CN1595607A - Manufacturing system and method for semiconductor device - Google Patents
Manufacturing system and method for semiconductor device Download PDFInfo
- Publication number
- CN1595607A CN1595607A CNA2004100746155A CN200410074615A CN1595607A CN 1595607 A CN1595607 A CN 1595607A CN A2004100746155 A CNA2004100746155 A CN A2004100746155A CN 200410074615 A CN200410074615 A CN 200410074615A CN 1595607 A CN1595607 A CN 1595607A
- Authority
- CN
- China
- Prior art keywords
- mentioned
- quality
- processing unit
- information
- semiconductor substrate
- 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.)
- Pending
Links
- 239000004065 semiconductor Substances 0.000 title claims abstract description 320
- 238000000034 method Methods 0.000 title claims abstract description 253
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 196
- 238000012545 processing Methods 0.000 claims abstract description 359
- 230000008569 process Effects 0.000 claims abstract description 193
- 239000000758 substrate Substances 0.000 claims abstract description 97
- 238000007689 inspection Methods 0.000 claims abstract description 57
- 230000003647 oxidation Effects 0.000 claims description 105
- 238000007254 oxidation reaction Methods 0.000 claims description 105
- 238000007726 management method Methods 0.000 claims description 63
- 238000004092 self-diagnosis Methods 0.000 claims description 60
- 238000012360 testing method Methods 0.000 claims description 57
- 238000004088 simulation Methods 0.000 claims description 35
- 230000008859 change Effects 0.000 claims description 34
- 238000004364 calculation method Methods 0.000 claims description 24
- 230000007547 defect Effects 0.000 claims description 18
- 238000011002 quantification Methods 0.000 claims description 16
- 230000008439 repair process Effects 0.000 claims description 15
- 238000005259 measurement Methods 0.000 claims description 8
- 238000005530 etching Methods 0.000 claims description 7
- 239000000284 extract Substances 0.000 claims description 7
- 230000010354 integration Effects 0.000 claims description 6
- 238000003745 diagnosis Methods 0.000 claims description 3
- 238000012372 quality testing Methods 0.000 claims description 3
- 239000010409 thin film Substances 0.000 abstract 1
- 235000012431 wafers Nutrition 0.000 description 234
- 239000010408 film Substances 0.000 description 76
- 238000003860 storage Methods 0.000 description 39
- 230000000875 corresponding effect Effects 0.000 description 33
- 230000002950 deficient Effects 0.000 description 15
- 230000007774 longterm Effects 0.000 description 14
- 230000005856 abnormality Effects 0.000 description 13
- 239000012528 membrane Substances 0.000 description 13
- 238000009792 diffusion process Methods 0.000 description 12
- 238000006243 chemical reaction Methods 0.000 description 11
- 239000013256 coordination polymer Substances 0.000 description 11
- 238000012937 correction Methods 0.000 description 11
- 238000012423 maintenance Methods 0.000 description 11
- 238000001514 detection method Methods 0.000 description 10
- 230000001276 controlling effect Effects 0.000 description 9
- 238000009434 installation Methods 0.000 description 9
- 238000012546 transfer Methods 0.000 description 9
- 238000011144 upstream manufacturing Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 239000012535 impurity Substances 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 238000000227 grinding Methods 0.000 description 7
- 230000004044 response Effects 0.000 description 7
- 238000012544 monitoring process Methods 0.000 description 6
- 206010061619 Deformity Diseases 0.000 description 5
- 239000013071 indirect material Substances 0.000 description 5
- 238000001259 photo etching Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 238000012550 audit Methods 0.000 description 4
- 230000015572 biosynthetic process Effects 0.000 description 4
- 150000001875 compounds Chemical class 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 238000009713 electroplating Methods 0.000 description 4
- 230000000977 initiatory effect Effects 0.000 description 4
- 239000000463 material Substances 0.000 description 4
- 239000000203 mixture Substances 0.000 description 4
- 230000000630 rising effect Effects 0.000 description 4
- 238000000427 thin-film deposition Methods 0.000 description 4
- 230000032258 transport Effects 0.000 description 4
- 238000012384 transportation and delivery Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 239000013078 crystal Substances 0.000 description 3
- 238000001312 dry etching Methods 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000010438 heat treatment Methods 0.000 description 3
- 238000010926 purge Methods 0.000 description 3
- 238000001039 wet etching Methods 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- BOTDANWDWHJENH-UHFFFAOYSA-N Tetraethyl orthosilicate Chemical compound CCO[Si](OCC)(OCC)OCC BOTDANWDWHJENH-UHFFFAOYSA-N 0.000 description 2
- 238000000137 annealing Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000000151 deposition Methods 0.000 description 2
- 239000013070 direct material Substances 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000008676 import Effects 0.000 description 2
- 238000002347 injection Methods 0.000 description 2
- 239000007924 injection Substances 0.000 description 2
- 239000011229 interlayer Substances 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 239000010410 layer Substances 0.000 description 2
- 238000003754 machining Methods 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000002203 pretreatment Methods 0.000 description 2
- 239000000523 sample Substances 0.000 description 2
- 230000002000 scavenging effect Effects 0.000 description 2
- 229910021332 silicide Inorganic materials 0.000 description 2
- FVBUAEGBCNSCDD-UHFFFAOYSA-N silicide(4-) Chemical compound [Si-4] FVBUAEGBCNSCDD-UHFFFAOYSA-N 0.000 description 2
- 238000004528 spin coating Methods 0.000 description 2
- 238000001771 vacuum deposition Methods 0.000 description 2
- JJLJMEJHUUYSSY-UHFFFAOYSA-L Copper hydroxide Chemical compound [OH-].[OH-].[Cu+2] JJLJMEJHUUYSSY-UHFFFAOYSA-L 0.000 description 1
- 101001047746 Homo sapiens Lamina-associated polypeptide 2, isoform alpha Proteins 0.000 description 1
- 101001047731 Homo sapiens Lamina-associated polypeptide 2, isoforms beta/gamma Proteins 0.000 description 1
- 102100023981 Lamina-associated polypeptide 2, isoform alpha Human genes 0.000 description 1
- 238000003326 Quality management system Methods 0.000 description 1
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 1
- 230000035508 accumulation Effects 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000003556 assay Methods 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000003990 capacitor Substances 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000011960 computer-aided design Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000008020 evaporation Effects 0.000 description 1
- 238000001704 evaporation Methods 0.000 description 1
- 238000013401 experimental design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000005468 ion implantation Methods 0.000 description 1
- 150000002500 ions Chemical class 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000001590 oxidative effect Effects 0.000 description 1
- 238000013439 planning Methods 0.000 description 1
- 238000009832 plasma treatment Methods 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 238000010992 reflux Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 229910052710 silicon Inorganic materials 0.000 description 1
- 239000010703 silicon Substances 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- WVLBCYQITXONBZ-UHFFFAOYSA-N trimethyl phosphate Chemical compound COP(=O)(OC)OC WVLBCYQITXONBZ-UHFFFAOYSA-N 0.000 description 1
- 238000009966 trimming Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L21/00—Processes or apparatus adapted for the manufacture or treatment of semiconductor or solid state devices or of parts thereof
- H01L21/67—Apparatus specially adapted for handling semiconductor or electric solid state devices during manufacture or treatment thereof; Apparatus specially adapted for handling wafers during manufacture or treatment of semiconductor or electric solid state devices or components ; Apparatus not specifically provided for elsewhere
- H01L21/67005—Apparatus not specifically provided for elsewhere
- H01L21/67242—Apparatus for monitoring, sorting or marking
- H01L21/67253—Process monitoring, e.g. flow or thickness monitoring
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41875—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by quality surveillance of production
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L21/00—Processes or apparatus adapted for the manufacture or treatment of semiconductor or solid state devices or of parts thereof
- H01L21/67—Apparatus specially adapted for handling semiconductor or electric solid state devices during manufacture or treatment thereof; Apparatus specially adapted for handling wafers during manufacture or treatment of semiconductor or electric solid state devices or components ; Apparatus not specifically provided for elsewhere
- H01L21/67005—Apparatus not specifically provided for elsewhere
- H01L21/67242—Apparatus for monitoring, sorting or marking
- H01L21/67276—Production flow monitoring, e.g. for increasing throughput
-
- 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
- H01L22/26—Acting in response to an ongoing measurement without interruption of processing, e.g. endpoint detection, in-situ thickness measurement
-
- 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
- H01L22/30—Structural arrangements specially adapted for testing or measuring during manufacture or treatment, or specially adapted for reliability measurements
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Computer Hardware Design (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
- General Factory Administration (AREA)
- Exposure Of Semiconductors, Excluding Electron Or Ion Beam Exposure (AREA)
Abstract
The present invention provides a solution for interleaving data frames, in a semiconductor device manufacturing system in which the processing apparatus for conducting a process on any one of a semiconductor substrate and a thin film on a surface thereof; a self-diagnostic system for diagnosing a state of the processing apparatus; and a parameter fitting apparatus for maintaining a parameter of the self-diagnostic system when an inspection result of the semiconductor substrate having undergone the process has been determined to be correct, and for changing the parameter of the self-diagnostic system when the inspection result has been determined to be incorrect.
Description
Technical field
The present invention relates to the manufacturing system of semiconductor device and the manufacture method of semiconductor device, specially refer to the control method of manufacturing installation, and the analogy method and the analogue means that adopt the process for fabrication of semiconductor device of this device and these class methods.
Background technology
In the past, formed substrate step, formation trap, insulation, formation transistor repeatedly, formed bit line, form capacitor and form and connect up, made semiconductor device such as DRAM by semiconductor-fabricating device.This semiconductor fabrication is by photoetching treatment, etch processes, heat treatment (oxidation, annealing, diffusion), ion injection processing or film formation processing (CVD, sputter, evaporation), clean (remove resist, cleaned by solution) and inspection processing are made the appropriate combination formation.
Generally have such system, it is constant down in the atmosphere that keeps all process chambers, substrate is transported into/transports process chamber, carry out PROCESS FOR TREATMENT, the instrumentation of checking in the processing or the processing metacoxal plate obtains is checked that data send central control system to, carry out the record management and the record of substrate and process chamber, output is carried out the suitable indication (referring to patent documentation 1) of self diagnosis to chambers and manufacturing installation.
The manufacturing installation of existing semiconductor device has the oxidation furnace of heat-treating, the oxidation furnace controller of controlling this oxidation furnace and is connected the thickness of oxidation film controller that carries out machining control with the oxidation furnace controller with oxidation furnace.
This thickness of oxidation film controller comprises the thickness of oxidation film calculating part with thickness of oxidation film computing function and has the calculating thickness detection unit that calculates the thickness decision-making function, it is in utilizing the predetermined semiconductor fabrication of thermal chemical reaction, the initial setting of carrying out based on predefined procedure begins semiconductor fabrication, carry out the predetermined system's atmosphere state and the variation thereof of thermal chemical reaction by preset time measuring space and analysis, this analysis result is fed back to semiconductor fabrication (referring to patent documentation 2).
But in the manufacturing installation of this existing semiconductor device, promptly enable to carry out the self diagnosis and the process simulation of each processing device, the thickness of the oxide-film on the semiconductor substrate and wiring width and diffusion of impurities concentration are controlled, but, can produce difference between its rate of finished products and the rate of finished products of trying to achieve by simulation to the wafer finished through a plurality of processing devices semiconductor device by the probe detection assay.Must adjust the production schedule of semiconductor device and user's order number for this reason, spend the production schedule work of appending that the labour carries out trouble continually repeatedly, increase the inspection operation of wafer, spin out the manufacturing schedule of semiconductor device.
[patent documentation 1] WO96/25760 communique (the 36th page ,～37 pages of the 25th row, the 2nd row).
[patent documentation 2] spy open the 2002-299336 communique (the 11st hurdle, from the 5th the row to the 48th the row, Fig. 1).
Summary of the invention
The invention provides such system for manufacturing semiconductor device and method, semi-conductor device manufacturing method, they can manage interval maintenance time of various processing devices rightly, prolong the running time of each processing device, also can cut down the wafer inspection step that each semiconductor fabrication ending phase is implemented.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed of semiconductor substrate; The processing control apparatus of control and treatment device; Carry out the processed of semiconductor substrate, monitor the state of processing unit, the internal information of integral processing device is carried out the simulation of processed, infers the real time simulator of the progress of semiconductor substrate processing according to processed.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed that adopts semiconductor substrate; The self-diagnosable system of inferring mass value from processing unit receiving system information calculations processed; Check processed result's testing fixture; The audit by comparison result with infer mass value, when to inferring mass value when making effective judgements, keep the parameter of self-diagnosable system, when inferring the CALCULATION OF PARAMETERS machine that then changes self-diagnosable system when mass value is invalid judgement.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed that adopts semiconductor substrate; Carry out the self-diagnosable system of processing unit from processing unit receiving system information from management; Check processed result's testing fixture; Whether repair automatically according to check result determination processing device, keep the parameter of self-diagnosable system when result of determination when effectively judging and work as the CALCULATION OF PARAMETERS machine that then changes self-diagnosable system when result of determination is invalid judgement.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed of semiconductor substrate; Self-diagnosis system based on self diagnosis parameter diagnosis processing unit; Check the testing fixture of processed; Be connected with self-diagnosis system and testing fixture, when the check result of semiconductor substrate is effective, keep the self diagnosis parameter and when check result is invalid change self diagnosis CALCULATION OF PARAMETERS machine.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the semiconductor substrate processed; Obtain the device information of processing unit, infer the quality of semiconductor substrate quality and infer portion; To the quality examination device that carries out quality examination through the semiconductor substrate of processed; Relatively quality is inferred the comparator of the quality management information of inferring qualitative data and the actual measurement of quality examination device of portion.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed of semiconductor substrate; Obtain the device information of processing unit, the quality of inferring quality information of output semiconductor substrate is inferred portion; Export the quality examination device of the quality information of treated semiconductor substrate; Infer quality information and quality information the quality relevant treatment infer quality management sector; Simulate the yield prediction device of prediction finished semiconductor device product rate according to the quality information of inferring of inferring quality management sector's output; Check processing through processing unit, at least be the rate of finished products testing fixture of having finished the finished semiconductor device product rate after the wafer operation, carry out the rate of finished products relevant treatment that the rate of finished products to the rate of finished products of yield prediction device and rate of finished products testing fixture compares.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed of semiconductor substrate; Export the quality examination device of the quality information of treated semiconductor substrate; According to simulate the yield prediction device of prediction finished semiconductor device product rate from the quality information of quality examination device; Check processing through processing unit, at least be the rate of finished products testing fixture of having finished the finished semiconductor device product rate after the wafer operation, carry out the rate of finished products relevant treatment that the rate of finished products to the rate of finished products of yield prediction device and rate of finished products testing fixture compares.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed of semiconductor substrate; Monitor the device information of processing unit, the quality of inferring quality information of output semiconductor substrate is inferred portion; Export the quality examination device of the quality information of treated semiconductor substrate; Infer quality information and quality information the quality relevant treatment infer quality management sector; According to simulate the first yield prediction device of prediction finished semiconductor device product rate from the quality information of quality examination device; According to from the quality information of inferring of inferring quality management sector, except that the simulation of the first yield prediction device, carry out the second yield prediction device of simulation and forecast finished semiconductor device product rate; Semiconductor fabrication has been passed through in inspection, at least be the rate of finished products testing fixture of having finished the finished semiconductor device product rate of wafer operation, compare the second rate of finished products relevant treatment of the rate of finished products of the rate of finished products of the first rate of finished products relevant treatment, the comparison second yield prediction device of the rate of finished products of the rate of finished products of the first yield prediction device and rate of finished products testing fixture and rate of finished products testing fixture.
