CN1673991A - Defect control method - Google Patents
Defect control method Download PDFInfo
- Publication number
- CN1673991A CN1673991A CN 200410031726 CN200410031726A CN1673991A CN 1673991 A CN1673991 A CN 1673991A CN 200410031726 CN200410031726 CN 200410031726 CN 200410031726 A CN200410031726 A CN 200410031726A CN 1673991 A CN1673991 A CN 1673991A
- Authority
- CN
- China
- Prior art keywords
- wafer
- defect
- defective
- defects
- defectives
- 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
- 238000000034 method Methods 0.000 title claims abstract description 124
- 230000007547 defect Effects 0.000 title claims description 112
- 230000008569 process Effects 0.000 claims abstract description 77
- 238000004519 manufacturing process Methods 0.000 claims abstract description 76
- 239000004065 semiconductor Substances 0.000 claims abstract description 40
- 230000002950 deficient Effects 0.000 claims description 68
- 238000001514 detection method Methods 0.000 claims description 29
- 230000001066 destructive effect Effects 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 13
- 238000000059 patterning Methods 0.000 claims description 6
- 238000007689 inspection Methods 0.000 abstract description 6
- 235000012431 wafers Nutrition 0.000 description 36
- 238000012360 testing method Methods 0.000 description 13
- 238000012544 monitoring process Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000003203 everyday effect Effects 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 4
- 238000012937 correction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 230000035484 reaction time Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- LPXPTNMVRIOKMN-UHFFFAOYSA-M sodium nitrite Substances [Na+].[O-]N=O LPXPTNMVRIOKMN-UHFFFAOYSA-M 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Landscapes
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
Abstract
The fault controlling method for the routine inspection in semiconductor manufacture includes first providing one patterned wafer with several first faults in the surface, passing the wafer through semiconductor manufacture process to form several second faults on the wafer, detecting the faults in the wafer to find out the first faults and the second faults, classifying the detected faults according to one preset data base to separate the first faults and the second faults and separate the second faults into several fault types.
Description
Technical field
The invention provides a kind of defect control method, refer in particular to utilize in a kind of semi-conductor chip manufacture process one patterned wafers (patterned wafer) as the defect control method of monitoring wafer (monitor wafer).
Background technology
In semiconductor fabrication; tend to because some are unavoidable former thereby generate tiny particulate or defective; and along with constantly dwindling of component size in the semiconductor fabrication and improving constantly of circuit level; these atomic little defectives or particulate also are on the rise to the influence of integrated circuit quality; therefore for keeping the stable of product quality; usually when carrying out each semiconductor fabrication; also must constantly test at the semiconductor element of being produced; and adjust each procedure parameter according to test result; to suppress generation of defects, promote the yield rate and the fiduciary level of semiconductor fabrication.
Please refer to Fig. 1, Fig. 1 is the semiconductor fabrication synoptic diagram of a routine.As shown in Figure 1, semiconductor wafer must be through the multiple tracks semiconductor fabrication in manufacturing process, generally speaking, the manufacture process in thousands of roads that semiconductor wafer experiences in a chip factory mostly, for convenience of explanation, only describe defect control method in the conventional semiconductor fabrication among Fig. 1 with wherein several manufacture process.As shown in Figure 1, process A 10, process B 20, process C 30, process D 40 and process E 50 represent five road semiconductor fabrications respectively, and utilize different boards to carry out respectively.For fiduciary level and the degree of stability that promotes each manufacture process; these boards all must carry out some fixing maintenance flow processs; for example at set intervals; more detailed prevention and maintain (preventativemaintenance is carried out once in shutdown; PM) work; and in process of production; also can carry out every day routine inspection (dailycheck); it is in normal production procedure; add a plurality of nude films (bare wafer) as monitoring wafer; and (process A 10 as the aforementioned to carry out each semiconductor fabrication according to same procedure parameter; process B 20; process C 30; process D 40 and process E 50 etc.); respectively each nude film is carried out defect analysis again, whether meet the standard that continues production to differentiate each manufacture process board.
