Simplified Micro-Bridging and Roughness Analysis

• US 20110239167A1
• Filed: 03/26/2010
• Published: 09/29/2011
• Est. Priority Date: 03/26/2010
• Status: Active Grant

• Current Assignee: Tokyo Electron Limited
• Original Assignee: Tokyo Electron Limited
• Inventors: Rathsack, Benjamen
• Granted Patent: US 8,108,805 B2
• Time in Patent Office: Days
• US Class Current: 716/50
• CPC Class Codes: G03F 7/705 Modelling or simulating from physical phenomena up to complete wafer processes or whole workflow in wafer productions
• CPC Class Codes: G03F 7/70625 Dimensions, e.g. line width, critical dimension [CD], profile, sidewall angle or edge roughness
• CPC Class Codes: H01L 22/20 Sequence of activities consisting of a plurality of measurements, corrections, marking or sorting steps
• CPC Class Codes: H01L 2924/00 Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
• CPC Class Codes: H01L 2924/0002 Not covered by any one of groups H01L24/00, H01L24/00 and H01L2224/00

Abstract

The invention provides apparatus and methods for processing substrates using pooled statistically based variance data. The statistically based variance data can include Pooled Polymer De-protection Variance (PPDV) data that can be used to determine micro-bridging defect data, LER defect data, and LWR defect data.

Claims

1. A method for processing a substrate using statistical variance data comprising:

establishing a first defect type for analysis on a first patterned substrate selected from a group of patterned substrates, wherein the first defect type includes a micro-bridging defect, or a line edge roughness (LER) defect, or a line width roughness (LWR) defect;

obtaining defect analysis data from a plurality of analysis segments on the first patterned substrate, wherein the defect analysis data includes bright-field data, dark-field data, NLS data, intensity data, intensity difference data, contrast data, or hot-spot data, or any combination thereof;

identifying a first potentially-defective analysis segment at a first location on the first patterned substrate using the defect analysis data, wherein the first potentially-defective analysis segment defines a three-dimensional volume that includes the first defect type;

calculating first Pooled Polymer De-protection Variance [(PPDV, σ

_m)]_Calc data for the first potentially-defective analysis segment, wherein (m) is a simulation parameter and (σ

_m) is a standard deviation of the simulation parameter (m);

obtaining first estimated [(PPDV, σ

_m)]_Est data in real-time from a first Pooled Polymer De-protection Variance (PPDV, σ

_m) data library to compare with the first calculated [(PPDV, σ

_m)]_Calc data, wherein the [(PPDV, σ

_m)]_Est data comprises analysis variable data (AVD), average value (<

AVD>

) data, simulation data, photoresist model data, integrated circuit (IC) data, design rule data, processing sequence data, substrate data, evaluation zone data, or defect type data, or any combination thereof;

determining a first difference value using the first calculated [(PPDV, σ

_m)]_Calc data and the first estimated [(PPDV, σ

_m)]_Est data;

identifying a first defect in the first potentially-defective analysis segment on the first patterned substrate using at least one defect type associated with the first estimated [(PPDV, σ

_m)]_Est data, when the first difference value is less than or equal to a first accuracy limit; and

continuing to analyze the first patterned substrate, when the first difference value is greater than the first accuracy limit.

2. The method of claim 1, wherein calculating the first calculated [(PPDV, σ

_m)]_Calc data comprises;

establishing a first modified photoresist model for the first potentially-defective analysis segment on the first patterned substrate, wherein the first modified photoresist model includes a first simulation parameter (m₁), wherein the first simulation parameter (m₁) includes blocked polymer concentration data, de-blocked polymer concentration data, polymer de-protection concentration data, protected polymer concentration data, acid diffusion data, base diffusion data, NILS data, blocked chromophore data, or de-blocked chromophore data, or any combination thereof;

creating a first simulation data map using the first modified photoresist model and the first simulation parameter (m₁) for the first potentially-defective analysis segment on the first patterned substrate, wherein the first simulation data map includes a plurality of contours associated with different simulated data values;

establishing a first evaluation zone to associate with the first simulation data map, the first evaluation zone being configured to include a first set of the contours;

calculating first three-dimensional analysis variable data (AVD_vol)₁ using the first evaluation zone; and

calculating the first calculated [(PPDV, σ

_m)]_Calc data using the first three-dimensional analysis variable data (AVD_vol)₁.