A kind of form of the present invention is the manufacturing system of semiconductor device, and it has: the processing unit of carrying out the processed of semiconductor substrate; The device information that detects processing unit is extracted the characteristic quantity that the non-controlling value of handling is carried out characteristic quantification from detected value, with compare by intrinsic defect image being carried out the characteristic quantity that category classification quantizes in the correlation table of pre-stored, the characteristic quantity of judging this characteristic quantification be can with characteristic quantity in this correlation table stage with the value of journeyization, Interrupt Process is criticized semiconductor substrate the quality testing portion of notice as substandard products.
A kind of form of the present invention is the manufacturing system of semiconductor device, it has the quality of inferring automatic renewing device, this updating device has and quantizes characteristic quantity in the correlation table of pre-stored can not criticize with the substandard products of the new feature amount of journeyization the time by the inherent shortcoming image being carried out category classification when detecting, the processing record information of the semiconductor substrate of criticizing according to these substandard products, extract these semiconductor substrate substandard products of expression and criticize the device information of processing device condition when generating, this device information is fed back to quality automatically infer portion.
A kind of form of the present invention is the manufacture method of semiconductor device, the method by processing unit carry out the processed of semiconductor substrate, the processed of carrying out semiconductor substrate, the state that monitors processing unit, integral processing device internal information, carry out the processed simulation, infer the carrying out that semiconductor substrate is handled by processed.
The present invention proposes such system for manufacturing semiconductor device and method, semi-conductor device manufacturing method, they can manage interval maintenance time of various processing devices rightly, prolong the running time of each processing device, also can cut down the wafer inspection operation that each semiconductor fabrication ending phase is implemented, shorten during making the manufacturing process of semiconductor device.
Description of drawings
Fig. 1 is the system diagram of manufacturing system model of the semiconductor device of the present invention's first form of implementation.
Fig. 2 is the flow chart of manufacture method model of the semiconductor device of explanation the present invention first and second form of implementation.
Fig. 3 is the block diagram of manufacturing system model of the semiconductor device of the present invention's the 3rd form of implementation.
Fig. 4 is the flow chart of manufacture method model of the semiconductor device of explanation the present invention the 4th form of implementation.
Fig. 5 is the flow chart of manufacture method model of the semiconductor device of explanation the present invention the 5th form of implementation.
Fig. 6 is the key diagram of the method for the manufacturing installation that specific generation substandard products are criticized in the semiconductor device of the present invention's the 6th form of implementation.
Fig. 7 is the flow chart of manufacture method model of the semiconductor device of explanation the present invention the 7th form of implementation.
Fig. 8 is the system diagram of manufacturing system of the semiconductor device of the present invention's the 8th form of implementation.
Fig. 9 is the flow chart of manufacture method of the semiconductor device of the present invention's the 8th form of implementation.
Figure 10 is the profile of the used semiconductor device of the present invention's the 8th form of implementation.
Figure 11 is the key diagram of the used decision method of the present invention's the 8th form of implementation.
Figure 12 is the key diagram of the used decision method of the present invention's the 8th form of implementation.
Figure 13 is the key diagram of the used decision method of the present invention's the 8th form of implementation.
Figure 14 is the used temperature rising curve figure of the present invention's the 8th form of implementation.
The meaning of each label is as follows among the figure:
5a, self-diagnosis system; 10, manufacturing system; 11, computer; 12, storage device; 13, database; 14, processing unit; 15, direct material-indirect material; 16, gate; 17, wafer; 18, link; 19, testing fixture; 20, mask; 21, detect and handle; 22, determination processing; 23, phenomenon is judged; 24, determination processing; 25, design information; 30, manufacturing system; 31, self-diagnosable system; 32, warning device; 33, the parameter feedback device; 34, self-diagnosable system; 35, the yield prediction system; 36, the rate of finished products collection unit; 37, request portion; 38, maintenance indication picture; 39, disfigurement discovery report portion; 40, delivery forecasts portion; 43, treatment process; 44, check operation; 44c checks operation; 45, check operation; 46, treatment process; 47, criticize and finish operation; 48, the D/S treatment process; 51, manufacturing system; 52, the oxidation furnace controller; 53, the thickness of oxidation film real time simulator; 54, oxidation furnace; 56, the thickness of oxidation film calculating part; 61, quality is inferred portion; 65, database; 66, infer quality management sector; 68, quality is inferred portion; 69, the device simulation device; 70, device performance yield prediction portion; 75, yield prediction is handled; 77, analyzer.
Embodiment
(first form of implementation)
The manufacturing system of the semiconductor device of the present invention's first form of implementation is to adjust the system of semiconductor fabrication stage inspection frequency, has as shown in Figure 1: carry out the processing unit of handling 14 for wafer 17 or its lip-deep film as semiconductor substrate; As make processing unit 14 carry out the computer 11a of the self-diagnosable system of management certainly by the EES of plant engineering system (EES is made in following letter); Whether the check result determination processing device 14 according to the wafer 17 of 19 pairs of treated semiconductor substrates of testing fixture will be repaired automatically, when result of determination is to keep the parameter (or little withering) of self-diagnosable system when effectively judging (or suitable), when result of determination is invalid judgement (or inappropriate), then change the parameter (for example increase and check number of times) of self-diagnosable system, be fit to the computer 11 of device as the parameter under this situation.
Here " EES " is from processing unit 14 acquisition device information, and the data of device information are carried out the statistical analysis, and the situation of determination processing device 14 is normal or unusual system.
EES carries out on the computer 11a as self-diagnosable system, obtains the internal state of the processing unit 14 of handling wafer 17 in real time.Owing to can infer the internal state of processing unit 14, so also can obtain the machining state of the wafer 17 of inter-process in real time.
As the computer 11a of self-diagnosable system, the processing that processing unit 14 can be carried out through the time change corresponding and be stored among the database 13a with wafer 17.
Specifically, the manufacturing system 10 of semiconductor device has: the computer 11 of controlling this manufacturing system integral body; Be connected with computer 11, store the storage device 12 of the data processing algorithm relevant with semiconductor fabrication; Storage is used for the APC (senior process controller is designated hereinafter simply as APC) of computer 11 connections and the master data base 13 of the middle data of using of MES (manufacturing execution system is designated hereinafter simply as MES); Handle processing unit 14 as the wafer 17 of semiconductor substrate; Check the testing fixture 19 of the wafer 17 that this processing device 14 is handled.
The APC here is meant the contents processing of the wafer of handling according to processing unit 14 17, system by computer 11 change semiconductor fabrications, when checking that by testing fixture 19 result of wafer 17 does not reach the quality of expectation, reference treatment conditions in the past change to new treatment conditions with the treatment conditions of processing unit 14.When so-called " MES " then is meant by many batches of wafers 17 of processing unit 14 processing, carry out the production management of semiconductor device by computer 11, the a collection of wafer of selecting 17 is flowed to processing unit 14, in processing unit 14, handle, carry out the system that checks by testing fixture 19 again.
For illustrated processing unit 14 is oversimplified it as the single form illustration, but understand easily that from the manufacturing process of current LSI the manufacturing system 10 of semiconductor device generally is provided with many processing unit 14 more than 10.These many processing unit 14 are connected with computer 11a through holding wire 14a, the management certainly of practicable EES.These many processing unit 14 can directly or by computer 11a be connected with computer 11, send the device information of processing unit 14, and computer 11 can carry out unified management to semi-conductor manufacturing system integral body according to the device information that receives under the management of APC, MES.
Above-mentioned processing unit 14 can be corresponding with the various processing unit of processed semiconductor device, for example obviously can be with following processing unit as corresponding object: carry out the one-tenth membrane treatment appts that film forming is handled; The diffusion processing apparatus of impurity; The thin film deposition device of CVD; Make the heating furnace device of backflows (fusion) such as psg film, bsg film, bpsg film (dielectric film); The thermal chemical reaction processing unit of fine and close amount, silicide film (electrode) thickness etc. such as adjusting CVD oxide-film; The sputter equipment of depositing metal wiring layer and vacuum deposition apparatus; The electrolytic plating apparatus of electroplating in addition in addition; Semiconductor substrate is carried out the CMP processing unit of grinding chemistry, machinery; Semiconductor substrate surface is carried out etched doing or the processing unit of wet etching; The spin coating device relevant with photoetching treatment; Exposure-processed devices such as stepping; The electrode that is cut into the semiconductor device of sheet is connected to adds the processing unit that the bonding line processing unit carries out being applied to various semiconductor fabrications on the lead frame.
The manufacturing system 10 of semiconductor device of the present invention can be applied to batch formula device or one chip device.All forms of implementation described later can be used for batch formula device or one chip device too.
Store manufacturing process data among the database 13a of computer 11a inside, upgrade the contents processing data of the processing resume of contents processing data relevant and relevant wafer 17 one by one with the lot number of wafer 17, the optimization process state offers processing unit 14 during with current (in real time), can whether change to detect to processing unit 14 inside simultaneously and handle 21, this internal state is fed back to semiconductor fabrication.
For example when processing unit 14 be vacuum flush system one-tenth membrane treatment appts, diffusion processing apparatus, thin film deposition device and so on have the processing unit of chamber the time, then, carry out semiconductor fabrication according to the various parameters of all conditions such as aperture of the valve of the pressure of the temperature of temperature, the base-plate temp in a plurality of places in the decision stove, chamber outer wall many places, the indoor vacuum degree of expression, gas flow, control gaseous flow.
When processing unit 14 be plasma process system device for dry etching, ion implantation apparatus and so on have the processing unit of electrode the time, then except the various parameters of above-mentioned vacuum flush system, also to carry out semiconductor fabrication according to the various parameters of positional information of the matched position of RF, RF voltage (row wave voltage, reflected wave voltage), wafer and so on.
When processing unit 14 is wet etching process device, spin-coat process device, stepping exposure-processed device, the lead-in wire bonding processing unit of atmospheric pressure treatment system, then, carry out semiconductor fabrication based on the various parameters of processing time or positional information of wafer or chip and so on.
In the manufacturing system 10 of semiconductor device, processing unit 14 is into the situation that membrane treatment appts, diffusion processing apparatus, thin film deposition device and so on are used gas and compound, this gas and compound are supplied with by gate 16, direct material-indirect material 15 that the direct material of gas supplied and compound and so on and container and so on indirect material is formed quantizes,, in master data base 13, carry out as the material model data record from management.By this material modelization, can be to gas and compound and so on material whether to the influential determination processing 22 in real time of carrying out of semiconductor fabrication.
As mentioned above, the manufacturing system 10 of the semiconductor device of form of implementation of the present invention is owing to being the APC system that the processing modelization of processing unit 14, direct material, indirect material 15, wafer 17 is constituted computer 11, and can carry out real-time quality management, owing to can implement TCAD (process computer Aided Design) and YMS (rate of finished products management system), just advantageously predict final finished semiconductor device product rate in the intermediate treatment operation before last procedure finishes.Below with the feature of the manufacturing system 10 of the semiconductor device of first to the 7th form of implementation explanation the invention process form.
Fig. 2 is the model flow chart of manufacture method of the semiconductor device of explanation the present invention first form of implementation.The manufacturing system 30 of semiconductor device is according to its semiconductor fabrication order, drop into operation 42 through wafer batch, process first treatment process 43 that A handles with respect to wafer or its surface film by processing unit as semiconductor substrate, carry out the first inspection operation 44 of the online QC processing of first testing fixture, carry out the second inspection operation 45 of the surfacial pattern defect inspection processing of second testing fixture, process second treatment process 46 that B handles to wafer or to its surface film by processing unit, also has abridged the 3rd treatment process among the figure in addition, the science and engineering preface everywhere, multiple tracks treatment process and inspection operation such as the 3rd inspection operation are finished operation 47 until criticizing of wafer state end.Criticize the formed a plurality of semiconductor device of the wafer of finishing operation 47 thus, check finished semiconductor device product rate by the rate of finished products testing fixture in D/S (tube core classification) treatment process 48 (following brief note is " D/S " treatment process).D/S handles thus, with the electrical property of pin check chip, distinguishes finished product and substandard products, can discern with the ink sign on the substandard products chip.
For example carry out first treatment process 43 of processing A and handle situation in film forming, the a succession of film forming such as discharge that wafer is applied gas injection, temperature treatment, stress management, RF voltage management, gas is handled, and simultaneously device information 43a such as the having or not of the gas pressure of processing unit, temperature, RF voltage, spark, deposit amount is sent to self-diagnosable system 31.
When first treatment process 43 is the CMP processing, the quantity of monitoring grinding agent, the revolution of grinding table, the degenerate state of grinding table, simultaneously wafer is carried out milled processed, the device information 43a such as replacing period that supply with grinding agent, the grinding table of processing unit are sent to self-diagnosable system 31.
