Using Multi-Layer/Multi-Input/Multi-Output (MLMIMO) Models for Metal-Gate Structures
First Claim
1. A method for using a Multi-Layer/Multi-Input/Multi-Output (MLMIMO) model comprising:
- receiving a first set of patterned wafers and associated wafer data, each patterned wafer having a first patterned soft-mask layer and a plurality of additional layers, the first patterned soft-mask layer including a plurality of metal-gate-related soft-mask features and at least one first periodic evaluation structure, the wafer data including real-time integrated metrology (IM) data for the at least one first periodic evaluation structure in the first patterned soft-mask layer;
establishing a first Multi-Layer-Multi-Step (MLMS) processing sequence, wherein the first MLMS processing sequence comprises a first set of Poly-Etch procedures and is configured to establish a first gate-width control pattern in a first set of the additional layers using the first patterned soft-mask layer;
creating a second set of patterned wafers using the first MLMS processing sequence;
creating first simulation data for the first MLMS processing sequence using a first Multi-Layer/Multi-Input/Multi-Output (MLMIMO) model for the first MLMS processing sequence, wherein the first MLMIMO model includes a first number (Na) of first Controlled Variables (CV1a, CV2a, . . . CVNa), a first number (Ma) of first Manipulated Variables (MV1a, MV2a, . . . MVMa), and a first number (La) of first Disturbance Variables (DV1a, DV2a, . . . DVLa), wherein (La, Ma, and Na) are integers greater than one;
establishing a second MLMS processing sequence, wherein the second MLMS processing sequence is configured to create a first controlled pattern of metal-gate structures by patterning a second set of the additional layers using the first gate-width control pattern;
creating a third set of patterned wafers using the second MLMS processing sequence;
creating second simulation data for the second MLMS processing sequence using a second MLMIMO model for the second MLMS processing sequence, wherein the second MLMIMO model includes a second number (Nb) of second Controlled Variables (CV1b, CV2b, . . . CVNb), a second number (Mb) of second Manipulated Variables (MV1b, MV2b, . . . MVMb), and a second number (Lb) of second Disturbance Variables (DV1b, DV2b, . . . DVLb), wherein (Lb, Mb, and Nb) are integers greater than one.obtaining evaluation data for at least one of the third set of patterned wafers;
identifying the third set of patterned wafers as verified wafers when the evaluation data is less than a first metal-gate limit; and
performing a corrective action when the evaluation data is not less than the first metal-gate limit.
1 Assignment
0 Petitions
Accused Products
Abstract
The invention provides a method of processing a wafer using multilayer processing sequences and Multi-Layer/Multi-Input/Multi-Output (MLMIMO) models and libraries that can include one or more measurement procedures, one or more Poly-Etch (P-E) sequences, and one or more metal-gate etch sequences. The MLMIMO process control uses dynamically interacting behavioral modeling between multiple layers and/or multiple process steps. The multiple layers and/or the multiple process steps can be associated with the creation of lines, trenches, vias, spacers, contacts, and gate structures that can be created using isotropic and/or anisotropic etch processes.