3. The method of claim 1, wherein continuing to analyze the first patterned substrate further comprises:

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real-time from the first (PPDV, σ

_m) data library to compare with the first calculated [(PPDV, σ

_m)]_Calc data, wherein the new estimated [(PPDV, σ

_m)]_NewEst data comprises new analysis variable data (AVD)_New, new average value (<

AVD>

)_New data, new simulation data, new photoresist model data, new IC data, new design rule data, new processing sequence data, new substrate data, new evaluation zone data, or new defect type data, or any combination thereof;

determining a new difference value using the first calculated [(PPDV, σ

_m)]_Calc data and the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data;

identifying a new defect in the first potentially-defective analysis segment on the first patterned substrate using at least one new defect type associated with the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data, when the new difference value is less than or equal to a new accuracy limit; and

continuing to analyze the first patterned substrate, when the new difference value is greater than the new accuracy limit.

4. The method of claim 1, wherein continuing to analyze the first patterned substrate further comprises:

calculating new Pooled Polymer De-protection Variance [(PPDV, σ

_m)]<sub>New—

Calc</sub> data for the first potentially-defective analysis segment;

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real time from the first (PPDV, σ

_m) library to compare with the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data, wherein the new estimated [(PPDV, σ

_m)]_NewEst data comprises new analysis variable data (AVD)_New, new average value (<

AVD>

)_New data, new simulation data, new photoresist model data, new IC data, new design rule data, new processing sequence data, new substrate data, new evaluation zone data, or new defect type data, or any combination thereof;

determining a new difference value using the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data and the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data;

identifying a new defect in the first potentially-defective analysis segment on the first patterned substrate using at least one new defect type associated with the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data, when the new difference value is less than or equal to a new accuracy limit; and

continuing to analyze the first patterned substrate, when the new difference value is greater than the new accuracy limit.

5. The method of claim 4, wherein calculating the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data comprises;

establishing a new calibrated photoresist model for the first potentially-defective analysis segment on the first patterned substrate, wherein the new calibrated photoresist model includes a new simulation parameter (m)_New, wherein the new simulation parameter (m)_New includes new blocked polymer concentration data, new de-blocked polymer concentration data, new polymer de-protection concentration data, new protected polymer concentration data, new acid diffusion data, new base diffusion data, new NILS data, blocked chromophore data, or new de-blocked chromophore data, or any combination thereof;

establishing at least one new evaluation zone to associate with a new simulation data map, the new evaluation zone being configured to include a new set of new contour lines;

calculating new three-dimensional analysis variable data (AVD_vol)_New using the new set of new contour lines in the new evaluation zone; and

calculating the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data using the new three-dimensional analysis variable data (AVD_vol)_New.

6. The method of claim 1, wherein continuing to analyze the first patterned substrate further comprises:

identifying a new potentially-defective analysis segment at a new location on the first patterned substrate using the defect analysis data, wherein the new potentially-defective analysis segment defines a new three-dimensional volume that includes a new defect type;

calculating new Pooled Polymer De-protection Variance [(PPDV, σ

_m)]<sub>New—

Calc</sub> data for the new potentially-defective analysis segment;

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real-time from the first (PPDV, σ

_m) library to compare with the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data, wherein the new estimated [(PPDV, σ

_m)]_NewEst data comprises new analysis variable data (AVD)_New, new average value (<

AVD>

)_New data, new simulation data, new photoresist model data, new IC data, new design rule data, new processing sequence data, new substrate data, new evaluation zone data, or new defect type data, or any combination thereof;

determining a new difference value using the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data and the new estimated [(PPDV, σ

_m)]_NewEst data;

identifying a new defect in the new potentially-defective analysis segment on the first patterned substrate using at least one new defect type associated with the new estimated [(PPDV, σ

_m)]_NewEst data, when the new difference value is less than or equal to a new accuracy limit; and

continuing to analyze the first patterned substrate, when the new difference value is greater than the new accuracy limit.