Self-diagnosable system 31 is in order to check whole wafers in real time, a plurality of pressure detectors, Temperature Detector, spark tester and deposit film thickness detector etc. can be set in processing unit, also can constitute to be provided with to receive and show that the where there is light manages the detection signal of the unit state of device with the corresponding a plurality of detectors of semiconductor fabrication.
The event information 43b in the maintenance timing of first treatment process, 43 used processing unit, cleaning period, parts swap period and so on is sent to the self-diagnosable system 34 of carrying out second treatment process 46 that processing B handles.Can infer that based on this event information 43b current (in real time) batch wafer in first treatment process 43 is handled can proceed in second treatment process 46 with which kind of state (or characteristic).
Self-diagnosable system 31 is pressed preset time (for example 1 second interval) mensuration and analysis at interval, propelling by semiconductor fabrication is selected the stage of scheduled volume to generate in this amount and is repaired request signal automatically in processing unit 14 inner accumulations shown in Figure 1 or the deposit or the amount of consumption.This repairs request signal automatically and sends to warning device 32 as QC information 31a.In the starting stage of manufacturing system 30 running of semiconductor device, the automatic reparation request signal that is comprised among the response QC information 31a, the timing indication information 32a that warning device 32 will be repaired automatically sends to the processing unit 14 of oxidation furnace of operation in first treatment process 43 and so on.
When first treatment process 43 of processing A is illustrative is when carrying out thermal chemical reaction and handling, press preset time (for example 1 second interval) mensuration and the oxidation deposit amount of analyzing as the inner deposit of oxidation furnace of processing unit 14 at interval, reach the stage of being scheduled to the deposit amount in this oxidation deposit amount, generate and repair request signal automatically.This repairs request signal automatically and sends to warning device 32 as QC information 31a.The manufacturing system 30 of semiconductor device can constitute, contained automatic reparation request signal among the response QC information 31a, and the timing indication information 32a that will be repaired automatically by warning device 32 sends to oxidation furnace.
Illustrative oxidation furnace can the timing indication information that receives automatic reparation 32a's or in first treatment process 43 in the executory stage, when the semiconductor fabrication of 1 unit finishes, wait for automatic repair process, after finishing to transport wafer outside the oxidation furnace, purge gas is imported oxidation furnace inside when the wafer batch processing in the pre-treatment.
At this moment, the wafer that transports continues to check that by first the online QC of operation 44 enforcements handles that the QC information 44a that will comprise the thickness information that generates on the wafer sends to parameter and is fit to device 33.
Check the wafer of having finished online QC processing in the operation 44 first, by having or not graphic defects on second testing fixture inspection wafer, the QC information 45a that will comprise the defect inspection result sends to parameter and is fit to device 33 in the second inspection operation 45.In this " online QC ", mainly carry out film thickness and measure the mensuration of equal thickness directioin parameter.In addition, " defect inspection " be primarily aimed at the inspection of defective on the planar graph that forms by photo-mask process.
Parameter is fit to device 33 is fit to the model of above-mentioned self-diagnosable system 31 and its parameter.This form of implementation be fit to be meant a kind of like this function: this that can generate being received from the device information 43a of processing unit based on self-diagnosable system 31 with predetermined time interval (for example 1 second at interval) repaired timing of request signal automatically, change or reconstruct to (for example after receiving 3 batches of inappropriate signals of continuous processing unit, beginning to generate automatic reparation request signal) during suitable.
Model and the uncomfortable timing of its parameter when self-diagnosable system 31, even if in device information 43a, add the improper parameter that produces improper result in the oxidizing process that has processing unit, but but send from processing unit in batch wafer end process in pre-treatment, and the check result of the first inspection operation 44 enters normal value.In addition, second check that the check result of operation 45 also enters normal value.Improper parameter itself is wrong among the device information 43a of decidable processing unit transmission thus, and reliability is low.
This form of implementation adopts such structure to improve the reliability of the device information 43a that contains the low improper parameter of reliability, promptly carry out quality examination to sending improper parameter batch wafer constantly, device information 43a and wafer state result is obtained in statistics ground whether consistent, by the suitable device 33 of parameter the parameter 33a of correction etc. is fed back to self-diagnosable system 31.
Generally can do such control, check that by first and second operation 44,45 obtains the QC information 44a and the QC information 45a of former 10 batches of wafers,, change the model parameter 33a of self-diagnosable system 31 simultaneously one by one with the improper parameter comparison of device information 43a.
In the manufacturing system with the semiconductor device that adopts into membrane treatment appts is that example is when describing, also can carry out such control: check in the operation 44 in first and check thickness with the thickness testing fixture, in the second inspection operation 45, check defective with flaw detection apparatus, obtain each check result information, compare with the improper parameter of device information 43a, change the parameter 33a of the model of self-diagnosable system 31 one by one.
Self-diagnosable system 31 is revised the model parameter of self-diagnosable system 31 between manufacturing system 30 on-stream periods of semiconductor device, carry out the self diagnosis of processing unit.That is the state of supervision processing unit, respond the acceptance of improper signal, generate the high QC information 31a of reliability.This QC information 31a is to impel the signal that sends the timing indication information 32a that repairs automatically with respect to the warning device 32 that gives a warning to the MES that is positioned at the downstream.
For example when processing unit shown in Figure 1 14 is into membrane treatment appts,, just can reduces wash number and prevent to reduce quality, the manufacture of semiconductor device is increased by optimizing the importing number of times that cleans furnace gas.
The reception of warning device 32 response QC information 31a sends repair message 32b, informs that operating personnel repair generation automatically.For example can be in the purge chamber near the set processing unit scintillator of configuration warning usefulness notify, or also can in the purge chamber outside, monitor and show on operating personnel's watch-dog of whole semiconductor fabrication that maintenance indicates picture 38.
In addition, warning device 32 also can respond to the inspection frequency correction indication information 33b that other route receives, and controls above-mentioned first and second and checks the number of processes of operation 44,45, adjusts the inspection frequency of wafer batch.Specifically, can automatically form semiconductor fabrication again, semiconductor fabrication can be reduced in real time/stage of exhaustive test necessity along the learning curve stabilised quality, response comes autoregressive parameter to be fit to the inspection frequency correction indication information 33b of device 33, to omit the quality examination that first and second checks operation 44,45, can reduce the inspection operation, improve output as the wafer of semiconductor device.
Have again, parameter is fit to device 33 and can generates by the predetermined new disfigurement discovery report information 33c that wafer batch produced, the fact of the new disfigurement discovery of the wafer that operation exists in the middle of reporting in real time for the technical staff's who cleans outdoor activity disfigurement discovery report portion 39.
On the other hand, completion rate/rate of finished products collection unit 36 obtains D/S treatment process 48 and checks crowd rate of finished products information 48a that finishes the wafer of operation 47, for example when judging that rate of finished products is low, automatically strengthen the quality management system immediately, will show that the low rate of finished products information 36b of rate of finished products sends to completion rate/yield prediction system 35.
Completion rate/yield prediction system 35 handles the quality management of wafer in real time, with in the past rate of finished products information and current rate of finished products information 36b relatively, and the increase and decrease tendency of prediction rate of finished products.The yield prediction information 35a that will show this tendency sends to the suitable device 33 of above-mentioned parameter, can be fit to device 33 from parameter thus and will check that frequency correction indication information 33b sends to warning device 32.
Above-mentioned inspection frequency correction indication information 33b can adjust automatically, now its control method of illustration.The process capability index CP of general available processing unit by 43 operations of first treatment process adjusts the inspection frequency automatically.The thickness that forms in this wafer or the diffusion of impurities degree of depth or etch-rate etc. are stipulated each variable as desired value.If with respect to this desired value the upper limit specification of the processing unit that quantitatively increases be Su, with respect to this desired value the processing unit lower limit specification that quantitatively reduces be SI, the mean value of the processing unit that the measured value of number batch processing is average is x, expression is D with respect to the standard deviation of the processing unit of the deviation of each measured value of this mean value X, when adding up the process capability index CP of these value calculation processing apparatus, based on upper limit specification Su, process capability index CP sets up the relation of (Su-x)/3D, and based on lower limit specification SI, process capability index CP sets up the relation of (x-SI)/3D.
When above-mentioned process capability index CP for example is CP＜1.33, then reconstructs semiconductor fabrication and check frequency to increase.When process capability index CP is 1.33≤CP＜1.67, then controls to make and check that frequency remains unchanged.And when process capability index CP satisfies 1.67≤CP, then can reconstruct semiconductor fabrication with controlling automatically and check frequency to reduce.
In addition, also inspection mean value X, the upper limit specification Su of processing unit of the processing unit of 10 times thickness in the past that can transfer out according to processing unit and check result gained such as the diffusion of impurities degree of depth and etch-rate or the lower limit specification SI of processing unit come the process capability index of calculation processing apparatus.At this moment, CP is corresponding with process capability index, changes current 2 times, 1 times and 0.5 times even will check frequency, also can carry out the semiconductor fabrication management equal with above-mentioned inspection FREQUENCY CONTROL.
When oxidation furnace being come example as processing unit and the desired value of thickness is set at 10nm, upper limit specification Su can be made as 12nm, lower limit specification SI is set at 8nm, checks the control of frequency.
Completion rate/rate of finished products collection unit 36 sends to the GDS data with rate of finished products information 36a and cuts out request portion 37, can impel the mask information 37a that will cut out a used mask data part in the photo-mask process to send to above-mentioned parameter and be fit to device 33.For example the rejected region on the wafer of decision semiconductor device yield is illustrated as under the situation of locality, helps the feedback effects to the model of self-diagnosable system 31 sometimes.
(second form of implementation)
Referring now to Fig. 2, the workflow of manufacturing system 30 of the semiconductor device of second form of implementation of the present invention is described.Then omitted for parts or specification with above-mentioned first form of implementation repetition.
The manufacturing system 30 of semiconductor device has: handle by batch first treatment process 43 of the wafer that input operation 42 is sent into; What the wafer that first treatment process 43 was handled was checked first checks operation 44; To the second inspection operation of checking through the wafer of the first inspection operation 44 45; To second treatment process of handling through the wafer of the second inspection operation 45 46; Through a plurality of treatment process and inspection operation, in wafer, finish criticizing of semiconductor device and finish operation 47; Check the D/S treatment process 48 of wafer.
System for manufacturing semiconductor device 30 also has: the diagnostic system 31 that first treatment process 43 is carried out self diagnosis; Second treatment process 46 is carried out the diagnostic system 34 of self diagnosis; Completion rate/rate of finished products collection unit 36 that the rate of finished products information 48a that D/S treatment process 48 is obtained receives; Completion rate/rate of finished products collection unit 36 connects therewith, completion rate/yield prediction system 35 that the completion rate/rate of finished products of semiconductor device is predicted.
The manufacturing system 30 of semiconductor device constitutes: will be checked by D/S treatment process 48 by the above-mentioned batch of wafer of finishing operation 47, send to completion rate/rate of finished products collection unit 36, generate new rate of finished products information 36b to send to completion rate/yield prediction system 35.
The manufacturing system 30 of semiconductor device informs the semiconductor fabrication in downstream with the state of the semiconductor fabrication of upstream, to the finished semiconductor device product rate in every batch of prediction of wafer.Specifically, the item of the processing unit that disposes in the management upstream process, send this event information 43b that handles the maintenance time, scavenging period, part replacement time and so on of the replacing running stores of device to be connected self-diagnosable system 34 with the processing unit of carrying out processing by second treatment process 46, whether 43b is suitable by decision event information, the state of the processing unit of operation in first treatment process 43 can be informed the semiconductor fabrication of being carried out by second treatment process 46 in real time.
For example, be arranged at the item of the one-tenth membrane treatment appts of upstream process by management, this class event information 43b of the part replacement time of the maintenance time of the running stores of membrane treatment appts, the scavenging period that becomes membrane treatment appts, one-tenth membrane treatment appts will be replaced with, send the self-diagnosable system 34 that etch processes device that another semiconductor fabrication of operation in second treatment process 46 of lower procedure uses is connected to, just the state of the one-tenth membrane treatment appts of operation in first treatment process 43 can be reflected in real time to semiconductor fabrication by 46 execution of second treatment process.
The self-diagnosable system 31 that is disposed in the upstream process is connected with self-diagnosable system 34 in being located at lower procedure, sends batch information 49 about the quality management of the wafer handled by processing unit.
34 couples of above-mentioned event information 43b of self-diagnosable system carry out comprehensive numerical Evaluation with batch information 49.The wafer batch of being handled by second treatment process 46 is by good upstream process of processing device condition or the upstream process that worsens by processing device condition, according to numerical value they are made objective quality judging, apply the device information 46a that the processing unit of processing of processing B sends in real time based on 46 pairs of wafers of second treatment process, can carry out the high self diagnosis of reliability and handle.
Self-diagnosable system 34 application affairs information 43B, batch information 49 and the device information 46a that carries out the processing unit of handling by second treatment process 46, fortune is calculated the rate of finished products (for example 10 rate of finished products) through the wafer of second treatment process 46, then this operation result is sent to delivery forecasts portion 40.
At this moment, can not in second treatment process 46, repair by the rejected region that the improper processing of first treatment process 43 causes, thereby, the rate of finished products of the wafer during to the semiconductor fabrication lower procedure is low, so the quantity that becomes final semiconductor device certified products just can or be judged in operation prediction in the middle of the semiconductor fabrication by delivery forecasts portion 40.
Delivery forecasts portion 40 can automatically carry out the wafer input amount for the production schedule of manufacturing system 30 work plans of formulating semiconductor device with computer (not shown) and control extra urgent application processing 41, can before substandard products are found in the final inspection of wafer, the wafer batch that reclaims be sent in the wafer operation, thereby can expect to reach maximization as the semiconductor manufacturing factory man's of enterprise fund is mobile.