Except the inspection of board itself,, also can carry out some defect inspections usually at the product on the production line for the stability of safeguarding manufacture process and the fiduciary level of product.Because the manufacture process of semiconductor wafer is quite numerous and diverse, the manufacture process that has often comprised thousands of roads, therefore usually when carrying out defects detection, be a station how with several manufacture process, utilize the mode of substation that each product wafer is carried out sampling Detection, for example the defects detection among Fig. 1 60 promptly is to be used for manufacture process A, B, C and D are detected.Please refer to Fig. 2, Fig. 2 is the method synoptic diagram of defects detection 60 in the routine techniques.As shown in Figure 2, scanning manufacture process 130 after the method for normal defect detection 60 is carried out a prescan manufacture process 110 and a manufacture process respectively in predetermined one or multi-channel semiconductor fabrication 120 front and back, so that wafer is carried out defects detection, and the result that this secondary defect is detected compares, to draw the newly-increased defective 140 that in this one or multi-channel semiconductor fabrication 120, produces, then again in artificial mode, allow the slip-stick artist that these newly-increased defectives 140 are carried out SEM one by one and detect (SEM review) 150 again, to carry out follow-up defect cause analytical work, and after finishing the defect cause analysis, more can further revise the procedure parameter of manufacture process board again, take place once more to avoid same defective according to the result who analyzes.
Yet in routine techniques; utilize nude film to carry out routine inspection every day of board no matter be; or product on the line is carried out substation take a sample test; all still there are many shortcomings; for instance; the former need expend a large amount of nude films; and therefore meeting reduces the rate of capacity utilization of board on the production line; cause the significantly lifting of manufacturing cost; and owing to utilize nude film as monitoring wafer; therefore some comprehensive defectives that produce because of the multiple tracks manufacture process can't be discovered in the nude film test; and some sporadic unusual conditions that easily are present in the product often also can can't be handled because of being difficult for observing; even find sometimes such defective also regular meeting and be difficult to effectively carry out defect analysis and eliminating because data deficiencies.
Compare with the former, though the latter does not have above-mentioned shortcoming, but its method is quite complicated, respectively carry out a defective scanning (scanning process 130 after pre-scanning process 110 and the process) before and after the semiconductor fabrication that not only needs desire is detected, after drawing newly-increased defective 140, still need carry out a large amount of SEM and detect 150 work again, because it needs great amount of manpower and time, therefore when carrying out defects detection and analyzing, mostly the mode that must utilize substation to take a sample test is carried out, and can't carry out large-scale detection to the per pass manufacture process, under this situation, often after defective produces, still to pass through several manufacture process, just can carry out defects detection 60, the detection sensitivity of defects detection 60 also can therefore and significantly reduce naturally, and can't control each defective on the product wafer effectively.In addition because conventional defects detection 60 modes need the quite long time just can obtain a result, therefore in case in defects detection 60, find to have on the product wafer unusual condition to take place, and when further the problem of analyzing is certain one manufacture process that comes from wherein (as process B), often all passed through a couple of days, change speech, the wafer of interior manufacturing during this period of time all has same flaw probably, this will cause the decline of manufacture process yield rate and the raising of cost, and along with chip size rises to 12 inches by 8 inches, these problems are also with even more serious.
Therefore, press at present a kind of with low-cost, to unusual condition rapid reaction and highly sensitive defective control mode, to address the above problem.
Summary of the invention
Fundamental purpose of the present invention be to provide a kind of tool have low cost concurrently, to unusual condition rapid reaction and highly sensitive defective control mode, to address the above problem.
Most preferred embodiment of the present invention discloses a kind of semiconductor most of manufacturing defects control method, one wafer of patterning at first is provided, after this wafer is carried out this semiconductor fabrication, this wafer is carried out a defects detection, and utilize a preset database with detected a plurality of automatic defect classifications, filter out to be created on and carry out the preceding preceding layer defects of this semiconductor fabrication, and the defective that this semiconductor fabrication causes is divided into destructive defective and non-destructive defective, when this defective belongs to destructive defective, will send the Email and the defect analysis report of unusual caution automatically, to assist operator's makeover process parameter rapidly, to get rid of this unusual generation.
Because defect control method of the present invention utilizes a product wafer to come defects detection, and utilize a database to carry out the automatic defect classification, therefore not needing to carry out pre-scanning process can separate preceding layer defects and newly-increased defect area come, and can further reduce the burden that the SEM defective detects again, and significantly shorten the reaction time and promote flaw sensitivity, reach the purpose that promotes finished product rate and fiduciary level.
Description of drawings
The defect control method synoptic diagram of the semiconductor fabrication of Fig. 1 and Fig. 2 routine.
Fig. 3 and Fig. 4 are semiconductor most of manufacturing defects control method synoptic diagram among the present invention.
