CN100590832C - Method for detecting silicon chip edge automatically - Google Patents

Method for detecting silicon chip edge automatically Download PDF

Info

Publication number
CN100590832C
CN100590832C CN200710094319A CN200710094319A CN100590832C CN 100590832 C CN100590832 C CN 100590832C CN 200710094319 A CN200710094319 A CN 200710094319A CN 200710094319 A CN200710094319 A CN 200710094319A CN 100590832 C CN100590832 C CN 100590832C
Authority
CN
China
Prior art keywords
silicon chip
coordinate
klarf
virtual
file
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.)
Active
Application number
CN200710094319A
Other languages
Chinese (zh)
Other versions
CN101452867A (en
Inventor
王亚东
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Huahong Grace Semiconductor Manufacturing Corp
Original Assignee
Shanghai Huahong NEC Electronics Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Huahong NEC Electronics Co Ltd filed Critical Shanghai Huahong NEC Electronics Co Ltd
Priority to CN200710094319A priority Critical patent/CN100590832C/en
Publication of CN101452867A publication Critical patent/CN101452867A/en
Application granted granted Critical
Publication of CN100590832C publication Critical patent/CN100590832C/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a method for automatically detecting the edge of a silicon chip. The method comprises: step one, an automatic defect detector scans the silicon chip and stores a scanning result to form a file with a KLARF format; step two, a virtual crystal element is added to the KLARF file stored in the step one; the virtual crystal element integrally covers the edge of the whole siliconchip; step three, a coordinate of a detection point of the edge of the silicon chip is calculated in a coordinate system adopted by the KLARF file; step four, the coordinate of the detection point calculated in the step three is used as virtual defect and is added to the KLARF file edited in the step two; and step five, an electron scanning microscope is used to carry out automatic detection on the virtual defect of the KLARF file edited in the step four. Through functional integration of data analytic software and a detection machine, the method realizes automatic detection on the position of the edge of the silicon chip, saves manpower and material resource, improves efficiency and strengthens data accuracy and analyzability.

