Advanced Micro Devices Corporation

Globalfoundries Inc.

Globalfoundries U.S. Inc.

Richard P Good

James Broc Stirton

Eugene Barash

Kevin R Lensing

Brian K Cusson

Gregory A Cherry

Uwe Schulze

Jin Wang

Dirk Wollstein

J Broc Stirton

Matthias Ruhm

Timothy L Jackson

Kevin Andrew Chamness

Detlef Pabst

Daniel Kost

Houssam Lazkani

Roberto Schiwon

Daniel Zschaebitz

Lothar Waetzold

Eyup Cinar

Dirk Gleitsmann

Thomas Depaly

Ian R Krumanocker

Dinesh Balasubra Manian

The present disclosure generally relates to a system and a method for detecting a defect in a specimen. More particularly, the present disclosure relates to a lithography exposure system and a method for detecting a dispensing error in a wafer The present disclosure provides a system for detecting a defect in a specimen having a lithography exposure tool including a measurement unit and a stage, the measurement unit is configured to obtain topography data of the specimen placed on the stage by illumination of a surface of the specimen with an optical signal, and a processor configured to generate a statistical data from the topography data and produce a defect notification if the statistical data is outside of a control limit.

System and method for detecting a defect in a specimen

The present disclosure generally relates to semiconductor structures and, more particularly, to overlay optimization and methods of manufacture. The method includes performing, by a computing device, an exposure with a correction parameter to a first wafer; performing, by the computing device, a decorrection of the correction parameter; collecting, by the computing device, overlay data in response to the exposure and the decorrection; estimating, by the computing device, an optimal parameter from the overlay data; and applying, by the computing device, the optimal parameter to a second wafer to align an overlay in the second wafer.

Overlay optimization

A method for initializing individual exposure field parameters of an overlay controller is disclosed including initializing a first control thread having a first context associated with a first product type, wherein a first layout of first exposure fields is defined for the first product type for processing in a stepper. The method further includes remapping a set of previous control state data for a set of control threads associated with other product types different than the first product type into the first layout. The other product types have layouts of second exposure fields different than the first layout. An initial set of control state data for the first control thread associated with the first product type is generated using the remapped previous control state data. The stepper is configured for processing a first substrate of the first product type using the initial set of control state data.

Initializing individual exposure field parameters of an overlay controller

Methods of controlling an across-wafer profile of a semiconductor process, as well as related systems and computer program products. A target profile of a semiconductor process over a radius of a wafer is fit to a polynomial. A plurality of gain matrices between a first plurality of process inputs and a plurality of polynomial coefficients of the polynomial are determined. An offset is estimated between the plurality of polynomial coefficients and an effect of the first plurality of process inputs. An objective function is defined as an integral of a squared deviation between an estimated profile and the target profile over the radius of the wafer. A second plurality of process inputs are mapped to the objective function by vector convolution using the plurality of gain matrices. The objective function is solved to optimize the second plurality of process inputs.

Across-wafer profile control in semiconductor processes

Methods according to the disclosure include: identifying at least one non-candidate tool which previously processed at least one prior semiconductor wafer, and represented in a set of historical data, the set of historical data including manufacturing settings for performing an operation on the at least one prior semiconductor wafer, and a candidate tool which has not previously performed the operation on at least one prior semiconductor wafer, the candidate tool not being represented in the set of historical data; determining whether the set of historical data predicts the performing of the operation with the candidate tool based on the manufacturing settings for performing the operation on the at least one prior semiconductor wafer; and in response to the set of historical data predicting the performing of the operation with the candidate tool, selecting a manufacturing setting for the candidate tool based on the set of historical data.

Selecting manufacturing settings based on historical data from manufacturing tools

Methods and computer program products for performing automatically determining when to shut down a fabrication tool, such as a semiconductor wafer fabrication tool, are provided herein. The methods include, for example, creating a measurement vector including process parameters of semiconductor wafers, creating a correlation matrix of correlations between measurements of parameters obtained of each wafer, creating autocorrelation matrixes including correlations between measurements of the parameter obtained for pairs of wafers; creating a combined matrix of correlation and autocorrelation matrixes, obtaining a Tvalue from the measurement vector and combined matrix, and stopping a semiconductor wafer fabrication tool if the Tvalue exceeds a critical value.