The reference numeral explanation
10 process A, 20 process B
30 process C, 40 process D
50 process E, 60 defects detection
110 pre-scanning process, 120 semiconductor fabrications
150 SEM defectives detect 210 process A again
220 process B, 230 process C
240 process D, 250 process E
260 defects detection, 310 defects detection
320 automatic defects classification 330 caution unusually
340 procedure parameter corrections
Embodiment
Please refer to Fig. 3 and Fig. 4, Fig. 3 and Fig. 4 are the defect control method synoptic diagram among the present invention.As shown in Figure 3, this wafer is in manufacturing process, need five road semiconductor fabrications such as experience process A 210, process B 220, process C230, process D 240 and process E 250, and each semiconductor fabrication all can form a plurality of defectives on this wafer, below is that example illustrates defect control method of the present invention with wherein process B 220.At first, the present invention does not need with extra nude film as monitoring wafer, and directly with on the production line one the wafer of patterning test, change speech, test source of the present invention is product wafer (product wafer), therefore, and after nondestructive method is tested with some, still it can be put back the manufacture process of carrying out next stage on the production line, and can any influence not arranged output.In addition, method of the present invention also can be used to check at different platform or the indoor wafer of differential responses, so that destructive board of potential high finished product rate or reaction chamber are carried out defective control.
And when this wafer behind complete process B 220, will carry out a defects detection 260.As shown in Figure 4, in defect control method of the present invention, defects detection 260 is carried out a defects detection 310 to this wafer earlier, utilize the mode of full-wafer scanning, all defect on this wafer is detected, utilize a preset database to detected a plurality of defective automatic defect classification (automatic defectclassification that carry out again, ADC) 320, and with the setting in these defective one databases, divide into different defect types, in preferred embodiment of the present invention, according to shape, size, parameters such as position will be divided into defect type A in detected a plurality of scarce lines, defect type B, four classes such as defect type C and defect type D.
It should be noted that in this preset database, include all issuable defect types and corresponding defect information thereof in each manufacture process, therefore after finishing defects detection 310, only need compare with the data in this database, newly-increased defective that can will in process B 220, produce with carrying out before the process B 220 separating with regard to the defective that existed (for example process A 210 caused defective), as shown in Figure 4, in preferred embodiment of the present invention, defect type A, defect type B and defect type C are the newly-increased defective that process B 220 is caused, defect type D is the defective for having existed before process B 220 then, that is layer defects before so-called.
In defect control method of the present invention, defect information in the database also include all types of defectives to the influence degree of this manufacture process yield rate with and may occurrence cause, to give different disposal according to the influence degree to this manufacture process yield rate in the various newly-increased defectives (defect type A, defect type B and defect type C).For instance, when carrying out classification of defects, can each defect kind foundation be divided into destructive defective (killer defect) (as defect kind A and B) and non-destructive defective (non-killer defect) (as defect kind C) two classes once more to the influence degree of this manufacture process yield rate according to the data in this database, and further dispose at the part of destructive defective.In preferred embodiment of the present invention, this defect control method will be when detecting destructive defective, can be further going to search in the database it according to detected defect type may occurrence cause (may lead because of in reason A as defect kind A), and send unusual caution 330 at once, the slip-stick artist that notice is responsible for, for example can be (for example: the kind of defective according to the defect type of detected destructive defective and related data, quantity, the position and may generate reason) make defect analysis report, mode with E-mail offers the slip-stick artist, make the slip-stick artist promptly carry out suitable processing according to these data, in the shortest time, carry out procedure parameter correction 340, to solve this unusual condition.For instance, when having ready-made data in the database, the slip-stick artist can directly come the makeover process parameter to avoid the next group product similar situation to take place according to the data that database provided, and needing to determine whether this batch product is heavily processed (rework) or direct discarding according to the situation of defective.And if when there is no proper data in the database, then can carry out further manual analysis to detected defective according to circumstances, and analysis result is updated to this database.
Defect control method of the present invention utilize one the wafer of patterning be used as monitoring wafer, and need not use nude film to carry out routine inspection every day, so can significantly reduce the consumption of nude film, and because can directly test with the product wafer on the line, therefore on the sampling scope, be not limited to them, visual situation is carried out pick test to the product on the line arbitrarily, even therefore some accidental unusual conditions also can effectively detect, and carries out defect analysis and eliminating.In addition, in routine techniques, some comprehensive defectives often are difficult for finding, even if observed, also being difficult to find out its problem is to come from which road manufacture process on earth, if desire to make at short notice judgement, be easy to do the analysis result that makes mistake, and if desire increases the fiduciary level of defect analysis, then need spend a large amount of time and manpower and remove to carry out large-scale Test And Checkout, just can find out the true cause of problem.The present invention then cooperates the use of a database, and utilizes this database to provide quite detailed data to allow the operator further analyze, and judges rightly so can promptly make the occurrence cause of this defective, and suitably handles.