7. The method of claim 1, wherein continuing to analyze the first patterned substrate further comprises:

obtaining new defect analysis data from the first patterned substrate, wherein the new defect analysis data includes new bright-field data, new dark-field data, new NLS data, new intensity data, new intensity difference data, new contrast data, or new hot-spot data, or any combination thereof;

identifying a new potentially-defective analysis segment at a new location on the first patterned substrate using the new defect analysis data, wherein the new potentially-defective analysis segment defines a new three-dimensional volume that includes a new defect type;

calculating new [(PPDV, σ

_m)]<sub>New—

Calc</sub> data for the new potentially-defective analysis segment;

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real-time from the first (PPDV, σ

_m) library to compare with the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data, wherein the new estimated [(PPDV, σ

_m)]_NewEst data comprises new analysis variable data (AVD)_New, new average value (<

AVD>

)_New data, new simulation data, new photoresist model data, new IC data, new design rule data, new processing sequence data, new substrate data, new evaluation zone data, or new defect type data, or any combination thereof;

determining a new difference value using the new calculated [(PPDV, σ

_m)]_Calc data and the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data;

identifying a new defect in the new potentially-defective analysis segment on the first patterned substrate using at least one new defect type associated with the new estimated [(PPDV, σ

_m]<sub>New—

Est</sub> data, when the new difference value is less than or equal to a new accuracy limit; and

continuing to analyze the first patterned substrate, when the new difference value is greater than the new accuracy limit.

8. The method of claim 1, wherein continuing to analyze the first patterned substrate further comprises:

establishing a new defect type for analysis on the first patterned substrate, wherein the new defect type includes a new micro-bridging defect, a new LER defect, or a new LWR defect, or any combination thereof;

obtaining new defect analysis data from the first patterned substrate, wherein the new defect analysis data includes new bright-field data, new dark-field data, new NLS data, new intensity data, new intensity difference data, new contrast data, or new hot-spot data, or any combination thereof;

identifying a new potentially-defective analysis segment at a new location on the first patterned substrate using the new defect analysis data, wherein the new potentially-defective analysis segment defines a new three-dimensional volume that includes the new defect type;

calculating new [(PPDV, σ

_m)]<sub>New—

Calc</sub> data for the new potentially-defective analysis segment;

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real-time from the first (PPDV, σ

_m) library to compare with the new calculated [(PPDV, σ

_m)]<sub>New—

Calc</sub> data, wherein the new estimated [(PPDV, σ

_m)]_NewEst data comprises new analysis variable data (AVD)_New, new average value (<

AVD>

)_New data, new simulation data, new photoresist model data, new IC data, new design rule data, new processing sequence data, new substrate data, new evaluation zone data, or new defect type data, or any combination thereof;

determining a new difference value using the new calculated [(PPDV, σ

_m)]_Calc data and the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data;

identifying a new defect in the new potentially-defective analysis segment on the first patterned substrate using a first new defect type associated with the new estimated [(PPDV, σ

_m)] New<sub>—

Est</sub> data, when the new difference value is less than or equal to a new accuracy limit; and

continuing to analyze the first patterned substrate, when the new difference value is greater than the new accuracy limit.

9. The method of claim 1, wherein calculating the first calculated [(PPDV, σ

_m)]_Calc data comprises;

determining at least one probability value to associate with the first potentially-defective analysis segment and the first patterned substrate.

