CN103777460A - Method for improving precision of optical proximity effect correction model - Google Patents
Method for improving precision of optical proximity effect correction model Download PDFInfo
- Publication number
- CN103777460A CN103777460A CN201410076758.3A CN201410076758A CN103777460A CN 103777460 A CN103777460 A CN 103777460A CN 201410076758 A CN201410076758 A CN 201410076758A CN 103777460 A CN103777460 A CN 103777460A
- Authority
- CN
- China
- Prior art keywords
- residual error
- sampled point
- mean square
- model
- value
- 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
Images
Abstract
The invention provides a method for improving precision of an optical proximity effect correction model. The method comprises the following steps: sampling on a wafer to obtain a group of measurement data values; processing the measurement data values, filtering measurement data of incredible sampling points; building an initial optical proximity effect (OPC) model on the basis of the processed measurement data; computing the original residual error of the model data value of each sampling point; obtaining corrected residual error mean square root by using a squared value or a cubic value of the original residual error; and simulating an optical system by using a corrected residual error mean square root guide algorithm, so as to obtain the corrected OPC model. By adopting the method provided by the invention, the optical system is effectively simulated by using the corrected residual error mean square root guide algorithm, and the residual error of the overall model is effectively reduced in the simulating process. Thus, the precision of the corrected OPC model is improved, and the stability of a photoetching technology is further improved.
Description
Technical field
The present invention relates to technical field of semiconductors, particularly a kind of method that improves optical proximity correction model accuracy.
Background technology
Along with the sustainable development of integrated circuit, manufacturing technology is constantly to the development of less size, and lithographic process has become limit ic to the Main Bottleneck of small-feature-size development more.In the semiconductor fabrication process of deep-submicron, key graphic size has been far smaller than the wavelength of light source, can cause like this diffraction of light effect, thereby the figure that causes light shield to be projected on silicon chip can change a lot, as various optical approach effects such as the shortenings of the sphering of the variation of live width, corner, line length.
The error producing in order to compensate these effects, we can directly revise the figure designing, and then carry out the plate-making work of reticle, for example, line tail is modified as to the figure of tup (hammer head) and so on etc.The iterative process of this correction is just photoetching Proximity effect correction, i.e. so-called OPC(Optical Proximity Correction).In general 0.18 micron of following lithographic process need to be aided with OPC and just can obtain good photoetching quality.
In OPC process, setting up of model is most important, and the correction of mask (mask) is all the emulation based on model and realizes by a large amount of iteration.OPC model simulates according to information such as photoresistance on optical system correlation parameter, wafer and rete information parameter, reticle transmissivity parameters the pattern forming on wafer after the pattern projection through designing in reticle.Adopt after the projection that OPC pattern die draws up pattern, with actual, the pattern forming that exposes to the sun of pattern in reticle is existed to error on wafer, this error is called model residual error (model residual error, MRE).The existence of error has limited the precision of OPC model.
On wafer, the characteristic dimension of pattern (CD) measures under believable prerequisite, conventionally wish the residual error of model to reduce as far as possible, thus lift scheme precision.
In existing method, the method for setting up OPC model comprises:
Step L01: sample on wafer, obtain one group of metric data value;
Step L02: above-mentioned metric data value is processed, filtered out the metric data of incredible sampled point; Here, incredible sampled point can be, but not limited to comprise: larger sampled point, the metric data of the metric data in homogeneous measurement process fluctuation is not significantly less than the sampled point that sampled point, trend and the homogeneous data of design rule size obviously depart from, etc.
Step L03: the metric data value based on above-mentioned processing is set up initial OPC model, calculates the residual error of the model data value of each sampled point.Wherein, the model data value of each sampled point and above-mentioned metric data value are compared one by one, obtain the residual error of the model data value of each sampled point; For example, measCD1, measCD2 ... measCDn is respectively the 1st sampled point, the 2nd sampled point ... the metric data value of n sampled point, modelCD1, modelCD2 ... modelCDn is respectively first sampled point, the 2nd sampled point ... the model data value of n sampled point, the residual error MRE1 of the model data value of each sampled point, MRE2 ... MREn is respectively:
MRE1=modelCD1-measCD1
MRE2=modelCD2-measCD2
…
MREn=modelCDn-measCDn
Step L04: calculate the residual error root mean square of above-mentioned model data value, the formula of employing is:
M=[(MRE1)
2+(MRE1)
2+…+(MREn)
2]
1/2
Wherein, M represents residual error root mean square.
Step L05: the guiding numerical value using M value as the algorithm of OPC model, bootstrap algorithm is made effective simulation to optical system.
But, in actual simulation process, for the less or larger data point of residual error, can also further revise, the residual error of OPC model is reduced, thereby further improve the precision of OPC model.
Summary of the invention
In order to overcome the deficiencies in the prior art, the present invention aims to provide a kind of residual error and root mean square method thereof revised, thereby improves the precision of OPC model.
The invention provides a kind of method that improves optical proximity effect correction model precision, it comprises the following steps:
Step S01: sample on wafer, obtain one group of metric data value;
Step S02: described metric data value is processed, filtered out the metric data of incredible sampled point;
Step S03: the metric data value based on described processing is set up initial OPC model, calculates the original residual error of the model data value of each sampled point;
Step S04: utilize the square value of described original residual error or a cube value to obtain described revised residual error root mean square;
Step S05: utilize described revised residual error root mean square bootstrap algorithm to carry out the simulation of optical system, thereby obtain revised OPC model.
Preferably, in described step S04, while utilizing the square value of described original residual error to obtain revised residual error root mean square, adopt following formula: M=[(MRE1
2)
2+ (MRE
2)
2+ ... + (MREn
2)
2]
1/2, wherein, MRE1, MRE2 ..., MREn represent respectively the 1st sampled point, the 2nd sampled point ..., a n sampled point described original residual error, M represents described revised residual error root mean square.
Preferably, in described step S04, while utilizing cube value of described original residual error to obtain revised residual error root mean square, adopt following formula: M=[(MRE1
3)
2+ (MRE
3)
2+ ... + (MREn
3)
2]
1/2, wherein, MRE1, MRE2 ..., MREn represent respectively the 1st sampled point, the 2nd sampled point ..., a n sampled point described original residual error, M represents described revised residual error root mean square.
Preferably, described figure of sampling on wafer comprises some one dimension figures and some X-Y schemes simultaneously.
Preferably, described one dimension figure and described X-Y scheme comprise several different feature pattern.
Preferably, described step S02 comprises: after the several different repetitive position of same figure is repeatedly measured, average, and remove the metric data value of the sampled point that in repeatedly measurement process, fluctuation is larger, and remove the metric data of the sampled point that is significantly less than design rule size.
The method of raising OPC model accuracy of the present invention, square value or cube value with the initial residual error of initial OPC model are revised initial residual error mean square root, further utilize revised residual error root mean square to carry out bootstrap algorithm optical system is made to effective simulation, thereby obtain the OPC model that revised precision is improved, correcting optical adjacency effect, the precision of raising etching technics.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of the method for the raising OPC model accuracy of a preferred embodiment of the present invention
Fig. 2 is the distribution schematic diagram of the original residual error of the metric data of above-mentioned preferred embodiment of the present invention
Fig. 3 be above-mentioned preferred embodiment of the present invention metric data original residual error through square after the distribution schematic diagram of residual error
Fig. 4 be above-mentioned preferred embodiment of the present invention metric data original residual error through cube after the distribution schematic diagram of residual error
Embodiment
For making content of the present invention more clear understandable, below in conjunction with Figure of description, content of the present invention is described further.Certainly the present invention is not limited to this specific embodiment, and the known general replacement of those skilled in the art is also encompassed in protection scope of the present invention.
Below in conjunction with specific embodiments and the drawings 1 and accompanying drawing 2, the method for raising OPC model accuracy of the present invention is described in further detail.Fig. 1 is the schematic flow sheet of the method for the raising OPC model accuracy of a preferred embodiment of the present invention.It should be noted that, accompanying drawing all adopts very the form simplified, uses non-ratio accurately, and only in order to object convenient, that reach lucidly the aid illustration embodiment of the present invention.
As previously mentioned, in the simulation process of existing OPC model, adopt the residual error root mean square bootstrap algorithm of model to carry out the simulation of optical system, thereby obtain the OPC model of revising, this residual error is directly the model data value of each sampled point and metric data value to be compared and obtained, because model residual error has limited the precision of OPC model, therefore, the present invention revises the root mean square algorithm of model residual error, thereby bootstrap algorithm carries out the simulation of optical system, to obtain the OPC model that revised precision is higher.
The method of raising OPC model accuracy of the present invention, refers to Fig. 1, comprises the following steps:
Step S01: sample on wafer, obtain one group of metric data value;
Concrete, in the present embodiment, the figure of sampling on wafer can comprise some one dimension figures and some X-Y schemes simultaneously, these one dimension figures and X-Y scheme can comprise again several different feature pattern.One group of metric data value generally reaches hundreds of or several thousand data points.
Step S02: metric data value is processed, filtered out the metric data of incredible sampled point;
Concrete, in the present embodiment, owing to can have error in the measurement process to wafer, therefore employing is averaged after the several different repetitive position of same figure is repeatedly measured conventionally, and remove the metric data of the sampled point that in repeatedly measurement process, fluctuation is larger, and remove the metric data of the sampled point that is significantly less than design rule size.
This be because: in the present invention, incredible sampled point can be, but not limited to point, the point that extraction model had little significance larger for metric data fluctuation or the sampled point obviously departing from homogeneous data, for example, for a certain sampled point, metric data fluctuation through repeatedly sampling is larger, thinks that such sampled point is insincere; Again for example, after a certain sampled point is sampled, the metric data that is significantly less than design rule size obtaining has little significance to extraction model, and such sampled point is thought incredible too; Again for example, it is not abnormal that certain data point seems itself, but while putting together with homogeneous data, its trend obviously deviates from homogeneous data, and it is insincere that such sampled point is also thought, etc.Therefore, in the present invention, the metric data of incredible sampled point need to be filtered out.
Step S03: the metric data value based on processing is set up initial OPC model, calculates the original residual error of the model data value of each sampled point;
Concrete, in the present embodiment, adopt the above-mentioned metric data that filters out insincere sampled point to set up an initial OPC model, then calculate each model data value, again each model data value and metric data value are compared, obtain the original residual error of the model data value of each sampled point.Such as, measCD1, measCD2 ... measCDn is respectively the 1st sampled point, the 2nd sampled point ... the metric data value of n sampled point, modelCD1, modelCD2 ... modelCDn is respectively first sampled point, the 2nd sampled point ... the model data value of n sampled point, the original residual error MRE1 of the model data value of each sampled point, MRE2,, MREn is respectively:
MRE1=modelCD1-measCD1
MRE2=modelCD2-measCD2
…
MREn=modelCDn-measCDn
Step S04: utilize the square value of original residual error or a cube value to obtain revised residual error root mean square;
Concrete, in the present embodiment, can utilize the square value of original residual error to obtain revised residual error root mean square time, adopt following formula: M=[(MRE1
2)
2+ (MRE
2)
2+ ... + (MREn
2)
2]
1/2, wherein, MRE1, MRE2 ..., MREn represent respectively the 1st sampled point, the 2nd sampled point ..., a n sampled point original residual error, M represents revised residual error root mean square.In the present embodiment, can also utilize cube value of original residual error to obtain revised residual error root mean square time, adopt following formula: M=[(MRE1
3)
2+ (MRE
3)
2+ ... + (MREn
3)
2]
1/2, wherein, MRE1, MRE2 ..., MREn represent respectively the 1st sampled point, the 2nd sampled point ..., a n sampled point original residual error, M represents revised residual error root mean square.
Step S05: utilize revised residual error root mean square bootstrap algorithm to carry out the simulation of optical system, thereby obtain revised OPC model.
In the present embodiment, refer to Fig. 2-4, Fig. 2 is the distribution schematic diagram of the original residual error of the metric data of above-mentioned preferred embodiment of the present invention, Fig. 3 be above-mentioned preferred embodiment of the present invention metric data original residual error through square after the distribution schematic diagram of residual error, the original residual error of the metric data that Fig. 4 is above-mentioned preferred embodiment of the present invention through cube after the distribution schematic diagram of residual error.Horizontal ordinate in figure represents metric data point, and ordinate represents corresponding residual error value.The contrast situation of residual error before and after revising is: for the less sampled point of residual error, residual error through square or cube after, the contribution of overall M value is diminished, and, through cube after contribution be less than through square after contribution; For the larger sampled point of residual error, residual error through square or cube after, overall M value contribution is become to large, and, through cube after contribution be greater than through square after contribution, like this, can bootstrap algorithm in simulation process, strengthen weight to the larger sampled point of residual error through revised residual error root mean square, effectively reduce the residual error of whole OPC model, thereby improved the precision of OPC model.
In sum, the method of raising OPC model accuracy of the present invention, by residual error root mean square is revised, adopt original residual error square value or a cube value to calculate revised residual error root mean square, then utilize revised residual error root mean square bootstrap algorithm to carry out the simulation of optical system, in simulation process, effectively reduce the residual error of model entirety, thereby the precision of revised OPC model is improved.
Although the present invention discloses as above with preferred embodiment; right described embodiment only gives an example for convenience of explanation; not in order to limit the present invention; those skilled in the art can do some changes and retouching without departing from the spirit and scope of the present invention, and the protection domain that the present invention advocates should be as the criterion with described in claims.
Claims (6)
1. a method that improves optical proximity effect correction model precision, is characterized in that, comprises the following steps:
Step S01: sample on wafer, obtain one group of metric data value;
Step S02: described metric data value is processed, filtered out the metric data of incredible sampled point;
Step S03: the metric data value based on described processing is set up initial OPC model, calculates the original residual error of the model data value of each sampled point;
Step S04: utilize the square value of described original residual error or a cube value to obtain described revised residual error root mean square;
Step S05: utilize described revised residual error root mean square bootstrap algorithm to carry out the simulation of optical system, thereby obtain revised OPC model.
2. the method for raising optical proximity effect correction model precision according to claim 1, it is characterized in that, in described step S04, while utilizing the square value of described original residual error to obtain revised residual error root mean square, adopt following formula: M=[(MRE1
2)
2+ (MRE
2)
2+ ... + (MREn
2)
2]
1/2, wherein, MRE1, MRE2 ..., MREn represent respectively the 1st sampled point, the 2nd sampled point ..., a n sampled point described original residual error, M represents described revised residual error root mean square.
3. the method for raising optical proximity effect correction model precision according to claim 1, it is characterized in that, in described step S04, while utilizing cube value of described original residual error to obtain revised residual error root mean square, adopt following formula: M=[(MRE1
3)
2+ (MRE
3)
2+ ... + (MREn
3)
2]
1/2, wherein, MRE1, MRE2 ..., MREn represent respectively the 1st sampled point, the 2nd sampled point ..., a n sampled point described original residual error, M represents described revised residual error root mean square.
4. the method for raising optical proximity effect correction model precision according to claim 1, is characterized in that, described figure of sampling on wafer comprises some one dimension figures and some X-Y schemes simultaneously.
5. the method for raising optical proximity effect correction model precision according to claim 4, is characterized in that, described one dimension figure and described X-Y scheme comprise several different feature pattern.
6. the method for raising optical proximity effect correction model precision according to claim 1, it is characterized in that, described step S02 comprises: after the several different repetitive position of same figure is repeatedly measured, average, and remove the metric data of the sampled point that in repeatedly measurement process, fluctuation is larger, and remove the metric data of the sampled point that is significantly less than design rule size.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410076758.3A CN103777460A (en) | 2014-03-04 | 2014-03-04 | Method for improving precision of optical proximity effect correction model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410076758.3A CN103777460A (en) | 2014-03-04 | 2014-03-04 | Method for improving precision of optical proximity effect correction model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103777460A true CN103777460A (en) | 2014-05-07 |
Family
ID=50569871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410076758.3A Pending CN103777460A (en) | 2014-03-04 | 2014-03-04 | Method for improving precision of optical proximity effect correction model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103777460A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109459911A (en) * | 2018-12-25 | 2019-03-12 | 上海微阱电子科技有限公司 | A method of improving OPC model precision |
| CN111146104A (en) * | 2019-11-29 | 2020-05-12 | 上海集成电路研发中心有限公司 | Key size error analysis method |
| CN111158210A (en) * | 2020-03-10 | 2020-05-15 | 长江存储科技有限责任公司 | Optical proximity correction method for photomask, photomask and semiconductor manufacturing method |
| CN116071319A (en) * | 2023-01-28 | 2023-05-05 | 合肥新晶集成电路有限公司 | Model building method, device, computer equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030082463A1 (en) * | 2001-10-09 | 2003-05-01 | Thomas Laidig | Method of two dimensional feature model calibration and optimization |
| US20070074142A1 (en) * | 2005-09-27 | 2007-03-29 | Applied Materials, Inc. | Integrated circuit layout methods |
| CN101482697A (en) * | 2008-01-07 | 2009-07-15 | 中芯国际集成电路制造(上海)有限公司 | Method for reducing OPC model residual error |
| CN102103324A (en) * | 2009-12-17 | 2011-06-22 | 中芯国际集成电路制造(上海)有限公司 | Optical proximity effect correction method |
| CN102521670A (en) * | 2011-11-18 | 2012-06-27 | 中国电力科学研究院 | Power generation output power prediction method based on meteorological elements for photovoltaic power station |
| CN103472672A (en) * | 2012-06-06 | 2013-12-25 | 中芯国际集成电路制造(上海)有限公司 | Correction method of optical proximity correction model |
-
2014
- 2014-03-04 CN CN201410076758.3A patent/CN103777460A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030082463A1 (en) * | 2001-10-09 | 2003-05-01 | Thomas Laidig | Method of two dimensional feature model calibration and optimization |
| US20070074142A1 (en) * | 2005-09-27 | 2007-03-29 | Applied Materials, Inc. | Integrated circuit layout methods |
| CN101482697A (en) * | 2008-01-07 | 2009-07-15 | 中芯国际集成电路制造(上海)有限公司 | Method for reducing OPC model residual error |
| CN102103324A (en) * | 2009-12-17 | 2011-06-22 | 中芯国际集成电路制造(上海)有限公司 | Optical proximity effect correction method |
| CN102521670A (en) * | 2011-11-18 | 2012-06-27 | 中国电力科学研究院 | Power generation output power prediction method based on meteorological elements for photovoltaic power station |
| CN103472672A (en) * | 2012-06-06 | 2013-12-25 | 中芯国际集成电路制造(上海)有限公司 | Correction method of optical proximity correction model |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109459911A (en) * | 2018-12-25 | 2019-03-12 | 上海微阱电子科技有限公司 | A method of improving OPC model precision |
| CN111146104A (en) * | 2019-11-29 | 2020-05-12 | 上海集成电路研发中心有限公司 | Key size error analysis method |
| WO2021103602A1 (en) * | 2019-11-29 | 2021-06-03 | 上海集成电路研发中心有限公司 | Critical dimension error analysis method |
| CN111146104B (en) * | 2019-11-29 | 2023-09-05 | 上海集成电路研发中心有限公司 | Method for analyzing critical dimension error |
| CN111158210A (en) * | 2020-03-10 | 2020-05-15 | 长江存储科技有限责任公司 | Optical proximity correction method for photomask, photomask and semiconductor manufacturing method |
| CN116071319A (en) * | 2023-01-28 | 2023-05-05 | 合肥新晶集成电路有限公司 | Model building method, device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108828896B (en) | Method for adding sub-resolution auxiliary graph and application of method | |
| CN101349863B (en) | Method for correcting optical approach effect of polygon rim dynamic cutting by configuration sampling | |
| CN103309149B (en) | Optical proximity correction method | |
| CN106933028A (en) | The modification method of mask plate patterns | |
| JP2014524135A (en) | Flare calculation and compensation for EUV lithography | |
| CN105825036A (en) | Method and system for optimizing layout design rules | |
| JP2008026822A (en) | Method for manufacturing photomask and method for manufacturing semiconductor device | |
| CN103186030A (en) | Optical proximity correction method | |
| CN103777460A (en) | Method for improving precision of optical proximity effect correction model | |
| CN106469235A (en) | Integrated circuit method and IC design system | |
| CN106773544B (en) | A kind of OPC modeling methods for controlling secondary graphics signal rate of false alarm | |
| CN103163728A (en) | OPC correction method based on photoetching process window | |
| TW201923443A (en) | Method for mask making | |
| CN109840342A (en) | The method executed by computing system | |
| CN103365071B (en) | The optical adjacent correction method of mask plate | |
| CN102117010B (en) | Optical adjacent correcting method | |
| CN104749896A (en) | Optical proximity correction method | |
| CN105334694A (en) | Prediction and improvement method of photoresist side wall angle | |
| CN109459911B (en) | Method for improving OPC model precision | |
| CN105093809A (en) | Optical proximity correction method for enhancing lithography process window | |
| CN104765246B (en) | The method of integrated targeted graphical optimization and optical proximity correction | |
| CN103163727A (en) | Mask pattern correction method | |
| CN114326288A (en) | Method for enlarging photoetching process window, electronic equipment and storage medium | |
| CN103744265B (en) | Improve the optical proximity correction method of process window | |
| CN101482697B (en) | Method for reducing OPC model residual error |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140507 |
|
| WD01 | Invention patent application deemed withdrawn after publication |