efficient tool utilization using previous scan data | |
Ben-Porath et al. | Advanced process development and control based on a fully automated SEM with ADC | |
JP2022187070A (en) | Image recognition system | |
Pilon et al. | Intelligent line monitor: maximum productivity through an integrated and automated line monitoring strategy |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: APPLIED MATERIALS, INC., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:YARON, GAD;LAMEY, PATRIC;MAIMON, AMOTZ;REEL/FRAME:009433/0705;SIGNING DATES FROM 19980716 TO 19980726 |
|
AS | Assignment |
Owner name: APPLIED MATERIALS INC., CALIFORNIA Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE NAME OF THE ASSIGNOR, FILED ON 08-31-98 RERECORD ON REEL 9433 FRAME 0705;ASSIGNORS:LAMEY, PATRICK H., JR.;MAIMON, AMOTZ;YARON, GAD;REEL/FRAME:009537/0275;SIGNING DATES FROM 19980716 TO 19980726 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
FPAY | Fee payment |
Year of fee payment: 12 |