(the 3rd form of implementation)
The manufacturing system 51 of the semiconductor device of the 3rd form of implementation explanation, it can be corresponding with the various processing unit of processed semiconductor device, for example obviously can be with following processing unit as corresponding object: the one-tenth membrane treatment appts of carrying out the film forming processing; The diffusion processing apparatus of impurity; The thin film deposition device of CVD; Make the heating furnace device of backflows (fusion) such as psg film, bsg film, bpsg film (dielectric film); The thermal chemical reaction processing unit of fine and close amount, silicide film (electrode) thickness etc. such as adjusting CVD oxide-film; The sputter equipment of depositing metal wiring layer and vacuum deposition apparatus; The electrolytic plating apparatus of electroplating in addition in addition; Semiconductor substrate is carried out the CMP processing unit of grinding chemistry, machinery; Semiconductor substrate surface is carried out etched doing or the processing unit of wet etching; The spin coating device relevant with photoetching treatment; Exposure devices such as stepping; The electrode that is cut into the semiconductor device of sheet is connected to the processing unit that bonding wire processing unit on the lead frame etc. can be applied to various semiconductor fabrications.
Fig. 3 is the model flow chart of manufacturing system 51 of the semiconductor device of the present invention's the 3rd form of implementation.The manufacturing system 51 of semiconductor device for example comprises: carry out the heat treated oxidation furnace 54 as processed portion that utilizes thermal chemical reaction; Oxidation furnace controller 52 as processed portion control device; Control the work of this oxidation furnace controller 52 and do not work and the computer 11 of operating state; To being arranged at the thickness of oxidation film real time simulator 53 that thickness of oxidation film tolerance is calculated on the wafer between this computer 11 and the oxidation furnace 54.This thickness of oxidation film real time simulator 53 is independently to constitute with oxidation furnace controller 52 in this form of implementation, but it obviously also can integrally constitute as an one function portion in the inside of oxidation furnace controller 52.
These thickness of oxidation film real time simulator 53 response oxidation furnace controllers 52 send to the initiating signal of oxidation furnace 54, the device internal information that shows temperature and pressure in the stove in that interior oxidation thickness calculating part 56 just sends in real time according to oxidation furnace 54 begins to calculate formed thickness of oxidation film on the wafer of sending in the oxidation furnace 54.This thickness of oxidation film calculated value sends to computer 11 in real time, is used to adopt the production management of the factory of MES.
In the manufacturing installation of in the past semiconductor device, do not monitor the stop signal that makes the oxide-film calculating part stop oxidation processes, whether represented that stop signal sends the information of back stove internal state so can not differentiate current device internal information, and the manufacturing system 51 of the semiconductor device of this form of implementation shown in Figure 3, owing to be the stop signal of sending from oxidation furnace controller 52 by 53 monitoring of thickness of oxidation film real time simulator, so can continue thickness of oxidation film calculating with the device internal information that pressure descends according to the internal temperature that expression stops the oxidation furnace 54 that oxidation processes rises constantly, handle in real time.
Behind the initiating signal that thickness of oxidation film calculating part 56 reception oxidation furnace controllers 52 send, begin to calculate the thickness of oxidation film that forms on the wafer in the oxidation furnace 54 according to the device internal information, the back receives the initiating signal that oxidation furnace controller 52 sends during predetermined processing, up to passing through the scheduled time again, continue the calculating of thickness of oxidation film in real time.
What the thickness of oxidation film calculated value of thickness of oxidation film calculating part 56 was represented is each instantaneous value that calculates constantly that the device internal information sends from oxidation furnace 54, this instantaneous value is according to DOE (experimental design) model 59 of planning of experiments method, carrying out integral processing in during stop signal this section time of reception of finishing oxidation processes, just can calculate the gross thickness of the oxide-film that forms on the wafer through the initiating signal time of reception of associating.
Be in the EES data 60 additional number of the having family information of handling for computer 11 with the different of manufacturing installation in the past according to the semi-conductive manufacturing system 51 of thickness of oxidation film real time simulator 53, be under the situation that constitutes by a plurality of chambers for example in oxidation furnace inside, make each corresponding different EES data 60 of these a plurality of chambers, just can improve the computational accuracy of thickness of oxidation film.Obviously, owing to can obtain the device internal information of each chamber in real time, can help the real-time Simulation of the thickness of oxidation film of monolithic processing.
In this form of implementation since with oxidation furnace 54 through the time factor be additional in real time in the plain course of reaction of thermalization, for example at the maintenance timing of changing running stores, receive the event information in cleaning period and so on of oxidation furnace 54, just can carry out the integrating function of integral processing by thickness of oxidation film real time simulator 53 with respect to later elapsed time realization.Like this, compare, can also implement accurate quality management with the manufacturing installation of handling the semiconductor device in the past of lots of wafers by identical semiconductor fabrication process condition.
Be that semi-conductive manufacturing system 51 is illustrated as oxidation treatment device above, but the present invention is not limited to oxidation treatment device.For example also can be constructed with manufacturing system with respect to processing unit such as Etaching devices as the semiconductor device of this real-time Simulation function of the real time simulator structural element of Fig. 3 illustrative " endpoint monitoring+etch-rate is inferred ".
(the 4th form of implementation)
Fig. 4 is the model flow chart of manufacture method of the semiconductor device of explanation the present invention the 4th form of implementation, wherein first treatment process 43 of implementation processing A, the first inspection operation 44 of carrying out online QC, the second inspection operation 45 of implementation defect inspection, second treatment process 46 of carrying out processing B, the disfigurement discovery report 39a of portion that informs the technical staff, QC information 44a, 45a, repair message 32b etc. are identical with above-mentioned first form of implementation, the omission duplicate explanation.
The difference of quality management value and measured value is inferred in the manufacturing system utilization of semiconductor device shown in this form of implementation, constitutes in order to diagnose the model of each processing unit, various transducer and self-diagnosable system.Quality is inferred portion 61 and is received the device information 43a that comprises a plurality of variable X (for example EES data) that is carried out the processing unit of handling by first treatment process 43, contained a plurality of variable X among the model of the self-diagnosable system of representing according to function f and parameter a thereof, parameter b and the device information 43a that receives in real time, calculating comprise corresponding to a plurality of output Y data of the wafer batch in the current thermochemical reaction process infer qualitative data 61a, this is inferred qualitative data 61a sends to the comparator 62 that plays the abnormal detector function.
This infer in the face that comprises the wafer oxide thickness among the qualitative data 61a distribute, the D data of the standard deviation value of the μ data of the mean value of thickness of oxidation film, expression thickness of oxidation film deviation, can real-time quality of semiconductor devices information be offered the computer 11 that the execution MES that is connected with comparator 62 controls.
This computer 11 is checked the online QC processing unit of operation 44 with implementation first and is carried out second and check that the flaw detection apparatus of operation 45 is connected, according to infer qualitative data 61a respectively with first check operation 44 QC information 44a and with the second QC information 45a different information relatively of checking operation 45, check that with first operation 44 or second checks that the optimization information 64 of the inspection frequency of operation 45 exports to online QC processing unit or flaw detection apparatus.Can omit the first inspection operation 44 or second like this and check operation 45 and the manufacturing engineering time of shortening semiconductor device.
Quality is inferred portion 61 and is for example received based on model and parameter information 63 as the self-diagnosable system of the Mahalanobis generalised distance of the quality management data of storage in the QC database 65, carry out the calculation process of the oxidation furnace internal state of processed, calculation qualitative data 61a by time dependent first treatment process 43.Infer qualitative data 61a and send to comparator 62 as mentioned above, also send to by first treatment process 43 as feedback information 61b concurrently and carry out the oxidation furnace of handling, send to the processing unit that the etch processes device that is arranged in this first treatment process, 43 lower procedures, second treatment process 46 processing B or photoetching treatment device and so on are carried out the semiconductor fabrication beyond the oxidation processes as feed-forward information 61c.
Comparator 62 constitute reception from quality infer portion 61 infer qualitative data 61a, from implement the first QC information 44a that checks the testing fixture of operation 44, from implement second check operation 45 testing fixture QC information 45a and from the repair message 32b of warning device 32.
Like this, comparator 62 can be inferred portion 61 from quality and be received and infer qualitative data 61a with what device information 43a and model and parameter information 63 carried out calculation process, can check that the operation 44 and the second inspection operation 45 receive measured value from each testing fixture of carrying out processed first, relatively more relevant with quality management presumed value and measured value, the while can be carried out the inferring of parameter of self-diagnosable system.In other words, can make the parameter of the self-diagnosable system of inferring be similar to measured value, infer parameter a, the parameter b that portion 61 sends the self-diagnosable system of revising from comparator 62 to quality simultaneously.
The semiconductor fabrication that quality YS feeds back to first treatment process 43 of inferring of quality being inferred portion 61 output makes steady quality.To infer quality YS simultaneously feedovers to second treatment process 46, make the parameter of processing unit of the processed of carrying out second treatment process 46 consistent thus with the quality YS that infers in advance, can carry out based on the processed of inferring quality YS to the wafer batch in the first current treatment process 43 in the stage of sending into second treatment process 46.
Comparator 62, as mentioned above owing to obtained and inferred qualitative data 61a and as QC information 44a, the 45a of measured value, the stage that can between presumed value and measured value, deviate from just, the processing unit, first of judging first treatment process 43 early check the testing fixture, second of operation 44 check operation 45 testing fixture certain or all produced improper situation and the state of the high quality management of reliability can not have been arranged.
Comparator 62 is based on inferring qualitative data 61a, check when in the testing fixture of operation 45 signal of improper situation taking place when detecting the processing unit that shows first treatment process 43 or second, device/sensor abnormality detection signal 62a can be sent to warning device 32 and operator's watch-dog, semiconductor fabrication is stopped, current quality management information being shown on operator's the maintenance indication picture 38.
(the 5th form of implementation)
Fig. 5 is the model flow chart of manufacture method of the semiconductor device of explanation the present invention the 5th form of implementation, wherein batch drop into operation 42, first treatment process 43, first are checked operation 44, second treatment process 46, batch are finished operation 47, D/S treatment process 48 and quality and infer portion 61 and equate with above-mentioned the 4th form of implementation, omit duplicate explanation.
The manufacture method of the semiconductor device of this form of implementation, comprise be arranged at the semiconductor fabrication upstream for example by first treatment process 43 that becomes membrane treatment appts to carry out, be arranged in this first treatment process 43 lower procedure second treatment process 46, through batch finishing the device performance of a plurality of semiconductor device that form on the wafer of operation 47 and the D/S treatment process 48 of rate of finished products inspection.
The device information 43a that first treatment process 43 of processing A will be carried out the processing unit of processed sends to quality and infers portion 61, and quality is inferred portion 61 and will be inferred qualitative data 61a according to device information 43a and send to and infer quality management sector 66.This infers quality management sector 66 after the batch processing of current execution processing A finishes, data 67 obtain the online QC result of wafer of the first inspection operation 44 as a comparison, can with infer qualitative data 61a relatively, can judge the reliability of inferring qualitative data 61a thus.
For example, infer quality management sector 66 quality is inferred the measured value comparison of inferring qualitative data 61a and 44 actual measurements of the first inspection operation that portion 61 infers, carry out the relevant treatment of device information 43a and measured value simultaneously.At this, the relevant treatment of this form of implementation is to instigate quality to infer the processing that qualitative data 61a is similar to measured value of inferring that portion 61 infers, the device information 43a that each batch can be obtained feeds back to the semiconductor fabrication of first treatment process 43, with can be at the stage means for correcting information 43a that handles next group and near measured value.Thereby can pass through relevant treatment, correction mass is inferred the parameter of the computation model of inferring qualitative data 61a of portion 61 rightly.
For example, infer quality management sector 66, the departing from of qualitative data 61a and measured value of inferring after carrying out 10 batches of relevant treatment reduces to first and checks that operation 44 (online QC) and the 3rd inspection operation 44c (online QC) become unwanted degree, send to and infer process and the device simulation device 69a that quality management sector 66 is connected by the rate of finished products information of first treatment process 43 being implemented the wafer batch of processing A, with regard to the rate of finished products of energy prediction processing operation 43 handled wafer batch.
Equally, the device information that second treatment process 46 of processing B will be carried out the processed of second treatment process 46 sends to quality and infers portion 68, quality is inferred portion 68 according to the device information that receives, and will send to as the QC information 68a that infers qualitative data and infer the 66a of quality management sector.This infers the 66a of quality management sector after the batch processing of current execution processing B finishes, will be as the 3rd QC information of checking the wafer inspection result data of operation 44c, data 67a obtains as a comparison, can with QC information 68a relatively.
Infer the 66a of quality management sector quality is inferred portion 68 QC information 68a that infers and the QC information 44d that is surveyed by the 3rd inspection operation 44c relatively, carry out the relevant treatment of device information and measured value simultaneously.
For example, infer the 66a of quality management sector, departing from of QC information 68a after carrying out 10 batches of relevant treatment and measured value, reduce to the 3rd of wafer that inspection transports from second treatment process 46 of implementing processing B and check that operation 44c (online QC) becomes unwanted degree, send to and infer process and the device simulation device 69a that quality management sector 66 is connected by the rate of finished products information of second treatment process 46 being implemented the wafer batch of processing B, just the rate of finished products of the wafer batch of energy prediction processing operation 46 processing.
Process and device simulation device 69a are based on as the rate of finished products information of first treatment process 43 of upstream process and rate of finished products information as second treatment process 46 of lower procedure, carry out the simulation of the device of fabrication of semiconductor device and semiconductor device, the rate of finished products information of inferring of each batch is sent to the device performance yield prediction 70a of portion in downstream.
The above-mentioned device performance yield prediction 70a of portion is according to inferring quality management sector 66 and inferring the information that quality management 66a infers, the prediction finished semiconductor device product rate, and then in this form of implementation, collect first of online QC processing gained and check that the QC information 44a and the 3rd of operation 44 checks the QC information 44d of operation 44c, can be according to measured value in intermediate treatment operation stage forecast finished semiconductor device product rate.
Process and device simulation device 69 merge the QC information 44a of the first inspection operation 44 of being located at upstream process (or prime operation) and are located at the QC information 44d that the 3rd of lower procedure (or back level operation) is checked operation 44c, the simulation of implementation and device will send to secondary device performance yield prediction portion 70 as the rate of finished products information of analog result.
Device performance yield prediction portion 70 sends to the above-mentioned device performance yield prediction 70a of portion with each yield prediction information 72 of each batch.The device performance yield prediction 70a of portion relatively infers rate of finished products information and yield prediction information 72, by carrying out finished semiconductor device product rate prediction processing 75, can also provide high-precision yield prediction; Parallel mutually with yield prediction processing 75, can set the first inspection operation 44 and the 3rd again and check the frequency of the online QC of operation 44c (for example omission or simplification are checked).