10. The method of claim 1, wherein the patterned substrates are created using a first set of design rules.

11. A method for verifying a set of design rules using pooled variance data comprising:

establishing a first defect type for analysis on a patterned substrate created using a first set of design rules, wherein the first defect type is associated with the first set of design rules and includes a micro-bridging defect, or a line edge roughness (LER) defect, or a line width roughness (LWR) defect;

obtaining defect analysis data from the patterned substrate, wherein the defect analysis data includes bright-field data, dark-field data, NLS data, intensity data, intensity difference data, contrast data, or hot-spot data, or any combination thereof;

identifying a first potentially-defective analysis segment at a first location on the patterned substrate using the defect analysis data, wherein the first potentially-defective analysis segment defines a three-dimensional volume that includes the first defect type;

calculating first Pooled Polymer De-protection Variance [(PPDV, σ

_m)]_Calc data for the first potentially-defective analysis segment, wherein (m) is a simulation parameter and (σ

_m) is a standard deviation of the simulation parameter (m);

obtaining first estimated [(PPDV, σ

_m)]_Est data in real-time from at least one Pooled Polymer De-protection Variance (PPDV, σ

_m) data library to compare with the first calculated [(PPDV, σ

_m)]_Calc data, wherein the first [(PPDV, σ

_m)]_Est data comprises historical data that includes a first set of defective design rules, at least one first defect type associated with the first set of defective design rules, and a first defective integrated circuit (IC) pattern associated with the first set of defective design rules;

identifying the first set of design rules as defective design rules using the first set of defective design rules, when the first estimated [(PPDV, σ

_m)]_Est data is substantially equal to the first calculated [(PPDV, σ

_m)]_Calc data; and

continuing to analyze the patterned substrate, when the first estimated [(PPDV, σ

_m)]_Est data is not equal to the first calculated [(PPDV, σ

_m)]_Calc data.

12. The method of claim 11, wherein calculating the first calculated [(PPDV, σ

_m)]_Calc data comprises;

establishing a first modified photoresist model for the first potentially-defective analysis segment on the patterned substrate, wherein the first modified photoresist model includes a first simulation parameter (m₁), wherein the first simulation parameter (m₁) includes blocked polymer concentration data, de-blocked polymer concentration data, polymer de-protection concentration data, protected polymer concentration data, acid diffusion data, base diffusion data, NILS data, blocked chromophore data, or de-blocked chromophore data, or any combination thereof;

creating a first simulation data map using the first modified photoresist model and the first simulation parameter (m₁) for the first potentially-defective analysis segment on the patterned substrate, wherein the first simulation data map includes a plurality of contour lines associated with different simulated data values;

establishing a first evaluation zone to associate with the first simulation data map, the first evaluation zone being configured to include a first set of the contours;

calculating first three-dimensional analysis variable data (AVD_vol)₁ using the first evaluation zone; and

calculating the first calculated [(PPDV, σ

_m)]_Calc data using the first three-dimensional analysis variable data (AVD_vol)₁.

13. The method of claim 11, wherein continuing to analyze the patterned substrate further comprises:

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real-time from the Pooled Polymer De-protection Variance (PPDV, σ

_m) data library to compare with the first calculated [(PPDV, σ

_m)]_Calc data, wherein the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data comprises new historical data that includes a new set of defective design rules, a new defect type associated with the new set of defective design rules, and a first defective integrated circuit (IC) pattern associated with the first set of defective design rules;

identifying the first set of design rules as defective design rules using the new set of defective design rules, when the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data is substantially equal to the first calculated [(PPDV, σ

_m)]_Calc data; and

continuing to analyze the patterned substrate, when the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data is not equal to the first calculated [(PPDV, σ

_m)]_Calc data.

14. The method of claim 11, wherein continuing to analyze the patterned substrate further comprises:

calculating new Pooled Polymer De-protection Variance [(PPDV, σ

_m)]<sub>New—

Calc</sub> data for the first potentially-defective analysis segment;

identifying the first set of design rules as defective design rules using a new set of defective design rules, when the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data is substantially equal to the first calculated [(PPDV, σ

_m)]_Calc data; and

continuing to analyze the patterned substrate, when the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data is not equal to the first calculated [(PPDV, σ

_m)]_Calc data.