Description

Automatically detect the method for silicon chip edge
Technical field
The present invention relates to a kind of semiconductor fabrication process, particularly relate to a kind of method that detects the defective of silicon chip edge
Background technology
Integrated circuit manufacturing enterprise (had promptly carried out some technology such as etching technics to graph silicon chip is arranged on silicon chip at present, and on silicon chip, formed the silicon chip of certain figure such as etching figure) the edge detect, be by electronic scanner microscope by manually the operation.
In a production procedure, the defective of silicon chip edge is to produce in which step process, how the already present defective of silicon chip edge develops in subsequent technique, just address these problems the defects detection of silicon chip edge being carried out repeatedly, repeating between need per two technologies in a production procedure.This detection manually need be expended lot of manpower and material resources, and efficient and precision are all very poor, can't accurately locate test point, and the analyticity of data is also relatively poor.
Summary of the invention
Technical problem to be solved by this invention provides a kind of method that the edge that graph silicon chip is arranged is detected automatically by electronic scanner microscope.
For solving the problems of the technologies described above, the present invention detects the method for silicon chip edge automatically, described silicon chip edge, and promptly the no graphics field of silicon chip edge and abandon surveyed area comprises the steps:
The first step, the automatic defect detector scans silicon chip, and scanning result is saved as the file of KLARF form;
Second step added virtual brilliant unit in the KLARF file that the first step is preserved, described virtual brilliant unit intactly covers the whole silicon wafer edge;
The 3rd step, the coordinate of test point in the coordinate system that the KLARF file is adopted of calculating silicon chip edge;
In the 4th step, add the 3rd coordinate that goes on foot the test point that calculates in the second step editor's the KLARF file as virtual defective;
In the 5th step, electronic scanner microscope detects automatically to the virtual defective in the 4th step editor's the KLARF file.
As a further improvement on the present invention, also comprise between the 4th step of this method and the 5th step, electronic scanner microscope is according to the coordinate of the one or more real defects in the KLARF file of first step preservation, and the KLARF file that the 4th step was edited utilizes the coordinate of identical real defect to calibrate.
The present invention detects the method for silicon chip edge automatically, not only can carry out defects detection to the silicon chip edge that figure is arranged, more can in a production procedure, do complete monitoring to silicon chip edge, for example understand defective and produce in which step process, how the defective that has produced develops in subsequent technique etc.And by automatic detection and the monitoring that electronic scanner microscope carried out, more can use manpower and material resources sparingly, and improve and detect and monitoring efficiency, strengthen the accuracy of test point location, improve the analyticity of the data that obtain.
Description of drawings
The present invention is further detailed explanation below in conjunction with drawings and Examples:
Fig. 1 is the method flow diagram that the present invention detects silicon chip edge automatically;
Fig. 2 is the schematic diagram of automatic defect scanner to the silicon chip scan image;
Fig. 3 is the schematic diagram that increases virtual brilliant unit for silicon chip;
Fig. 4 is the schematic diagram of coordinate that calculates the test point of silicon chip edge;
Fig. 5 is the schematic diagram that increases virtual defective for silicon chip;
Reference numeral is among the figure: the 10-silicon chip; The brilliant unit of 11-; 12-does not have the graphics field; 13-abandons surveyed area; The 14-defective; The virtual defective of 15-; 20-silicon chip actual boundary; The border of institute overlay area behind the 21-KLARF file modification; The 30-test point; The search window of 31-electronic scanner microscope.
Embodiment
See also Fig. 1, the method that the present invention monitors silicon chip edge comprises the steps:
The first step scans silicon chip by the automatic defect detector, and scanning result is saved as the file of KLARF form.Can write down the size and the coordinate of each real defect on the silicon chip during automatic defect detector scanning silicon chip.
Second step added virtual brilliant unit in the KLARF file that the first step is preserved, described virtual brilliant unit intactly covers the whole silicon wafer edge.Add after the virtual brilliant unit, the actual brilliant unit of this KLARF file and virtual brilliant first total size that is covered are just greater than the actual range of silicon chip.
The 3rd step, the coordinate of test point in the coordinate system that the KLARF file is adopted of calculating silicon chip edge.The coordinate system that the KLARF file is adopted is that the automatic defect detector is self-defining when each silicon chip of scanning, and its origin of coordinates often is not the center of silicon chip.For circular silicon chip, it is known being positioned at each test point of silicon chip edge and the distance in the center of circle.If detect the silicon chip edge outermost, the distance in the test point and the center of circle just is the silicon chip radius so; If detect the inner slightly position of silicon chip edge, the distance in the test point and the center of circle is slightly smaller than the silicon chip radius so.In order to detect, need a unlimited test point in theory to the whole silicon wafer edge.In the practical operation, can reach the purpose of data acquisition, so only need limited test point of setting to get final product at ad-hoc location to indivedual positions sampling Detection in a certain zone.Usually, set a plurality of test points equably at silicon chip edge, for example evenly set 12 test points in 360 degree circumference, the formed angle of line in every adjacent two test points and the center of circle is 30 degree.
In the 4th step, the 3rd coordinate that goes on foot the test point that calculates is added in the second step editor's the KLARF file as virtual defective.Like this, both comprised in the KLARF file, comprised the virtual defective that the user adds again by the real defect of automatic fault detection system scanning.
In the 5th step, the KLARF file that the 4th step was edited imports to automatic fault detection system, is resent to the board of electronic scanner microscope, is converted to the file that electronic scanner microscope detects automatically.Electronic scanner microscope detects the virtual defective in the 4th step editor's the KLARF file automatically as test point.
As a further improvement on the present invention, also comprise between the 4th step of this method and the 5th step, electronic scanner microscope is according to the coordinate of the one or more real defects in the KLARF file of first step preservation, and the KLARF file that the 4th step was edited utilizes the coordinate of identical real defect to calibrate.Because in the data file that the automatic defect scanner generates, brilliant first coordinate, brilliant first initial point, defect coordinate etc. all are explications, in the operation in the 5th step, need again editor KLARF file is sent to electronic scanner microscope, at this moment require the detection initial point of electronic scanner microscope and the initial point inregister that the automatic defect scanner is set, so this improvement project can be used for the calibration of coordinate.