It should be noted that defect control method of the present invention also includes a classification of defects step, because the type of defective is very many, a lot of defectives seriously influence manufacture process, but also there are a lot of defectives that manufacture process be there is no too big influence, and the critical defect that can influence each manufacture process also is not quite similar, therefore under the use that cooperates a presetting database and an automatic defect classification tool, the step of utilizing this classification of defects is directly with preceding layer defects, destructive defective is separated with the non-destructive defective, compare with routine techniques, not only can dispense a defective scanning work (pre-scanning process), effectively reduce and detect the defective scanning work of board when carrying out defects detection 260, and can allow the operator can be absorbed in destructive defective to handle, and needn't on the non-destructive defective, lose time, significantly to reduce the working load that the SEM defective detects again, reach the effect in the reaction time of reduction defect analysis and eliminating.
Because the working load of ground defects detection can significantly fall in automatic defect classification instrument used in the present invention, therefore defect control method of the present invention can be applied to the multiple tracks semiconductor fabrication simultaneously, (as process B) monitors one by one to the destructive semiconductor fabrication of each high finished product rate, make wafer one finish the semiconductor fabrication of institute's pre-monitoring after, to directly carry out defects detection 260, and do not need as after carrying out the multiple tracks semiconductor fabrication, just carrying out one time defects detection in the conventional technology, therefore can significantly promote the flaw detection sensitivity in the defect control method of the present invention.
In addition, defective control mode of the present invention is used as testing the target except a product wafer, also can use nude film according to circumstances instead tests, for instance, in case when generation equipment abnormal occurrence maybe will be carried out prevention and maintain, also can cooperate making of nude film to be used for carrying out defect control method of the present invention, in the hope of reaching a better detecting effect.
Compare with the defective control mode in the routine techniques, defective control mode of the present invention directly utilizes on the line the wafer of patterning to test, so need not utilize nude film to carry out customary detection every day, therefore can effectively reduce the use of nude film and propose production can utilization factor.In addition, for the comprehensive defective of some sporadic unusual conditions and the generation of multiple tracks manufacture process, the present invention obviously has more higher sensitivity, under the use that cooperates database and automatic defect classification tool, more can significantly shorten the time of defect analysis and eliminating, effectively reduce the manufacturing cost of semiconductor wafer.
The above only is preferred embodiment of the present invention, and all equalizations of being done according to the present patent application claim change and modify, and all should belong to the covering scope of patent of the present invention.
Claims (7)
1. semiconductor most of manufacturing defects control method, it includes the following step:
One wafer of patterning is provided, and this wafer surface includes a plurality of first defectives;
This wafer is carried out the semiconductor manufacture process, and this semiconductor fabrication forms a plurality of second defectives on this wafer;
This wafer is carried out a defects detection, with in a plurality of first defectives and these a plurality of second defectives of detecting on this wafer; And
Detected these defectives are carried out classification of defects according to preset database, should a plurality of first defectives separating with these a plurality of second defectives, and should a plurality of second classifications of defects be a plurality of defect types according to this database.
2. method as claimed in claim 1 wherein includes the mode classification of various defect types and corresponding to the defect information of this defect type respectively in this database.
3. method as claimed in claim 2, wherein respectively this defect information of this defect type includes respectively this defect type to the yield rate influence degree of this semiconductor fabrication.
4. method as claimed in claim 3, wherein this method can be divided into destructive defective and non-destructive defective to the yield rate influence degree of this semiconductor fabrication with detected these second defect areas according to this defect type respectively after carrying out classification of defects.
5. method as claimed in claim 4, wherein this method also includes the following step when detecting destructive defective:
Carry out the defective generation analysis of causes according to this defect type under in detected this defective; And
Notify the operator of this semiconductor fabrication, the procedure parameter of this semiconductor fabrication is revised to assist this operator.
6. method as claimed in claim 1, wherein this method utilizes an on-line automatic classification of defects instrument to carry out classification of defects.