-
Citations
32 Claims
-
1. A method for using a Multi-Layer/Multi-Input/Multi-Output (MLMIMO) model comprising:
-
receiving a first set of patterned wafers and associated wafer data, each patterned wafer having a first patterned soft-mask layer and a plurality of additional layers, the first patterned soft-mask layer including a plurality of metal-gate-related soft-mask features and at least one first periodic evaluation structure, the wafer data including real-time integrated metrology (IM) data for the at least one first periodic evaluation structure in the first patterned soft-mask layer; establishing a first Multi-Layer-Multi-Step (MLMS) processing sequence, wherein the first MLMS processing sequence comprises a first set of Poly-Etch procedures and is configured to establish a first gate-width control pattern in a first set of the additional layers using the first patterned soft-mask layer; creating a second set of patterned wafers using the first MLMS processing sequence; creating first simulation data for the first MLMS processing sequence using a first Multi-Layer/Multi-Input/Multi-Output (MLMIMO) model for the first MLMS processing sequence, wherein the first MLMIMO model includes a first number (Na) of first Controlled Variables (CV1a, CV2a, . . . CVNa), a first number (Ma) of first Manipulated Variables (MV1a, MV2a, . . . MVMa), and a first number (La) of first Disturbance Variables (DV1a, DV2a, . . . DVLa), wherein (La, Ma, and Na) are integers greater than one; establishing a second MLMS processing sequence, wherein the second MLMS processing sequence is configured to create a first controlled pattern of metal-gate structures by patterning a second set of the additional layers using the first gate-width control pattern; creating a third set of patterned wafers using the second MLMS processing sequence; creating second simulation data for the second MLMS processing sequence using a second MLMIMO model for the second MLMS processing sequence, wherein the second MLMIMO model includes a second number (Nb) of second Controlled Variables (CV1b, CV2b, . . . CVNb), a second number (Mb) of second Manipulated Variables (MV1b, MV2b, . . . MVMb), and a second number (Lb) of second Disturbance Variables (DV1b, DV2b, . . . DVLb), wherein (Lb, Mb, and Nb) are integers greater than one. obtaining evaluation data for at least one of the third set of patterned wafers; identifying the third set of patterned wafers as verified wafers when the evaluation data is less than a first metal-gate limit; and performing a corrective action when the evaluation data is not less than the first metal-gate limit. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A method for using a Multi-Layer/Multi-Input/Multi-Output (MLMIMO) model to create metal-gate structures on a plurality of wafers, the method comprising:
-
a) receiving a first set of send-ahead wafers and associated wafer data, the wafer data including real-time and historical data; b) establishing a first number (la) of disturbance variables DV(La) for a first multi-layer etch sequence and a second number (lb) of disturbance variables DV(Lb) for a second multi-layer etch sequence using real-time integrated metrology (IM) data associated with a patterned photoresist layer on one or more of the send-ahead wafers, wherein the real-time IM data includes critical dimension (CD) data, sidewall angle (SWA) data, thickness data, photoresist data, BARC data, the wafer data, and diffraction signal data from multiple sites in the patterned photoresist layer on each incoming wafer, wherein (La) and (Lb) are integers greater than two; c) establishing a first number (Ma) of manipulated variables MV(Ma) for the first multi-layer etch sequence and a second number (Mb) of manipulated variables MV(Mb) for a second multi-layer etch sequence, wherein (Ma) and (Mb) are integers greater than two; d) establishing a first number (Na) of controlled variables for the first multi-layer etch sequence and a second number (Nb) of controlled variables for the second multi-layer etch sequence, wherein (Na) and (Nb) are integer greater than two, wherein CV(Na) is defined as
CV(Na)=fNa{MV(1a) . . . MV(Ma-1), MV(Ma), DV(1), . . . DV(La-1), DV(La)}+OFFSETSNaand CV(Nb) is defined as
CV(Nb)=fNb{MV(1b) . . . MV(Mb-1), MV(Mb), DV(1), . . . DV(Lb-1), DV(Lb)}+OFFSETSNbe) calculating optimized process settings using a first quadratic objective function, wherein first target deviations t(Na) for the first multi-layer etch sequence are defined as;
t(Na)={DV(Na}−
target CV(Na)};f) calculating optimized process settings using a second quadratic objective function, wherein second target deviations t(Nb) for the second multi-layer etch sequence are defined as;
t(Nb)={DV(Nb}−
target CV(Nb)};g) defining adjusted process recipes for the first multi-layer etch sequence and/or the second multi-layer etch sequence using one or more calculated manipulated variables established during nonlinear programming; h) processing one or more of the first set of send-ahead wafers using the adjusted process recipes; i) obtaining additional measurement data for one or more of the send-ahead wafers, wherein new controlled variable data are obtained and filtered; j) calculating one or more process errors using differences between the new controlled variable data and predicted controlled variable data; k) calculating feedback data items, wherein errors are used to update the OFFSETSNa for the first multi-layer etch sequence and/or the OFFSETSNb for the second multi-layer etch sequence using an exponentially weighted moving average (EWMA) filter; l) updating the OFFSETSNa for the first multi-layer etch sequence and/or the OFFSETSNb for the second multi-layer etch sequence in an optimizer unit; and m) repeating steps a)-l) using each wafer in the first set of send-ahead wafers. - View Dependent Claims (28, 29, 30, 31, 32)
-
Specification