Methods of error detection in fabrication processes

A method for estimating a state of a process implemented by a tool for fabricating workpieces includes collecting metrology data associated with a subset of workpieces processed in the tool. The collecting exhibits an irregular pattern. Metrology data associated with a selected state observation is received for a selected run of the process. A tuning factor for the selected run is determined based on the irregular pattern. The selected state observation is discounted based on the determined tuning factor. A state estimate of the process is determined based on the discounted selected state observation. At least one process tool operable to implement the process is controlled based on the state estimate.

Tuning a process controller based on a dynamic sampling rate

By performing a two-step approach for predicting a quality distribution during the fabrication of semiconductor devices, enhanced flexibility and efficiency may be accomplished. The two-step approach first models electrical characteristics on the basis of measurement data, such as inline measurement data, and, in a second step, an appropriate distribution for the electrical characteristics may be established, thereby obtaining modeled wafer sort data which may then be used for predicting a quality distribution of the semiconductor devices under consideration.

Method and system for a two-step prediction of a quality distribution of semiconductor devices

A method includes receiving measured values for a plurality of electrical test parameters associated with integrated circuit devices on at least one wafer measured prior to completion of the wafer. Values of the electrical test parameters are predicted. The measured values are compared to the predicted values to generate residual values associated with the electrical test parameters. At least one performance metric associated with the devices is generated based on the residual values.

Identifying manufacturing disturbances using preliminary electrical test data

A method includes acquiring metrology data associated with a process. Bias information associated with the process is determined. The metrology data is adjusted based on the bias information to generate bias-adjusted metrology data. The bias-adjusted metrology data is filtered to identify and reject outlier data. The process is controlled based on the metrology data remaining after the rejection of the outlier data.

Method and apparatus for compensating metrology data for site bias prior to filtering

A method includes defining a plurality of simple sampling rules for selecting material for metrology. Each simple sampling rule has an associated penalty. At least one combination sampling rule relating a subset of at least two simple sampling rules is defined. The combination sampling rule has an associated penalty. The penalties are assessed responsive to a previous material selection not satisfying the simple sampling rules or the combination sampling rule. Material is selected for subsequent metrology based on the sampling rules and the assessed penalties. At least one characteristic of the selected material is measured.

Method and apparatus for metrology sampling using combination sampling rules

A method for estimating a state associated with a process includes receiving a state observation associated with the process. The state observation has an associated process time. A weighting factor to discount the state observation is generated based on the process time. A state estimate is generated based on the discounted state observation. A system includes a process tool, a metrology tool, and a process controller. The process tool is operable to perform a process in accordance with an operating recipe. The metrology tool is operable to generate a state observation associated with the process. The process controller is operable to receive the state observation, the state observation having an associated process time, generate a weighting factor to discount the state observation based on the process time, generate a state estimate based on the discounted state observation, and determine at least one parameter of the operating recipe based on the state estimate.

Time weighted moving average filter

A method includes providing a plurality of sampling rules. Each sampling rule is associated with at least one of a plurality of sites on a workpiece. At least one penalty is assigned to each sampling rule. The penalty is assessed responsive to a previous site selection not satisfying the associated sampling rule. A subset of the sites is selected for subsequent metrology based on the sampling rules and the assessed penalties. At least one characteristic of the workpiece is measured at the selected subset of the sites.

Method and apparatus for selecting sites for sampling

A method includes processing a plurality of workpieces to form at least one feature on each workpiece. A plurality of characteristics of the feature is measured. A covariance matrix including diagonal and non-diagonal terms for the plurality of characteristics measured is constructed. At least the non-diagonal terms of the covariance matrix are monitored. A sampling plan for measuring the workpieces is determined based on the monitoring. A system includes a plurality of tools, at least one metrology tool, and a sampling controller. The tools are configured to process a plurality of workpieces to form at least one feature on each workpiece. The metrology tool is configured to measure a plurality of characteristics of the feature. The sampling controller is configured to construct a covariance matrix including diagonal and non-diagonal terms for the plurality of characteristics measured, monitor at least the non-diagonal terms of the covariance matrix, and determine a sampling plan for measuring the workpieces based on the monitoring.

Growing community of inventors

Austin, TX, United States of America

Richard P Good