The manufacturing system of the semiconductor device of this form of implementation according to based on first treatment process 43 and second treatment process 46 separately among setting processing unit device information infer quality, the prediction finished semiconductor device product rate, predict finished semiconductor device product rate according to the check result of the wafer that transfers out from each treatment process simultaneously, parallel mutually with the rate of finished products of this prediction, can ask the rate of finished products of device performance by the D/S treatment process 48 of pin check device with respect to criticizing the wafer of finishing operation 47.
Also can be to the processing 71 of making comparisons by the rate of finished products of D/S treatment process 48 actual measurement and the rate of finished products of device performance yield prediction portion 70, simulation precision according to its difference correction device performance yield prediction portion 70, to comparing processing 73, according to the precision of its difference correction device performance yield prediction 70a of portion by the rate of finished products of D/S treatment process 48 actual measurements and the rate of finished products of the device performance yield prediction 70a of portion.The also parameter of the simulation model of trimming process and device simulation device 69,69a rightly.
Manufacturing system according to the semiconductor device shown in this form of implementation, proofread and correct the prediction of each rate of finished products based on the finished semiconductor device product rate of actual measurement, simultaneously if repeat semiconductor fabrication, because high-precision yield prediction can be arranged, so can advantageously omit the NPW (non-product wafer) of the wafer of using as quality management, also can advantageously omit inspection operation, can also advantageously carry out simulation simultaneously every wafer in each semiconductor fabrication of each batch as online QC.
(the 6th form of implementation)
Fig. 6 illustrates the method for the manufacturing installation that specific generation substandard products are criticized in the semiconductor device of the present invention's the 6th form of implementation.With the treatment process of semiconductor device among the oxidation furnace illustration figure, but among the present invention in first treatment process 43 used processing unit obviously be not limited to oxidation furnace, also applicable to the processing unit of carrying out other semiconductor fabrications.
Analyzer 77 is judged the state of wafer in the semiconductor fabrication according to the various detection data that receive.For example can collect many data from the chamber of the processing unit inside of the semiconductor fabrication of carrying out first treatment process 43, can be in the each method for making of semiconductor fabrication specific handled wafer batch.
Analyzer 77 has the ability of collecting data in real time, for example can detected 9 kinds of analogue datas of while be transformed to numerical data according to predetermined time interval, writes down chronologically in each batch and these numerical datas of analysis.To be indoor design conditions that storing measurement device 77 is chronologically collected be transformed to scalar with the time series data of storage to the histogram of figure central authorities, is expressed as detected value 79 (probe value).The controllable low crest value of process in this detected value 79 and with the impossible non-controlling value 80 of the process control shown in the high crest value (value out of control) can be confirmed on display.
Analyzer 77 is according to the scalar shown in the non-controlling value 80, the constantly corresponding chamber state of processing that analyzing and processing and treatment process 43 are predetermined, with the numerical information of indoor element information characteristic quantification 81 (location) with graphical representation.This characteristic quantification 81 shows which kind of defective the wafer of certain sequence number in the criticizing of processing exists at privileged site.For example according to detected non-controlling value 80, to be set as 0.78 and the value of B parameter is made as-0.63 carries out calculation process as the value of the parameter A of indoor characteristic quantity (or element information), visit is carried out the same journey processing of handling as real-time defect recognition 82 (defective is equal to) in the inside of analyzer 77 or the not shown characteristic quantity database of the digital information of outside this characteristic quantification 81 of record.
Being used for above-mentioned characteristic quantity database with journey processing 82 can handle with the form of scheming correlation table 83, can be used for the correction of the unit state of oxidation furnace and so on treatment process 43.When for example carrying out with journey processing 82 corresponding to the amount of characteristic parameter A=0.78, the B=-0.63 of characteristic quantity (A=0.8/B=-0.6) in the right hurdle of X1 shown in the classification of defects hurdle of defect parameters shown in the correlation table 83 left hurdles and analyzer 77 calculation process, the state of wafer batch is as the situation of batch state 84 of scheming to have shown in upper left defective part, the statistics of exporting with the few wafer of the many wafer of the journal certified products of odd number and even number and certified products.The method for making information and the TEG information that also comprise semiconductor fabrication in the data of this statistics can also comprise the classified information of checking resume and semiconductor device in the data.
Processing according to the specific wafer of the characteristic value on the pairing right hurdle of X1 in the hurdle, a correlation table left side detects certain EES signal difference sometimes between a plurality of chambers in oxidation operation and so on treatment process 43, and can the particular procedure device infer bad mode.
In addition, carried out state according to the classification of defects X2 of defective patterns shown in the correlation table 83 left hurdles with the wafer batch of journey processing 82, can read in the characteristic quantity database as the value 0.7 of the value 0.2 of characteristic quantity stored parameters A and B parameter numerical information with 1 bit representation behind the decimal point, when relatively being approximate situation with the data of having carried out characteristic quantification 81, the state of this wafer batch is such as scheming batch state 85 shown in the left stage casing, can estimate treatment process 43 based on the such statistics of the few wafer of output certified products under stochastic regime.
Equally, carried out state according to the classification of defects X3 of defective patterns shown in the correlation table 83 left hurdles with the wafer batch of journey processing 82, can read in the characteristic quantity database as the value-0.3 of the value 0.4 of characteristic quantity stored parameters A and B parameter numerical information with 1 bit representation behind the decimal point, when with the data of having carried out characteristic quantification 81 relatively when approximate.The state of this wafer batch batch state 86 shown in the section of figure lower-left is such, can based on exported batch preceding partly have the few wafer of certified products and batch the later half many such statisticss of certified products that exist, estimate treatment process 43.In addition, corresponding to the wafer batch of X3 classification of defects number, for example can improve rate of finished products by the reparing process of eliminating defective that rises slowly of temperature in the stove that makes treatment process 43.
Analyzer 77 is as implied above, detect the chamber state in the treatment process 43 in real time, carry out characteristic quantification 81, with journey processing 82, just can by with journey processing 82 find to have a plurality of substandard products batch stage, the automatically specific treatment process 43 that becomes the substandard products reason.Thereby also can constitute when discovery has criticizing of a plurality of substandard products, stop immediately producing the treatment process 43 that substandard products are criticized, notifying operation personnel keep in repair.In addition, also can to the crystal that has carried out characteristic quantification 81 in the new figure criticize branch be equipped with among the figure " new mark " shown in the bottom right with its characteristic quantity with show that the information of bright wafer batch state is recorded in the aforesaid characteristic quantity database.
(the 7th form of implementation)
Fig. 7 is the flow chart of model of manufacture method of the semiconductor device of explanation the present invention the 7th form of implementation, wherein criticize and drop into that operation 42, first treatment process 43, first check that operation the 44, the 3rd is checked operation 44c, second treatment process 46, batch finished operation 47, D/S treatment process 48, quality are inferred portion 61, infer quality management sector 66, detected value 79, non-controlling value 80, characteristic quantification 81, identical with in journey processing 82, correlation table 83 and " new mark " and the foregoing description, omit its duplicate explanation.
The manufacturing system of semiconductor device, will be with known carrying out the relevant wafer 17b of the information of characteristic quantification 81 with defective part 84b and have the pairing defect parameters X1 of wafer 17c of defective part 84a, classification number and a plurality of characteristic quantity of X2, X3 is stored in the characteristic quantity database, though by the wafer that has carried out characteristic quantification 81 being done to find that with journey processing 82 substandard products criticize, even if but when having similar defective FRU 84a, carried out criticizing of wafer 17a processing for substandard products figure, also be difficult to criticize withdraw from and handle and maintenance process as the unknown.
In contrast, present embodiment will be inferred device information that portion 61 collects in real time from the processing unit of implementing treatment process 43 to quality corresponding to each wholesale sending, then when lower procedure finishes, crowd wafer 17a that finishes operation 47, specific final defective part 84a have been passed through by D/S treatment process 48 inspection.
The address of 48 storages of D/S treatment process or output lot number, defective part, the manufacturing system of semiconductor device is according to this lot number, the device information of equipment therefor in the treatment process 43 of the processing A of this batch of retrieval process (for example with reference to figure 2 43a), quality is inferred the device information 43a that retrieves in the portion 61 criticize information monitoring, quality is inferred the control that the parameter of portion 61 inside is automatically upgraded as substandard products.Infer quality management sector 66 as mentioned above, inferring portion 51 from quality accepts to infer qualitative data to infer finished semiconductor device product rate, thereby can be from the semiconductor fabrication after the current time, get rid of crystal early from the manufacturing system line of semiconductor device and criticize the inoperative chemical reagent etc. of avoiding waste effectively with wafer 17a and so on defective patterns.
In addition, the manufacturing system of semiconductor device obviously can be divided the characteristic quantity information of newfound substandard products figure and is equipped with new mark as the classification of defects X4 of defect parameters in the correlation table 83, automatically login in the characteristic quantity database.
(the 8th embodiment)
The system for manufacturing semiconductor device of eighth embodiment of the invention has as shown in Figure 8: the processing unit 14a that handles wafer 17a; Handle the processing unit 14b of wafer 17b; Handle the processing unit 14c of wafer 17c; Handle the processing unit 14d of wafer 17d; Be located at the self-diagnosis system 5a-5d among each processing unit 14a-14d respectively; Be located at the database 13a-13d among each self-diagnosis system 5a-5d respectively; Check the testing fixture 19a-19d of wafer 17a-17d respectively; 5a-5d is connected with self-diagnosis system, receives each the computer of inferring mass value 11 of the handled wafer 17a-17d of each processing unit 14a-14d.
The self-diagnosis system 5a-5d that the 8th embodiment uses can be made of the hardware resource identical with the used self-diagnosable system 11a of first embodiment and the software of collaborative work with it.Like this, as the device information of the self-diagnosis system 5a-5d receiving and processing device 14a-14d of self-diagnosable system, carry out self diagnosis.
When the quality management of computer 11 had many wafers in selected batch, testing fixture 19a-19d only checked the representative wafer, make check result with lot number, represent wafer number corresponding, receive by computer 11.Inferring out in the unchecked wafer with represent wafer has equal in quality, and computer 11 is stored corresponding to lot number and wafer number, and each batch and each wafer are carried out quality management.
Like this, computer 11 makes the mass value of inferring of all wafers of all batches store with lot number and wafer number are corresponding, each is checked that the value of checking on the quality of the wafer that the part of output is constantly criticized is corresponding with lot number and wafer number and stores.
APC, the MES data management manufacturing system integral body of storage in computer 11 application data bases 13 are carried out different processing to these a plurality of processing unit 14a-14d.For example control, make processing unit 14a carry out oxide-film and form processing for wafer 17a, processing unit 14b carries out etch processes for wafer 17b, and processing unit 14c carries out diffusion of impurities for wafer 17c and handles, and processing unit 14d carries out reflux (or fusion) for wafer 17d and handles.
But illustrated processing unit 14a-14d has oversimplified illustration 4, but understands easily from current LSI manufacturing process, and the manufacturing system of semiconductor device generally can be provided with similar processing unit many, is provided with the processing unit more than 10 on the whole.
Each processing unit 14a-14d is by the device information of various its internal states of transducer output expression.For example in the situation of plasma processing apparatus, with room pressure, temperature, High frequency power value, with the capacitance of high-frequency resistance coupling, import indoor gas flow etc. each handle in the device information of regulation export to the self-diagnosis system 5a-5d that is provided with respectively.
Self-diagnosis system 5a is connected with save set database of information 13a, from handle the processing unit 14a receiving system information of wafer 17a by EES, calculate based on device information and to infer mass value (thickness that on wafer 17a, forms when for example processing unit 14a is oxidation furnace) corresponding to wafer 17a, this is inferred mass value send to computer 11, in database 13a, store the infer mass value corresponding with wafer 17a.
So-called " inferring mass value " is for example when processing unit 14a is oxidation furnace, be meant and form from the device information X substitution thickness that processing unit 14a receives the formula (for example Ye=aX+b) of model, go up the thickness presumed value that forms in the wafer 17a that infers film thickness value Ye that calculates.Device information X can be a heter temperature, also can be gas flow, and each device information is stored in the formula that thickness forms model among the database 13a, and self-diagnosis system 5a imports calculation mass value in the formula that thickness forms model with device information.
Self-diagnosis system 5b is connected with save set database of information 13b, from handle the processing unit 14b receiving system information of wafer 17b by EES, calculate based on device information and to infer mass value (etch-rate when for example processing unit 14b is device for dry etching) corresponding to wafer 17b, this is inferred mass value send to computer 11, simultaneously correspondingly will infer mass value and be stored among the database 13b with wafer 17b.
Self-diagnosis system 5c is connected with save set database of information 13c, from handle the processing unit 14c receiving system information of wafer 17c by EES, calculate based on device information and to infer mass value (the diffusion of impurities degree of depth when for example processing unit 14c is diffusion furnace) corresponding to wafer 17c, this is inferred mass value send to computer 11, simultaneously correspondingly will infer mass value and be stored among the database 13c with wafer 17c.
Self-diagnosis system 5d is connected with save set database of information 13d, from handle the processing unit 14d receiving system information of wafer 17d by EES, calculate based on device information and to infer mass value (molten condition of dielectric film when for example processing unit 14d is annealing furnace) corresponding to wafer 17d, this is inferred mass value send to computer 11, simultaneously correspondingly will infer mass value and be stored among the database 13d with wafer 17d.
Referring to Fig. 8 and Fig. 9, the etching in the illustration wafer manufacturing process illustrates the operation of the manufacturing system of semiconductor device below.In manufacturing system, extract the value of the checking on the quality Yq that enters desired range of values, will infer mass value Ye and compare with the value of checking on the quality Yq, the model of inferring mass value Ye is upgraded along following flow process.
Here " desired value " is meant the value of the checking on the quality Yq in the past of storage in the storage device 12 and sequential parallel, and the tendency of calculating inspection matter quality Yq is in current value of checking on the quality for suitable expectation.
(reexamining flow process)
Below be that example describes with the processing unit 14b among processing unit 14a, 14b, 14c, the 14d as device for dry etching.
(a) computer 11 in the check result stage that the wafer 17b that processing unit 14b is handled in testing fixture 19b reception makes regular check on, is transformed into abnormal examination S91 in initial step 90 (following step is reduced to " S ").Computer 11 transfers the check result change to the value of checking on the quality Yq, and the desired value of storage in audit by comparison mass value Yq and the storage device 12 judges whether the value of checking on the quality Yq is unusual.Under the abnormality juding situation, transfer to and measure S92 again, will measure instruction again and send to testing fixture 19b.
(b) testing fixture 19b measures wafer 17b again in measuring S92 again, and the check result of measuring is again sent to computer 11 again.
(c) computer 11 is transformed to the value of checking on the quality Yq in measuring the check result that S92 will measure again again, when establishing with desired value relatively for abnormality juding, is transformed into letters and sends S93, and E-mail is sent to the director, and notice wafer 17b takes place unusual.
(d) computer 11 is transformed into resume and upgrades S94, and the exception record of wafer 17b is in storage device 12, corresponding with the processing unit 14b of the sequence number of wafer, wafer batch, processing wafer, more new database.
(measured value is flow process relatively)
(a) computer 11 will be in the time will measuring the value of the checking on the quality Yq that measures again that receives among the S92 and desired value again and relatively be made as suitable judgement, the resume of measuring again are associated with storing in storage device 12 with testing fixture 19b and wafer 17b, are transformed into given mass comparison S96 through node 95 simultaneously.
(b) computer 11 also is transformed into and infers mass ratio than S96 when abnormal examination S91 judges that the value of checking on the quality Yq is suitable.
(c) computer 11 in infer mass ratio than S96 relatively self-diagnosis system 5b calculate whether infer mass value Ye consistent with determinating reference with the value of checking on the quality Yq judgement, determinating reference is to be judged to be consistently when inferring reliable interval that mass value Ye is present in the regression straight line of inferring the mass value Ye and the value of checking on the quality, then is judged to be inconsistent in the time of outside the reliable interval.
(d) computer 11 is transformed into key element determination processing S97 in inferring mass ratio when having done inconsistent judgement than S96, proofreaies and correct the self diagnosis parameter.The self diagnosis parameter for example can be represented by following formula (1):
Ye1＝a×1+b (1)
(e) resume of computer 11 value of checking on the quality Yq of storage in key element judgement S97 reference storage device 12, the increase and decrease of the value of the checking on the quality Yq1c in past is compared with the mass value Ye1n that infers that is judged to be when inconsistent, infer mass value Ye1n and be similar to the value of checking on the quality Yq for making, judge and whether append the change of coefficient " a ", the change of coefficient " b ", the device information " Xn " different with device information " X1 ".
(f) computer 11 is relatively inferred mass value Ye1n and is judged with in the past the value of checking on the quality Yq1c when the outer change of imagination is arranged, append S98 in device information, append the new device information " Xn " different with device information " X1 ", be transformed into the formula of inferring and upgrade S101, new self diagnosis parameter is recorded among the database 13b via self-diagnosis system 5b.
When (g) computer 11 is judged to be monotone variation when relatively inferring mass value Ye1n and the value of checking on the quality Yq1c in the past, in a change S99 increase and decrease coefficient " a ", make and infer the approximate value of the checking on the quality Yq of mass value Ye1n, be transformed into the formula of inferring and upgrade S101, new self diagnosis parameter is recorded among the database 13b via self-diagnosis system 5b.
When (h) computer 11 is judged to be monotone variation when relatively inferring mass value Ye1n and the value of checking on the quality Yq1c in the past, in b change S100 increase and decrease coefficient " b ", make and infer the approximate value of the checking on the quality Yq of mass value Ye1n, be transformed into the formula of inferring and upgrade S101, new self diagnosis parameter is recorded among the database 13b via self-diagnosis system 5b.
(i) then, computer 11 is transformed into letters and sends S102, and E-mail is sent to the process director, carries out the updating record of new self diagnosis parameter.
(j) computer 11 is transformed into resume and upgrades S94, and the exception record of wafer 17b is in storage device 12, corresponding with the processing unit 14b of wafer number, wafer batch, processing wafer, more new database.
(k) computer 11 is made consistently when judging in inferring mass ratio than S96, transfers to specification and confirms S106, judges whether infer mass value Ye1n enters specification.When judgement enters specification, be transformed into resume and upgrade S94, in storage device 12, with wafer number, wafer batch, the processing unit 14b that handles wafer is corresponding, more new database with the normal recordings of wafer 17b.
Here so-called " specification " is unified specified reference value in each process of making semiconductor device, and be different with above-mentioned desired value.
(sensor abnormality notice flow process)
(a) computer 11, is transformed into transducer and checks S103 during when batch the sending into processing unit 14b and begin to handle of wafer 17b in beginning S90, and inspection is by the work of being located at the various transducers among the processing unit 14b of self-diagnosis system 5b.For example judge the transducer operate as normal whether of the etched device informations of output influence such as transducer, constant pressure transducer of flow sensor, the high frequency electric source of etching gas, infer Mass Calculation S104 being judged to be just often to be transformed into, even and when any sensor abnormality wherein, transfer to sensor abnormality notice S113.
(b) computer 11 receives in sensor abnormality notice S113 from self-diagnosis system 5b and the corresponding sensor abnormality notice of the device information of processing unit 14b, the transducer record information of the processing unit 14b of storage in the reference storage device 12 extracts the relevant record information that is judged to be unusual transducer.For example make such E-mail, wherein record and narrate have from last time occur to unusually current abnormality juding elapsed time (or normal work period) continuously, from processing unit 14b be arranged at the total degree of making back abnormality juding on the line, the normal recommended value of device information tendency during based on processing unit 14b operate as normal etc.
(c) computer 11 is transformed into letters and sends S114, will send to the process director at the E-mail that sensor abnormality notifies S113 to make, the sensor abnormality of notifier processes device 14b, end process.
(inferring the Mass Calculation flow process)
(a) self-diagnosis system 5b is transformed into transducer from beginning S90 and checks S103, when transducer all just often, be transformed into and infer Mass Calculation S104, wafer 17b is being sent into the stage that processing unit 14b begins to handle, for example also begin to obtain the device information " X1 " of the electric capacity that shows the impedance adjustment of carrying out the chamber, application formula (1) calculation mass value Ye1 is transformed into quality management S105.
(b) computer 11 is inferred mass value Ye1 in what quality management S105 accepted wafer 17b that self-diagnosis system 5b calculates, judges that according to lot number or wafer number wafer 17b checks not for testing fixture 19b simultaneously.When being judged to be when checking out, transfer to and infer mass ratio than S96, carry out relatively flow process sequence of measured value.
On the other hand, when wafer 17b does not check for testing fixture 19b, transfer to specification and confirm S106.Just owing to when handling many batches of wafers, checking that termly wafer 17b has unchecked wafer by processing unit 14b, even what also cross owing to inspected batch is not represent the wafer 17b of wafer not for the resume of testing fixture 19b inspection also are stored in the storage device 12, thereby the resume of computer 11 reference wafer are transferred to specification affirmation S106.
(c) computer 11 confirms that in specification S106 judges whether infer mass value Ye1 enters specification.When judgement enters specification, be transformed into resume and upgrade S94, with the normal recordings of wafer 17b in storage device 12, corresponding to wafer number, wafer lot number, handled the processing unit 14b of wafer, new database more, end process.
On the other hand, when computer 11 confirms that in specification S106 judges that inferring mass value Ye1 does not enter specification, shift to appending and check decision S107.
(d) computer 11 checks that in appending decision S107 appends when checking according to process director's instruction, transfer to and append quality management S108, check wafer 17b in testing fixture 19b, with conversion the value of the checking on the quality Ya of check result be stored in storage device 12 corresponding to wafer 17b.
(e) computer 11 to inferring mass ratio than S96, compares treatment conversion the value of the checking on the quality Yq that preserves in the storage device 12 and infers mass value Ye1 in appending quality management S108, carries out relatively flow process sequence of measured value.
(f) computer 11 is checked the instruction of decision S107 according to the process person of being responsible in appending, do not carry out appending when checking of wafer 17b, be transformed into upper composition report S109, relatively in the storage device 12 the given mass value Ye1c in the past of storage with confirm in specification to judge among the S106 do not enter specification infer mass value Ye1n, make the E-mail of inconsistent big upper composition (for example going up potential coefficient " a " and next coefficient " b ") in the model of inferring mass value Ye1 of recording and narrating calculating formula (1) expression.
In addition, computer 11 will confirm that S106 is judged to be the out-of-spec mass value Ye1n that infers and is stored in the database of storage device 12 in specification in upper composition report S109, the resume of having stored in the reference database of inferring mass value Ye1c, investigation has or not with in the past same unusual, makes the E-mail of recording and narrating investigation result.
(g) computer 11 is transformed into letters and sends S110, will send to the process person of being responsible at the E-mail that upper composition report S109 makes, and the mass value Ye1 that infers of notice wafer 17b takes place unusual.
Like this, in the manufacturing system of the semiconductor device of the 8th form of implementation, computer 11 compares the mass value Ye1 that infers that self-diagnosis system 5b calculates with the value of the checking on the quality Yq that obtains from testing fixture 19b termly, poor according to this Ye1 and Yq, the coefficient of the formula of inferring of mass value Ye1 is inferred in correction, or change and infer formula, thereby can improve the self diagnosis precision of processing unit 14b.
The profile that is used for the wafer of inferring Mass Calculation S104 of Fig. 9 is illustrated in Figure 10.Processing unit 14b sends into wafer indoor, and the wafer of sending into has the tetraethoxysilane TEOS118 of the interlayer dielectric that forms in silicon substrate 119 tops and goes up the membrane structure of the antireflection film 117 that forms with TEOS118.In addition, the interlayer dielectric among the present invention is not limited to TEOS, and can be object with TMPO, the TEM as other semiconductor film materials also.
The back side of antireflection film 117 has has finished zone that resist 116 that exposure imaging handles forms and has handled the zone of removing resist 116 and exposing antireflection film 117 through exposure imaging.
Processing unit 14b carries out the etching of antireflection film 117 with respect to the zone that antireflection film 117 exposes, up to appearing TEOS118.
Self-diagnosis system 5b receives the etching gas (CF of etching antireflection film 117 from processing unit 14b
4, HBr, O
2Deng) multiple device information Xi such as the relevant parameter of high frequency electric sources such as flow, the capable ripple of high frequency electric source, the reflected wave of high frequency electric source, pressure (i represents the number of the information of collecting).
Self-diagnosable system 5b for example infers mass value Ye1 for what calculate etch-rate, and the capacitance of adjusting the impedance of chamber is obtained as device information.Obtain the selection item that mass value Ye2 then is a manufacture process of inferring of calculating etch-rate as the etchant gas flow of device information X2.
Figure 11 shows that the bright Fig. 9's of being used for infers the curve chart of mass ratio than the determinating reference of S96.What the transverse axis of this curve chart was represented etch-rate that self-diagnosis system 5b calculates infers mass value Ye1.The longitudinal axis of curve chart then represents to check based on testing fixture 19b the measured value of wafer 17b, by the etch-rate value of checking on the quality Yq1 of computer 11 calculating.
When the formula that is used for asking the model of inferring mass value Ye1 (1) when suitable, occur and wafer, have in etch-rate 2.3-4.3 (nm/ second) scope and regression straight line 126 overlaids or be positioned near it quality with respect to the intersection point of inferring mass value Ye1 of the value of the checking on the quality Yq1 of each wafer number.
But when the indoor pressure in being located at processing unit 14b, capacitance, etchant gas flow change, the wafer quality between each batch can change sometimes.For example have the wafer of circle 120,121,122,123 intersection points that surrounded among the figure, with respect to the value of checking on the quality Yq1, produce difference with inferring mass value Ye1 respectively, needed to change the formula (1) of asking the model of inferring mass value Ye1, the reliability of inferring mass value Ye1 with raising for this reason.
Have the wafer by intersection point that circle 120 encloses, with respect to 2.8 of the value of checking on the quality Yq1, inferring mass value is 2.3, infers mass value Ye1, the parameter of the formula (1) of change model for what increase the regression straight line upside.As make coefficient a increase be similar to the value of checking on the quality Yq1.
Have the wafer by intersection point that circle 121 encloses, with respect to 3.1 of the value of checking on the quality Yq1, inferring mass value Ye1 is 2.6, for increase be positioned at regression straight line 126 upsides infer mass value Ye1, the parameter of the formula (1) of change model.For example make coefficient a increase the approximate value of checking on the quality Yq1.
Have the wafer by intersection point that circle 122 encloses, with respect to 2.3 of the value of checking on the quality Yq1, inferring mass value Ye1 is 2.8, infers mass value Ye1, the parameter of the formula (1) of change model for what reduce to be positioned at regression straight line 126 downsides.For example make coefficient a reduce the approximate value of checking on the quality Yq1.
Have the wafer by intersection point that circle 123 encloses, with respect to 2.5 of the value of checking on the quality Yq1, inferring mass value Ye1 is 3.0, infers mass value Ye1, the parameter of the formula (1) of change model for what reduce to be positioned at regression straight line 126 downsides.For example make coefficient a reduce the approximate value of checking on the quality Yq1.
As shown in figure 12, in the time-sequence curve chart of the value of checking on the quality Yq1, be shown the date that transverse axis represents to handle wafer, the longitudinal axis is represented the etch-rate value of the checking on the quality Yq1 of (nm/ second).
Figure 11 has a wafer batch of handling on February 4 that the wafer by intersection point that circle 120 encloses is equivalent to be enclosed with circle 130 among Figure 12, because the value of checking on the quality Yq1 increased monotonously from February 1, so computer 11 increases the coefficient a of formula (1), that upgrades etch-rate infers mass value Ye1n.
Figure 11 has a wafer batch of handling on February 5 that the wafer by intersection point that circle 121 encloses is equivalent to be enclosed with circle 131 among Figure 12, because the value of checking on the quality Yq1 is from dull increasing on February 4, so computer 11 increases the coefficient a of formula (1), that upgrades etch-rate infers mass value Ye1n.
Figure 11 has a wafer batch of handling on February 7 that the wafer by intersection point that circle 122 encloses is equivalent to be enclosed with circle 132 among Figure 12, owing to compare rapid minimizing with the value of the checking on the quality Yq1 on February 5, so computer 11 reduces the coefficient b of formula (1), that upgrades etch-rate infers mass value Ye1n.
Figure 11 has a wafer batch of handling on February 13 that the wafer by intersection point that circle 123 encloses is equivalent to be enclosed with circle 133 among Figure 12, because the value of checking on the quality Yq1 increased monotonously from February 7, so carrying out the change of the coefficient a of formula (1), computer 11 judges, for making the value of checking on the quality Yq1n be positioned at regression straight line 126 downsides the coefficient a of formula (1) is reduced, that upgrades etch-rate infers mass value Ye1n.
The wafer that has the intersection point that is surrounded by circle 134 among Figure 12 is equivalent to the wafer batch of processing on February 22, compares with the value of the checking on the quality Yq1c on February 13, increases sharp.The present situation of this batch crystal varies to outside the setting of formula (1), computer 11 is transformed into the formula of inferring of Fig. 9 and upgrades S101, new self diagnosis parameter is recorded among the database 13b via self-diagnosis system 5b, be transformed into letters and send S102 (with reference to figure 9), infer the change of formula with letters notification procedure director inferring quality.
In the 8th form of implementation, the self diagnosis parameter of computer 11 change database 13b, the pairing mass value Ye1e that infers of wafer number of storage in the storage device 12 is stored, so that the mass value Ye1n that infers of a plurality of wafers of multi-batch processing is similar to the value of checking on the quality Yq1n corresponding to wafer batch with the value of checking on the quality Yq1c.
So just can prevent from only to monitor etch-rate infer mass value Ye1 the time, the unusual flase drop of the processing unit 14b that can cause is surveyed and lets slip unusual.Promptly by preventing that the flase drop survey from rise the operation ratio of processing unit 14b, by preventing to let slip the generation that prevents that unusually low rate of finished products from criticizing.
Among processing unit 14a, 14b, 14c, 14d, be that example describes below with processing unit 14a with oxidation furnace.
Self-diagnosis system 5a will supply with the supply gas (O of processing unit 14a
2, N
2, H
2Deng) flow, the electrical power of heater, the air pressure around the oxidation furnace etc. receives as device information.The device information that receives is stored among the database 13a, uses following formula (2) to calculate the given mass value Ye2 of the thickness of oxidation film of wafer 17a simultaneously:
Ye2＝CX2+dX3+eX4+f (2)
Device information X2 represents the electrical power of heater, and device information X3 represents O
2Flow, device information X4 are represented the air pressure around the oxidation furnace.In addition, computer 11 is distributed to coefficient c, d, e and f with the self diagnosis parameter.
Though the thickness of oxidation film of wafer 17a can be subjected to the influence of all parameters, as shown in figure 13, long-term continuous out of service before (among the figure from April 13 to April 28 days) thickness of oxidation film (nm) infer the scope that mass value Ye2c concentrates on " 98-99 ".And quit work from morning night to May 6 on April 28 as processing unit 14a, long-term back out of service (among the figure after May 6), thickness of oxidation film (nm) infer the scope that mass value Ye2n then moves to " 96-97 ", the mass value Ye2 that infers of thickness of oxidation film has big change.
Self-diagnosis system 5a receives the device information of heater electrical power (w) as shown in figure 14 from processing unit 14a.The electrical power 140 of long-term heater before out of service continuously from start up process device 14a after 500 seconds below 900W, after 600 seconds below 200W, after 650 seconds below 100W.
In contrast, in long-term heater electrical power 141 after out of service continuously, be 600W after 500 seconds in start up process device 14a, after 600 seconds be zero W and after 650 seconds from the 200W that is raised to above freezing.750 seconds be below the 150-300W later on, with long-term heater electrical power 140 before out of service continuously relatively, different is, after 650 seconds, the electrical power of heater descends, and after 700 seconds, the rising of the electrical power of heater.
The temperature of processing unit 14a monitoring oxidation furnace rises.Long-term preceding temperature 142 out of service continuously rose to 800 ℃ in start up process device 14a after 560 seconds, rise to 840 ℃ after 700 seconds, 800 seconds later scopes that then move on to 840-850 ℃.Temperature 143 after long-time running stops after rose to 800 ℃ in 570 seconds, rises to 830 ℃ after 700 seconds at start up process device 14a, moves to 840-850 ℃ of scope after 800 seconds, compares before out of service continuously with long-term, and temperature rising curve descends.
Self-diagnosis system 5a can think that owing to the device information that has received the peak power of different device information X2 in long-term continuous front and back out of service from processing unit 14a temperature rising curve can change in long-term continuous front and back out of service.
So, computer 11 will be inferred the check on the quality value Yq2 of quality comparison process by the back long-term continuous out of service wafer 17a thickness of testing fixture 19a actual measurement based on implementation, with receive from the device 5a that autotomys long-term out of service continuously after wafer 17a thickness of oxidation film infer mass value Ye2 relatively, carry out inconsistent judgement.
Like this, by comparing termly based on the value of the checking on the quality Yq2 of the thickness of oxidation film measured value of wafer 17a and the mass value Ye2 that infers of the thickness of oxidation film that calculates by self-diagnosis system 5, just can grasp the state variation of the heater of long-term continuous front and back out of service (the non-regular operation of several in a year), and can prevent the thickness of oxidation film of wafer 17a to infer the unusual of mass value Ye2.
Effect that the invention process form is write down and effect only are to have enumerated the best use of and the effect that produce according to the present invention, and effect of the present invention and effect are not limited to be recorded and narrated in the invention process form.
The following describes the manufacturing system technology item relevant with semiconductor making method with the semiconductor device of the invention process form.
(a) manufacturing system of the semiconductor device of form of implementation of the present invention, it has: the processing unit of carrying out the processed that adopts semiconductor substrate; The self-diagnosable system of inferring mass value from processing unit receiving system information calculations processed; Check processed result's testing fixture; The audit by comparison result with infer the quality value, when to inferring mass value when making effective judgements, keep the parameter of self-diagnosable system, when inferring the CALCULATION OF PARAMETERS machine that then changes self-diagnosable system when quality is invalid judgement.
(b) computer of using in the form of implementation of the present invention is to inferring quality when doing invalid judgement, the change parameter formula.
(c) computer of using in the form of implementation of the present invention in the time of in check result is present in the reliable interval of inferring the regression straight line that mass value and measured value form, is effectively judged.
(d) computer of using in the form of implementation of the present invention calculates the quality tendency based on the measured value of storing in the storage device in the past, and audit by comparison result and quality tendency are judged.
(e) computer of using in the form of implementation of the present invention, the check result of based semiconductor substrate oxidation processes is judged.
(f) computer of using in the form of implementation of the present invention, the check result of based semiconductor substrate etching treatment procedure is judged.
(g) method for making semiconductor of form of implementation of the present invention has: the processing unit of the processed of semiconductor substrate is carried out in diagnosis, generates the automatic reparation request step of repairing request signal automatically in the stage of satisfying predetermined condition; Check the quality examination step of the quality of treated semiconductor substrate; The quality management information that obtains based on the quality examination step, judge that the generation of repairing request signal automatically is suitable constantly, when being effectively judgement, this result of determination keeps the parameter of self-diagnosable system, when this result of determination then changes the parameter of self-diagnosable system when being invalid, change the automatic inspection frequency correction step of the frequency of quality examination operation simultaneously.
(h) in the method, semi-conductor device manufacturing method of the invention process form, when result of determination was invalid judgement, computer appended the parameter of self-diagnosable system.
(i) method, semi-conductor device manufacturing method of the invention process form has: in the processing device condition that monitors the processed of carrying out semiconductor substrate, according to the device information integration of processing procedure model, carry out the semiconductor fabrication simulation to infer the real-time Simulation step of semiconductor substrate quality under this processing to this processing unit.
(j) method, semi-conductor device manufacturing method of the invention process form has: in the device information of the processing unit that monitors the processed of carrying out semiconductor substrate, export the quality of inferring quality information of this semiconductor substrate and infer step; Export the quality of information of checking on the quality of treated semiconductor substrate and check step; Relatively infer quality information and check on the quality information, judge whether the moment of quality examination step is suitable, when result of determination is invalid judgement, detect the unusual abnormality detection step of processing unit.
(k) method, semi-conductor device manufacturing method of the invention process form has: at the processing device condition that monitors the processed of carrying out semiconductor substrate, the quality of inferring quality information of output semiconductor substrate is inferred step; Export the quality of information of checking on the quality of treated semiconductor substrate and check step; What quality information and the quality of information relevant treatment of checking on the quality were inferred in implementation infers the quality management step; Carry out the semiconductor fabrication simulation, the yield prediction step of prediction semiconductor device yield according to inferring quality information; Check that the processing through processing unit is the rate of finished products inspection step of the finished semiconductor device product rate of wafer operation end at least; Relatively the rate of finished products of yield prediction step and rate of finished products are checked the rate of finished products relevant treatment step of the rate of finished products of step.
(l) form of implementation of the present invention is the manufacture method of semiconductor device, and it has: the processed step of carrying out the processed of semiconductor substrate; Output is through the quality examination step of the quality information of the semiconductor substrate of processed step; Reception is carried out the semiconductor fabrication simulation from the quality information of quality examination step, the yield prediction step of prediction finished semiconductor device product rate; Checking the processing through processing unit, is that the rate of finished products of having finished the finished semiconductor device product rate after the wafer operation is checked step at least; And the rate of finished products of yield prediction step and rate of finished products checked that the rate of finished products of step compares, improve the rate of finished products relevant treatment step of prediction accuracy of the rate of finished products of yield prediction step.
(m) form of implementation of the present invention is the manufacture method of semiconductor device, and it has: the treatment step of carrying out the processed of semiconductor substrate; Monitor the device information of processing unit, the quality of inferring quality information of output semiconductor substrate is inferred step; Output is through the quality examination step of the quality information of the semiconductor substrate of processed step; Infer quality information and quality information the quality relevant treatment infer the quality management step; Carry out the semiconductor fabrication simulation according to quality information, the first yield prediction step of prediction finished semiconductor device product rate from the quality examination step; According to from the quality information of inferring of inferring the quality management step, carry out the different simulation of simulation with the semiconductor fabrication of the first yield prediction step, through the second yield prediction step, the processed step of prediction finished semiconductor device product rate, check it is that the rate of finished products of having finished the finished semiconductor device product rate of wafer operation is checked step at least; Relatively the rate of finished products of the first yield prediction step and rate of finished products are checked the first rate of finished products relevant treatment step of the rate of finished products of step; Relatively the rate of finished products of the second yield prediction step and rate of finished products are checked the second rate of finished products relevant treatment step of the rate of finished products of step; Based on the correlated results of first and second rate of finished products relevant treatment step prediction rate of finished products step with the yield prediction reliability that improves first and second yield prediction step.
(n) form of implementation of the present invention is the manufacture method of semiconductor device, it has: the status detection value that detects the processing unit of the processed of representing the implementation semiconductor substrate, non-controlling value extraction from this detected value of this processing is obtained the characteristic quantification step of characteristic quantity, to the characteristic quantity that extracts by the characteristic quantification step with intrinsic defect image is carried out the characteristic quantity that category classification quantizes in the correlation table of pre-stored compare, judging that mutual characteristic quantity is the stage with the value of journeyization, interrupt semiconductor fabrication, semiconductor substrate is criticized the quality testing step of notice as substandard products.
Form of implementation of the present invention is the manufacture method of semiconductor device, when detect have with correlation table in characteristic quantity can not criticize with the substandard products of the new feature amount of journeyization the time, the processing record information of criticizing according to these substandard products, extract these substandard products of expression and criticize the device information of processing device condition when generating, infer the automatic step of updating of quality what this device information fed back to automatically that quality infers portion.
Claims (22)
1. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the processed of semiconductor substrate; Control the processing control apparatus of above-mentioned processing unit; Carry out the processed of above-mentioned semiconductor substrate, monitor the state of above-mentioned processing unit, the internal information of the above-mentioned processing unit of integration is carried out the simulation of above-mentioned processed, infers the real time simulator of the carrying out of above-mentioned semiconductor substrate processing according to above-mentioned processed.
2. system for manufacturing semiconductor device as claimed in claim 1 is characterized in that, above-mentioned real time simulator begins the above-mentioned internal information of integration from the processing stop signal due in that above-mentioned processing control apparatus sends, and carries out the simulation of above-mentioned processed.
3. system for manufacturing semiconductor device as claimed in claim 1 is characterized in that above-mentioned real time simulator is carried out the simulation of the oxidation processes of above-mentioned semiconductor substrate, by above-mentioned processed, infers the carrying out of the processing of above-mentioned semiconductor substrate.
4. system for manufacturing semiconductor device as claimed in claim 1 is characterized in that above-mentioned real time simulator is carried out the simulation of the etch processes of above-mentioned semiconductor substrate, infers the carrying out that above-mentioned semiconductor substrate is handled according to above-mentioned processed.
5. system for manufacturing semiconductor device as claimed in claim 1 is characterized in that above-mentioned real time simulator monitors the end points of above-mentioned semiconductor substrate, infers the carrying out that above-mentioned semiconductor substrate is handled.
6. system for manufacturing semiconductor device as claimed in claim 1, it is characterized in that, above-mentioned processing control apparatus receives the sign on of above-mentioned processed from the computer that is different from above-mentioned real time simulator, above-mentioned processing unit is sent handle commencing signal and handle stop signal.
7. system for manufacturing semiconductor device as claimed in claim 6 is characterized in that, above-mentioned real time simulator sends to the aforementioned calculation machine with the result that infers of the carrying out that above-mentioned semiconductor substrate is handled.
8. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the processed that adopts semiconductor substrate; The self-diagnosable system of inferring mass value from the above-mentioned processed of above-mentioned processing unit receiving system information calculations; Check above-mentioned processed result's testing fixture; More above-mentioned check result and the above-mentioned mass value of inferring are when when making effective judgement, keeping the parameter of above-mentioned self-diagnosable system to the above-mentioned mass value of inferring, when the above-mentioned mass value of inferring is the invalid CALCULATION OF PARAMETERS machine that then changes above-mentioned self-diagnosable system when judging.
9. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the processed that adopts semiconductor substrate; Carry out the self-diagnosable system of above-mentioned processing unit from above-mentioned processing unit receiving system information from management; Check above-mentioned processed result's testing fixture; Judge according to above-mentioned check result and to repair above-mentioned processing unit whether automatically, keep the parameter of above-mentioned self-diagnosable system when result of determination when effectively judging and work as the CALCULATION OF PARAMETERS machine that then changes above-mentioned self-diagnosable system when result of determination is invalid judgement.
10. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the processed of semiconductor substrate; Self-diagnosis system based on self diagnosis parameter diagnosis processing unit; Check the testing fixture of processed; Be connected with self-diagnosis system and testing fixture, when the check result of semiconductor substrate is effective, keep the self diagnosis parameter and when check result is invalid change self diagnosis CALCULATION OF PARAMETERS machine.
11. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the semiconductor substrate processed; Obtain the device information of processing unit, infer the quality of semiconductor substrate quality and infer portion; To the quality examination device that carries out quality examination through the semiconductor substrate of processed; Relatively quality is inferred the comparator of the quality management information of inferring qualitative data and the actual measurement of quality examination device of portion.
12. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the semiconductor substrate processed; Obtain the device information of processing unit, the quality of inferring quality information of output semiconductor substrate is inferred portion; Export the quality examination device of the quality information of treated semiconductor substrate; Implementation infer quality information and quality information the quality relevant treatment infer quality management sector; Simulate the yield prediction device of prediction finished semiconductor device product rate according to the quality information of inferring of inferring quality management sector's output; Check that processing through processing unit is the rate of finished products testing fixture of having finished the finished semiconductor device product rate after the wafer operation at least; The rate of finished products relevant treatment that implementation compares the rate of finished products of the rate of finished products of yield prediction device and rate of finished products testing fixture.
13. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the semiconductor substrate processed; Export the quality examination device of the quality information of treated semiconductor substrate; According to simulate the yield prediction device of prediction finished semiconductor device product rate from the quality information of quality examination device; Check that processing through processing unit is the rate of finished products testing fixture of having finished the finished semiconductor device product rate after the wafer operation at least; The rate of finished products relevant treatment that implementation compares the rate of finished products of the rate of finished products of yield prediction device and rate of finished products testing fixture.
14. a system for manufacturing semiconductor device is characterized in that it has: the processing unit of carrying out the semiconductor substrate processed; Monitor the device information of processing unit, the quality of inferring quality information of output semiconductor substrate is inferred portion; Export the quality examination device of the quality information of treated semiconductor substrate; Infer quality information and quality information the quality relevant treatment infer quality management sector; According to simulate the first yield prediction device of prediction finished semiconductor device product rate from the quality information of quality examination device; According to from the quality information of inferring of inferring quality management sector, carry out the simulation of the simulation that is different from the first yield prediction device; Semiconductor fabrication has been passed through in inspection, is the rate of finished products testing fixture of having finished the finished semiconductor device product rate of wafer operation at least; Compare the second rate of finished products relevant treatment of the rate of finished products of the rate of finished products of the first rate of finished products relevant treatment of rate of finished products of the rate of finished products of the first yield prediction device and rate of finished products testing fixture and the comparison second yield prediction device and rate of finished products testing fixture.
15. the manufacturing system of a semiconductor device, it has: the processing unit of carrying out the processed of semiconductor substrate; Detect the device information of processing unit, will from detected value, extract the characteristic quantity that the non-controlling value of handling is carried out characteristic quantification; With compare by intrinsic defect image being carried out the characteristic quantity that category classification quantizes in the correlation table of pre-stored, the characteristic quantity of judging this characteristic quantification be can with characteristic quantity in this correlation table stage with the value of journeyization, Interrupt Process is criticized semiconductor substrate the quality testing portion of notice as substandard products.
16. system for manufacturing semiconductor device, it is characterized in that, it has: infer the quality automatic renewing device, this updating device has and quantizes characteristic quantity in the correlation table of pre-stored can not criticize with the substandard products of the new feature amount of journeyization the time by the inherent shortcoming image being carried out category classification when detecting, the processing record information of the semiconductor substrate of criticizing according to these substandard products, extract these semiconductor substrate substandard products of expression and criticize the device information of processing device condition when generating, this device information is fed back to quality automatically infer portion.
17. method, semi-conductor device manufacturing method, it is characterized in that the method carries out the processed of semiconductor substrate by processing unit, carry out the processed of above-mentioned semiconductor substrate, the state that monitors above-mentioned processing unit, the above-mentioned processing unit of integration internal information, carry out above-mentioned processed simulation, infer the carrying out that above-mentioned semiconductor substrate is handled by above-mentioned processed.
18. method, semi-conductor device manufacturing method as claimed in claim 17 is characterized in that, above-mentioned processed is the oxidation processes of above-mentioned semiconductor substrate.
19. method, semi-conductor device manufacturing method as claimed in claim 17 is characterized in that, above-mentioned processed is that the etching and processing of above-mentioned semiconductor substrate is handled.
20. method, semi-conductor device manufacturing method as claimed in claim 17 is characterized in that, the simulation of above-mentioned processed is to begin to carry out above-mentioned internal information integration when the processing stop signal that above-mentioned processing control apparatus sends arrives.
21. method, semi-conductor device manufacturing method as claimed in claim 17 is characterized in that, above-mentioned simulation monitor the end points of above-mentioned semiconductor substrate appearance, the above-mentioned processing unit of integration internal information and infer the carrying out that above-mentioned semiconductor substrate is handled.
22. method, semi-conductor device manufacturing method as claimed in claim 17 is characterized in that, it is etch processes that above-mentioned semiconductor substrate is handled.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2003355684 | 2003-09-08 | ||
| JP355684/2003 | 2003-09-08 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102634018A Division CN101937836B (en) | 2003-09-08 | 2004-09-07 | Manufacturing system and method of semiconductor device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1595607A true CN1595607A (en) | 2005-03-16 |
Family
ID=34674780
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102634018A Expired - Fee Related CN101937836B (en) | 2003-09-08 | 2004-09-07 | Manufacturing system and method of semiconductor device |
| CNA2004100746155A Pending CN1595607A (en) | 2003-09-08 | 2004-09-07 | Manufacturing system and method for semiconductor device |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102634018A Expired - Fee Related CN101937836B (en) | 2003-09-08 | 2004-09-07 | Manufacturing system and method of semiconductor device |
Country Status (3)
| Country | Link |
|---|---|
| KR (1) | KR100733590B1 (en) |
| CN (2) | CN101937836B (en) |
| TW (1) | TWI280603B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103336518A (en) * | 2013-07-08 | 2013-10-02 | 上海华力微电子有限公司 | Method of obtaining measurement data from external auxiliary analysis system |
| CN103346105A (en) * | 2013-06-27 | 2013-10-09 | 上海华力微电子有限公司 | Defect casual inspection method capable of dynamically adjusting according to technology wafer number load |
| CN109375593A (en) * | 2012-09-21 | 2019-02-22 | 株式会社东芝 | Manufacturing management system and manufacturing management method |
| US11341441B2 (en) * | 2020-05-11 | 2022-05-24 | Shanghai Huali Integrated Circuit Corporation | Automatic product dispatching device and method in case of site abnormality in queue time section loop |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7289863B2 (en) * | 2005-08-18 | 2007-10-30 | Brooks Automation, Inc. | System and method for electronic diagnostics of a process vacuum environment |
| US9418413B1 (en) | 2009-07-06 | 2016-08-16 | Camtek Ltd. | System and a method for automatic recipe validation and selection |
| TWI497623B (en) * | 2009-07-06 | 2015-08-21 | Camtek Ltd | A system and a method for automatic recipe validation and selection |
| JP6063313B2 (en) | 2013-03-22 | 2017-01-18 | 株式会社東芝 | Electronic device manufacturing support system, manufacturing support method, and manufacturing support program |
| US20140303912A1 (en) * | 2013-04-07 | 2014-10-09 | Kla-Tencor Corporation | System and method for the automatic determination of critical parametric electrical test parameters for inline yield monitoring |
| CN106486391B (en) * | 2015-08-31 | 2020-08-07 | 北大方正集团有限公司 | Semiconductor manufacturing control method and system based on MES system |
| KR102525873B1 (en) * | 2015-10-16 | 2023-04-27 | 삼성전자주식회사 | Semiconductor process simulation device and simulation method thereof |
| US10310490B2 (en) * | 2016-02-01 | 2019-06-04 | Qoniac Gmbh | Method and apparatus of evaluating a semiconductor manufacturing process |
| JP6983966B1 (en) * | 2020-09-16 | 2021-12-17 | 株式会社東芝 | Thickness measuring device |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5076205A (en) * | 1989-01-06 | 1991-12-31 | General Signal Corporation | Modular vapor processor system |
| US5711843A (en) * | 1995-02-21 | 1998-01-27 | Orincon Technologies, Inc. | System for indirectly monitoring and controlling a process with particular application to plasma processes |
| US5910011A (en) * | 1997-05-12 | 1999-06-08 | Applied Materials, Inc. | Method and apparatus for monitoring processes using multiple parameters of a semiconductor wafer processing system |
| KR100216066B1 (en) * | 1997-05-20 | 1999-08-16 | 윤종용 | Control system and control method for ic test process |
| US6763130B1 (en) | 1999-07-21 | 2004-07-13 | Applied Materials, Inc. | Real time defect source identification |
| JP4409744B2 (en) | 2000-10-20 | 2010-02-03 | ライトロン株式会社 | Etching process progress display method and etching process monitoring apparatus |
| JP4213871B2 (en) | 2001-02-01 | 2009-01-21 | 株式会社日立製作所 | Manufacturing method of semiconductor device |
| JP3993396B2 (en) * | 2001-03-30 | 2007-10-17 | 株式会社東芝 | Manufacturing method of semiconductor device |
-
2004
- 2004-09-02 TW TW093126536A patent/TWI280603B/en not_active IP Right Cessation
- 2004-09-07 CN CN2010102634018A patent/CN101937836B/en not_active Expired - Fee Related
- 2004-09-07 CN CNA2004100746155A patent/CN1595607A/en active Pending
- 2004-09-08 KR KR1020040071858A patent/KR100733590B1/en not_active IP Right Cessation
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109375593A (en) * | 2012-09-21 | 2019-02-22 | 株式会社东芝 | Manufacturing management system and manufacturing management method |
| CN103346105A (en) * | 2013-06-27 | 2013-10-09 | 上海华力微电子有限公司 | Defect casual inspection method capable of dynamically adjusting according to technology wafer number load |
| CN103346105B (en) * | 2013-06-27 | 2015-09-30 | 上海华力微电子有限公司 | A kind of can according to the defect sampling observation method of technique wafer number load dynamic conditioning |
| CN103336518A (en) * | 2013-07-08 | 2013-10-02 | 上海华力微电子有限公司 | Method of obtaining measurement data from external auxiliary analysis system |
| CN103336518B (en) * | 2013-07-08 | 2016-03-30 | 上海华力微电子有限公司 | The method of metric data is obtained from external auxiliary analytic system |
| US11341441B2 (en) * | 2020-05-11 | 2022-05-24 | Shanghai Huali Integrated Circuit Corporation | Automatic product dispatching device and method in case of site abnormality in queue time section loop |
Also Published As
| Publication number | Publication date |
|---|---|
| TW200512794A (en) | 2005-04-01 |
| KR100733590B1 (en) | 2007-06-28 |
| KR20050025936A (en) | 2005-03-14 |
| CN101937836A (en) | 2011-01-05 |
| TWI280603B (en) | 2007-05-01 |
| CN101937836B (en) | 2012-08-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1199235C (en) | Method and device for searching flaw aggregation in mfg. semiconductor device, and program thereby | |
| JP4008899B2 (en) | Semiconductor device manufacturing system and semiconductor device manufacturing method | |
| CN1186700C (en) | Adaptive sampling method for improved control in semiconductor manufacturing | |
| CN1595607A (en) | Manufacturing system and method for semiconductor device | |
| CN100451888C (en) | Method and system for monitoring the processing tools in semiconductor production device | |
| JP7291255B2 (en) | PROCESSING APPARATUS, EQUIPMENT MANAGEMENT CONTROLLER, PROGRAM, AND SEMICONDUCTOR DEVICE MANUFACTURING METHOD | |
| CN1655325A (en) | Device for and method of creating a model for determining relationship between process and quality | |
| JP5032118B2 (en) | A system and method using a first principle simulation in a semiconductor manufacturing process. | |
| JP5095999B2 (en) | A system and method using a first principle simulation for analyzing a process performed by a semiconductor processing tool. | |
| JP4955393B2 (en) | A system and method using first principle simulation to control a semiconductor manufacturing process. | |
| CN1666203A (en) | Method and apparatus for automatic sensor installation | |
| CN1656599A (en) | Method and apparatus for monitoring tool performance | |
| CN1682165A (en) | Method and apparatus for the monitoring and control of a semiconductor manufacturing process | |
| JP2007507886A (en) | A system and method using a first principle simulation for controlling a semiconductor manufacturing process. | |
| CN107240564B (en) | Processing device, device management controller, and device management method | |
| CN1879195A (en) | Method for making semiconductor integrated circuit device | |
| CN1653598A (en) | Method of predicting processing device condition or processed result | |
| CN101031851A (en) | Semiconductor processing method using virtual modules | |
| CN1641841A (en) | Plasma processing method and apparatus | |
| WO2006041543A1 (en) | Method and system for dynamically adjusting metrology sampling based upon available metrology capacity | |
| CN1252795C (en) | Manufacturing method of semiconductor device, manufacturing equipment and control method of semiconductor device and control apparatus and semiconductor device | |
| CN1790614A (en) | Method of resetting substrate processing apparatus, storing program and substrate processing apparatus | |
| TWI708197B (en) | Predictive maintenance method for component of production tool and computer program product thereof | |
| CN1637756A (en) | Mass-production transfer support system and semiconductor manufacturing system | |
| JP2006505130A (en) | Method and apparatus for providing first-principles feedforward manufacturing control |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20050316 |