In above-mentioned improvement project, the defective that is useful on calibrating coordinates in order to make is real defect, need distinguish real defect and virtual defective.A kind of comparatively simple method is to distinguish from the size of defective.When the coordinate with test point joins the KLARF file as virtual defective, set the size of these virtual defectives simultaneously, for example all more than or equal to 10 μ m.And when selecting to be used for the real defect of calibrating coordinates, only select those real defects less than (preferably being significantly less than) all virtual defectives, for example only select virtual defective smaller or equal to 5 μ m.Can guarantee that like this defective that is used for calibrating coordinates is real defect.
See also Fig. 2, the blockage of dividing on the circular silicon chip 10 one by one is brilliant unit 11.The black at silicon chip 10 edges partly is no graphics field 12, because this zone is within the scope of photolithographic exposure, so the automatic defect detector does not detect this zone.The blockage that is filled with the fork-shaped mark near no graphics field 12 is to abandon surveyed area 13, the insufficient or other reasons owing to expose when photolithographic exposure in this zone, so the automatic defect detector is abandoned this zone is detected.Purpose of the present invention is exactly to no graphics field 12 and abandon surveyed area 13 and carry out defects detection.The a plurality of points that distribute on the silicon chip 10 are the detected real defective 14 of automatic fault detection system.Automatic fault detection system is at scanning all self-defined coordinate system during each silicon chip, and as the XY coordinate system that the X-axis among Fig. 2 and Y-axis are formed, the initial point of this XY coordinate system often is not positioned at the central authorities of silicon chip 10.
See also Fig. 3, circular silicon chip 10 is surrounded by silicon chip actual boundary 20, and in the KLARF file that the first step generated, the scope that the KLARF file is covered also is that silicon chip actual boundary 20 is with interior zone.Second step of the method for the invention increases virtual brilliant unit by revising the KLARF file, and the virtual brilliant unit of increase is with the 20 complete coverings of silicon chip actual boundary, and in other words, the every bit at the edge of silicon chip 10 all falls into the scope of virtual brilliant unit.The coverage of amended KLARF file is greater than the silicon chip actual range, and the border 21 of amended KLARF file institute overlay area is obviously enclosed than silicon chip actual boundary 20 big last.
See also Fig. 4 (a), acquiescence is the XY coordinate system record coordinate that adopts with the KLARF file on the silicon chip 10.Test point 30 is positioned at silicon chip 10 edges, and the distance of the central authorities (center of circle) of itself and silicon chip 10 is the radius r of silicon chip 10, and the coordinate in the center of circle of known silicon chip 10 is that (h, l), problem is the coordinate (x that how to find the solution test point 30 0, y 0).
At first, according to Pythagorean theorem (x is arranged 0-h) 2+ (y 0-l) 2=r 2(1)
Secondly, supposing has straight line from the initial point of silicon chip 10 to test point 30, and the slope of this straight line is k, so k=(y 0-l)/(x 0-h); (2)
Suppose that here slope k is known, its reason is in the back narration.The equation group of being formed by equation (1) and (2) promptly can solve the coordinate (x of test point 30 in the XY coordinate system so 0, y 0).
If wish to detect the slightly inner position of silicon chip edge, only need the smaller slightly coordinate that can find the solution the test point that makes new advances of change r.The solving equation group can be finished by manual calculations or coding, also can finish by means of relevant software for calculation.
According to the record rule of KLARF file, the coordinate (x of test point 30 0, y 0) need be converted to the numbering of the brilliant unit in this test point 30 places and this test point 30 with respect to the coordinate in brilliant first lower left corner, place, just can record in the KLAEF file as virtual defective then.
See also Fig. 4 (b), the KLARF file is that the initial point of XY coordinate system is defined as [0 to the definition rule of brilliant unit numbering, 0] lower left corner of brilliant unit, the brilliant unit on [0,0] brilliant unit the right is numbered [1,0], the brilliant unit on the left side is numbered [1,0], the brilliant unit of top is numbered [0,1], following brilliant unit is numbered [0 ,-1].
The coordinate of known test point 30 is (x 0, y 0), each brilliant unit all is α in length of X-axis projection, all is β in the length of Y-axis projection, need find the solution the numbering [m, n] of the brilliant unit in test point 30 places, and test point 30 is with respect to the coordinate (α in [m, n] brilliant first lower left corner 0, β 0).
According to the definition rule of KLARF file to brilliant unit numbering, m is x 0Round divided by α, n is y again 0Round α again divided by β 0,=x 0-m * α, β 0=y 0-n * β.
See also Fig. 5, be distributed with one or more real defectives 14 on the circular silicon chip 10, increased one or more virtual defectives 15 at the edge of silicon chip 10.The coordinate of these virtual defectives calculated by the 3rd step, can also set the size of these virtual defectives simultaneously.
Be distributed with one or more real defects 14 on the circular silicon chip 10, the edge distribution of silicon chip 10 has one or more virtual defectives 15, and these virtual defectives 15 are the test point 30 of electronic scanner microscope just.Virtual defective 15 among the figure evenly distributes every 30 degree along the edge of silicon chip 10.
Because in the practical operation, virtual defective 15 all evenly distributes along silicon chip edge, so the slope k of the line in the center of circle of each virtual defective 15 and silicon chip 10 is known.
Following automatic defect checkout equipment kla2351 cooperates automatic defect trace routine Odyssey, and electronic scanner microscope SEMVision G2 is example, introduces the concrete enforcement of the method for the invention.
The first step is used automatic defect checkout equipment kla2351 scanning silicon chip, opens scanning result in automatic defect trace routine Odyssey, and scanning result is saved as the KLARF file.
Second step, open the KLARF file that the first step generates with text mode, automatic defect trace routine Odyssey has defined the coordinate system of this KLARF file voluntarily when generating this KLARF file, from [00] at coordinate origin place brilliant unit, the X axis right side increases progressively, and Y-axis upwards increases progressively.Revise the coverage in this KLARF file, in other words, increase virtual brilliant unit in the periphery of the actual brilliant unit of silicon chip, the virtual brilliant unit that is increased at least will the complete covering with the edge of silicon chip.
The 3rd step, the coordinate in the defined coordinate system of KLARF file of the test point of accurate Calculation silicon chip edge.
In the 4th step, the coordinate of test point is joined in the KLARF file as virtual defective.
In the 5th step, the KLARF file that the 4th step was revised imports among the automatic defect trace routine Odyssey, is resent to electronic scanner microscope SEMVision G2.Electronic scanner microscope is calibrated according to the coordinate of real defect earlier, regulates the FOV value then and obtains suitable multiplying power and search window, saves as the file that electronic scanner microscope SEMVision G2 can automatically perform.

Claims (7)

1. method that automatically detects silicon chip edge, described silicon chip edge, i.e. the no graphics field of silicon chip edge and abandon surveyed area, it is characterized in that: this method comprises the steps:
The first step, the automatic defect detector scans silicon chip, and scanning result is saved as the file of KLARF form;
Second step added virtual brilliant unit in the KLARF file that the first step is preserved, described virtual brilliant unit intactly covers the whole silicon wafer edge;
The 3rd step, the coordinate of test point in the coordinate system that the KLARF file is adopted of calculating silicon chip edge;
In the 4th step, add the 3rd coordinate that goes on foot the test point that calculates in the second step editor's the KLARF file as virtual defective;
In the 5th step, electronic scanner microscope detects automatically to the virtual defective in the 4th step editor's the KLARF file.
2. the method for automatic detection silicon chip edge according to claim 1 is characterized in that: the KLARF file that the first step of this method is preserved comprises the size and the coordinate of each real defect on the silicon chip.
3. the method for automatic detection silicon chip edge according to claim 2, it is characterized in that: also comprise between the 4th step of this method and the 5th step, electronic scanner microscope is according to the coordinate of the one or more real defects in the KLARF file of first step preservation, and the KLARF file that the 4th step was edited utilizes the coordinate of identical real defect to calibrate.
4. the method for automatic detection silicon chip edge according to claim 1 is characterized in that: in the 4th step of this method, also be included in the size that virtual defective is set in the KLARF file.
5. the method for automatic detection silicon chip edge according to claim 4 is characterized in that: the size of the described virtual defective that in the KLARF file, is provided with, and all greater than the maximum of the size of the real defect that is used to calibrate.
6. the method for automatic detection silicon chip edge according to claim 1 is characterized in that: the second step editor's of this method KLARF file, the total size that its actual brilliant unit and virtual brilliant unit are covered is greater than the scope of silicon chip reality.
7. the method for automatic detection silicon chip edge according to claim 1 is characterized in that: described test point is one or more, and described test point is distributed in the silicon chip edge optional position.
CN200710094319A 2007-11-28 2007-11-28 Method for detecting silicon chip edge automatically Active CN100590832C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN200710094319A CN100590832C (en) 2007-11-28 2007-11-28 Method for detecting silicon chip edge automatically

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN200710094319A CN100590832C (en) 2007-11-28 2007-11-28 Method for detecting silicon chip edge automatically

Publications (2)

Publication Number Publication Date
CN101452867A CN101452867A (en) 2009-06-10
CN100590832C true CN100590832C (en) 2010-02-17

Family

ID=40735023

Family Applications (1)

Application Number Title Priority Date Filing Date
CN200710094319A Active CN100590832C (en) 2007-11-28 2007-11-28 Method for detecting silicon chip edge automatically

Country Status (1)

Country Link
CN (1) CN100590832C (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103531498B (en) * 2013-10-18 2016-04-20 上海华力微电子有限公司 Wafer defect analytical system
US10056224B2 (en) * 2015-08-10 2018-08-21 Kla-Tencor Corporation Method and system for edge-of-wafer inspection and review
CN107993955A (en) * 2017-11-24 2018-05-04 上海华力微电子有限公司 A kind of method for detecting defects of wafer edge

Also Published As

Publication number Publication date
CN101452867A (en) 2009-06-10

Similar Documents

Publication Publication Date Title
US7630539B2 (en) Image processing apparatus
US7483560B2 (en) Method for measuring three dimensional shape of a fine pattern
US7486814B2 (en) Local bias map using line width measurements
CN100431126C (en) User interface for quantifying wafer non-uniformities and graphically explore significance
US7420168B2 (en) Scanning electron microscope and CD measurement calibration standard specimen
US7310585B2 (en) Method of inspecting integrated circuits during fabrication
CN107516624A (en) A kind of sample position calibration method and device
US8090192B2 (en) Pattern misalignment measurement method, program, and semiconductor device manufacturing method
US20210231580A1 (en) Method for smart conversion and calibration of coordinate
JP4943616B2 (en) Straight line defect detection
CN100590832C (en) Method for detecting silicon chip edge automatically
CN105988305B (en) Wafer pre-alignment method
CN106500633A (en) A kind of measurement method of planeness
CN105719993A (en) Method for correcting position deviations of electron microscope electron beam and wafer
US20040057610A1 (en) Mask defect analysis for both horizontal and vertical processing effects
JP2003303868A (en) Inspection condition setting program, inspecting apparatus, and inspecting system
US8095896B2 (en) Method and system of displaying an exposure condition
US6115042A (en) Interface for model definition
JP2004536300A5 (en)
CN103646885B (en) A kind of method reducing electron microscope observation wafer defect error
CN102566322A (en) Multiple photoetching equipment correcting method
CN104851817B (en) Electron-beam measuring optimization method
US20190230517A1 (en) Methods, Systems and Computer Program Products for Automatic Calibration of Antennas
CN104598904B (en) The localization method and equipment at a kind of graphic code correction graph center
US20060147103A1 (en) Device for identifying a structure to be applied to a substrate, and corresponding methods

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
C14 Grant of patent or utility model
ASS Succession or assignment of patent right

Owner name: SHANGHAI HUAHONG GRACE SEMICONDUCTOR MANUFACTURING

Free format text: FORMER OWNER: HUAHONG NEC ELECTRONICS CO LTD, SHANGHAI

Effective date: 20131216

TR01 Transfer of patent right

Effective date of registration: 20131216

Address after: 201203 Shanghai city Zuchongzhi road Pudong New Area Zhangjiang hi tech Park No. 1399

Patentee after: Shanghai Huahong Grace Semiconductor Manufacturing Corporation

Address before: 201206, Shanghai, Pudong New Area, Sichuan Road, No. 1188 Bridge

Patentee before: Shanghai Huahong NEC Electronics Co., Ltd.

C41 Transfer of patent application or patent right or utility model
COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 201206 PUDONG NEW AREA, SHANGHAI TO: 201203 PUDONG NEW AREA, SHANGHAI