7. method as claimed in claim 1, wherein this wafer of patterning be product wafer on the line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410031726 CN1673991A (en) | 2004-03-24 | 2004-03-24 | Defect control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410031726 CN1673991A (en) | 2004-03-24 | 2004-03-24 | Defect control method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1673991A true CN1673991A (en) | 2005-09-28 |
Family
ID=35046544
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410031726 Pending CN1673991A (en) | 2004-03-24 | 2004-03-24 | Defect control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1673991A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101295659B (en) * | 2007-04-29 | 2010-06-09 | 中芯国际集成电路制造(上海)有限公司 | Method for detecting defect of semiconductor device |
| CN101079303B (en) * | 2006-05-22 | 2012-03-21 | 日立高新技术有限公司 | Method of detecting scratch defect in circumferential direction and magnetic disk detecting device |
| CN101581677B (en) * | 2008-05-16 | 2012-10-03 | 深超光电（深圳）有限公司 | Method for classifying grade of panel to be detected of lighting tester |
| CN103871921A (en) * | 2014-02-21 | 2014-06-18 | 上海华力微电子有限公司 | Wafer defect monitoring analysis method based on server |
| CN105204377A (en) * | 2014-06-18 | 2015-12-30 | 上海华力微电子有限公司 | Method for improving product standard |
| CN108364890A (en) * | 2018-02-24 | 2018-08-03 | 上海华力微电子有限公司 | The monitoring method of systematic defect |
-
2004
- 2004-03-24 CN CN 200410031726 patent/CN1673991A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101079303B (en) * | 2006-05-22 | 2012-03-21 | 日立高新技术有限公司 | Method of detecting scratch defect in circumferential direction and magnetic disk detecting device |
| CN101295659B (en) * | 2007-04-29 | 2010-06-09 | 中芯国际集成电路制造(上海)有限公司 | Method for detecting defect of semiconductor device |
| CN101581677B (en) * | 2008-05-16 | 2012-10-03 | 深超光电（深圳）有限公司 | Method for classifying grade of panel to be detected of lighting tester |
| CN103871921A (en) * | 2014-02-21 | 2014-06-18 | 上海华力微电子有限公司 | Wafer defect monitoring analysis method based on server |
| CN105204377A (en) * | 2014-06-18 | 2015-12-30 | 上海华力微电子有限公司 | Method for improving product standard |
| CN108364890A (en) * | 2018-02-24 | 2018-08-03 | 上海华力微电子有限公司 | The monitoring method of systematic defect |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9037280B2 (en) | Computer-implemented methods for performing one or more defect-related functions | |
| US7332359B2 (en) | Semiconductor device inspection method | |
| US6757621B2 (en) | Process management system | |
| US6635872B2 (en) | Defect inspection efficiency improvement with in-situ statistical analysis of defect data during inspection | |
| TW518645B (en) | Method and system of automatic wafer manufacture quality control | |
| US7728969B2 (en) | Methods and systems for identifying defect types on a wafer | |
| US20040228515A1 (en) | Method of inspecting defects | |
| US20040254752A1 (en) | System for identification of defects on circuits or other arrayed products | |
| US20020027653A1 (en) | Method for inspecting defects and an apparatus of the same | |
| KR101183133B1 (en) | System and method for haze control in semiconductor processes | |
| KR20070093046A (en) | Flexible hybrid defect classification for semiconductor manufacturing | |
| KR20110048528A (en) | Computer-implemented methods for inspecting and / or sorting wafers | |
| CN1673991A (en) | Defect control method | |
| JP4080087B2 (en) | Analysis method, analysis system, and analysis apparatus | |
| CN100375258C (en) | Method for detecting again fault | |
| KR102380099B1 (en) | Range-Based Real-Time Scanning Electron Microscopy Non-Visual Wiener | |
| JP2004296826A (en) | Quality control system and quality control method | |
| US20070031982A1 (en) | Method of classifying defects and apparatus for performing the method | |
| JP2001110867A (en) | Manufacture of electronic device and quality control system for electronic device | |
| US20050080572A1 (en) | Method of defect control | |
| CN107923856B (en) | Non-visual grid for range-based real-time scanning electron microscope | |
| CN1673992A (en) | A method for establishing defect data bank | |
| JP2004165395A (en) | Inspection data analysis program and inspecting program | |
| Bakker | Applied use of advanced inspection systems to measure, reduce, and control defect densities | |
| JPH10229110A (en) | Manufacture of semiconductor device and semiconductor device manufactured by the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |