公开(公告)号：US09412673B2

公开(公告)日：2016-08-09

申请号：US14459516

申请日：2014-08-14

Applicant: KLA-Tencor Corporation

Inventor： In-Kyo Kim ,  Xin Li ,  Leonid Poslavsky ,  Liequan Lee ,  Meng Cao ,  Sungchul Yoo ,  Andrei V. Shchegrov ,  Sangbong Park

IPC: G06F17/50 ,  H01L21/66 ,  G03F7/20

CPC classification number: H01L22/20 ,  G03F7/70616 ,  G03F7/70625 ,  G06F17/5068 ,  H01L21/67288 ,  H01L22/12

Abstract: Disclosed are apparatus and methods for characterizing a plurality of structures of interest on a semiconductor wafer. A plurality of models having varying combinations of floating and fixed critical parameters and corresponding simulated spectra is generated. Each model is generated to determine one or more critical parameters for unknown structures based on spectra collected from such unknown structures. It is determined which one of the models best correlates with each critical parameter based on reference data that includes a plurality of known values for each of a plurality of critical parameters and corresponding known spectra. For spectra obtained from an unknown structure using a metrology tool, different ones of the models are selected and used to determine different ones of the critical parameters of the unknown structure based on determining which one of the models best correlates with each critical parameter based on the reference data.

Abstract translation: 公开了用于在半导体晶片上表征感兴趣的多个结构的装置和方法。 产生具有浮动和固定关键参数和对应的模拟光谱的变化组合的多个模型。 生成每个模型以基于从这样的未知结构收集的光谱来确定未知结构的一个或多个关键参数。 基于包括多个关键参数和相应的已知光谱中的每一个的多个已知值的参考数据，确定哪一个模型与每个关键参数最佳相关。 对于使用计量工具从未知结构获得的光谱，选择不同的模型并且用于基于确定哪一个模型与每个关键参数最相关的来确定未知结构的关键参数的不同的参数 参考数据。

公开(公告)号：US20150058813A1

公开(公告)日：2015-02-26

申请号：US14459516

申请日：2014-08-14

Applicant: KLA-Tencor Corporation

Inventor： In-Kyo Kim ,  Xin Li ,  Leonid Poslavsky ,  Liequan Lee ,  Meng Cao ,  Sungchul Yoo ,  Andrei V. Shchegrov ,  Sangbong Park

IPC: G06F17/50

CPC classification number: H01L22/20 ,  G03F7/70616 ,  G03F7/70625 ,  G06F17/5068 ,  H01L21/67288 ,  H01L22/12

Abstract: Disclosed are apparatus and methods for characterizing a plurality of structures of interest on a semiconductor wafer. A plurality of models having varying combinations of floating and fixed critical parameters and corresponding simulated spectra is generated. Each model is generated to determine one or more critical parameters for unknown structures based on spectra collected from such unknown structures. It is determined which one of the models best correlates with each critical parameter based on reference data that includes a plurality of known values for each of a plurality of critical parameters and corresponding known spectra. For spectra obtained from an unknown structure using a metrology tool, different ones of the models are selected and used to determine different ones of the critical parameters of the unknown structure based on determining which one of the models best correlates with each critical parameter based on the reference data.

Abstract translation: 公开了用于在半导体晶片上表征感兴趣的多个结构的装置和方法。 产生具有浮动和固定关键参数和对应的模拟光谱的变化组合的多个模型。 生成每个模型以基于从这样的未知结构收集的光谱来确定未知结构的一个或多个关键参数。 基于包括多个关键参数和相应的已知光谱中的每一个的多个已知值的参考数据，确定哪一个模型与每个关键参数最佳相关。 对于使用计量工具从未知结构获得的光谱，选择不同的模型并且用于基于确定哪一个模型与每个关键参数最相关的来确定未知结构的关键参数的不同的参数 参考数据。

公开(公告)号：US10678226B1

公开(公告)日：2020-06-09

申请号：US15214371

申请日：2016-07-19

Applicant: KLA-Tencor Corporation

Inventor： Qiang Wang ,  Liequan Lee ,  Xin Li ,  Qiang Zhao

IPC: G05B19/418 ,  H01L21/67 ,  H01L21/66 ,  G01B11/00 ,  G01N21/956

Abstract: Systems and methods for providing efficient modeling and measurement of critical dimensions and/or overlay registrations of wafers are disclosed. Efficiency is improved in both spectral dimension and temporal dimension. In the spectral dimension, efficiency can be improved by allowing different numerical aperture (NA) models to be used for different wavelengths in electromagnetic calculations, effectively providing a balance between computation speed and accuracy. In the temporal dimension, different NA models may be used at different iterations/stages in the process, effectively improving the computation speed without sacrificing the quality of the final result.

公开(公告)号：US20160322267A1

公开(公告)日：2016-11-03

申请号：US15204461

申请日：2016-07-07

Applicant: KLA-Tencor Corporation

Inventor： In-Kyo Kim ,  Xin Li ,  Leonid Poslavsky ,  Liequan Lee ,  Meng Cao ,  Sungchul Yoo ,  Andrei V. Shchegrov ,  Sangbong Park

IPC: H01L21/66 ,  G06F17/50 ,  G03F7/20 ,  H01L21/67

CPC classification number: H01L22/20 ,  G03F7/70616 ,  G03F7/70625 ,  G06F17/5068 ,  H01L21/67288 ,  H01L22/12

Abstract: Disclosed are apparatus and methods for characterizing a plurality of structures of interest on a semiconductor wafer. A plurality of models having varying combinations of floating and fixed critical parameters and corresponding simulated spectra is generated. Each model is generated to determine one or more critical parameters for unknown structures based on spectra collected from such unknown structures. It is determined which one of the models best correlates with each critical parameter based on reference data that includes a plurality of known values for each of a plurality of critical parameters and corresponding known spectra. For spectra obtained from an unknown structure using a metrology tool, different ones of the models are selected and used to determine different ones of the critical parameters of the unknown structure based on determining which one of the models best correlates with each critical parameter based on the reference data.

公开(公告)号：US11537837B2

公开(公告)日：2022-12-27

申请号：US15883154

申请日：2018-01-30

Applicant: KLA-TENCOR CORPORATION

Inventor： Yuerui Chen ,  Xin Li

IPC: G06N3/04 ,  G06F17/16 ,  G06N3/08 ,  H01L21/66 ,  G03F7/20 ,  G01N21/47 ,  G01N21/21 ,  G01N21/95 ,  G06N20/10

Abstract: Techniques and systems for critical dimension metrology are disclosed. Critical parameters can be constrained with at least one floating parameter and one or more weight coefficients. A neural network is trained to use a model that includes a Jacobian matrix. During training, at least one of the weight coefficients is adjusted, a regression is performed on reference spectra, and a root-mean-square error between the critical parameters and the reference spectra is determined. The training may be repeated until the root-mean-square error is less than a convergence threshold.