15. The method of claim 11, wherein continuing to analyze the patterned substrate further comprises:

calculating new [(PPDV, σ

_m)]<sub>New—

Calc</sub> data for the first potentially-defective analysis segment;

obtaining new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data in real-time from the Pooled Polymer De-protection Variance (PPDV, σ

_m) data library to compare with the first calculated [(PPDV, σ

_m)]_Calc data, wherein the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data comprises new historical data that includes a new set of defective design rules, a new defect type associated with the new set of defective design rules, and a first defective integrated circuit (IC) pattern associated with the first set of defective design rules;

identifying the first set of design rules as defective design rules using the new set of defective design rules, when the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data is substantially equal to the new [(PPDV, σ

_m)]<sub>New—

Calc</sub> data; and

continuing to analyze the patterned substrate, when the new estimated [(PPDV, σ

_m)]<sub>New—

Est</sub> data is not equal to the new [(PPDV, σ

_m)]<sub>New—

calc</sub> data.

16. A method for creating pooled variance data comprising:

a) selecting an (i^th) potentially defective Integrated Circuit (IC) pattern from a group of (I) different potentially defective IC patterns, wherein i=1, 2, . . . I, and (I) is an integer greater than one;

b) selecting a (j_i^th) set of process parameters from a group of (J_i) different sets of process parameters, wherein i=1, 2, . . . I, (j=1, 2, . . . J_i), and (J_i) is an integer greater than one;

c) selecting a k_ij^th data-collection substrate (S_ijk) from a group of (K_ij) data-collection substrates, wherein (k_ij=1, 2, . . . K_ij), (i=1, 2, . . . I), (j=1, 2, . . . J_i) and (K_ij) is an integer greater than one;

d) processing the k_ij^th data-collection substrate (S_ijk) using the j_i^th set of process parameters;

e) obtaining at least one optical image from the k_ij^th data-collection substrate (S_ijk);

f) selecting an [l^th]_{i jk} analysis segment from a group of (L_ijk) analysis segments on the k_ij^th data-collection substrate (S_ijk), wherein (l_ijk=1, 2, . . . L_ijk) (i=1, 2, . . . I), (j=1, 2, . . . J_i), (k=1, 2, . . . K_ij), and (L_ijk) is an integer greater than one, each processed data-collection substrates (S_ijk) being divided into (L_ijk) analysis segments, wherein each intensity difference spot in the at least one optical image is used to establish at least one of the group of (L_ijk) analysis segments on each processed data-collection substrate (S_jk);

g) selecting an [m^th]_ijkl simulation parameter from a group of (M_ijkl) simulation parameters for each of the selected [l^th]_{i jk} analysis segment on each of the processed data-collection substrates (S_ijk), wherein (m_ijkl=1, 2, . . . M_ijkl), (i=1, 2, . . . I), (j=1, 2, . . . J_i), (k=1, 2, . . . K_ij), (l_ijk=1, 2, . . . L_ijk), and (M_ijkl) is an integer greater than one;

h) creating an [m^th]_ijkl simulation data map using the selected [m^th]_ijkl simulation parameter and a modified photoresist model;

i) selecting an [n^th]_ijklm evaluation zone in the [m^th]_ijkl simulation data map, wherein (i=1, 2, . . . I), (j=1, 2, . . . J_i), (k=1, 2, . . . K_ij), (l_ijk=1, 2, . . . L_ijk) (m_ijkl=1, 2, . . . M_ijkl), and ([N]_ijklm) is an integer greater than one;

j) calculating at least one Analysis Variable Data ([AVD_n^th]_ijklm) value, wherein (i=1, 2, . . . I), (j=1, 2, . . . J_i), (k=1, 2, . . . K_ij), (l_ijk=1, 2, . . . L_ijk) (m_ijkl=1, 2, . . . M_ijkl), and ([n^th]_ijklm=1, 2, . . . [N]_ijklm);

k) determining if a data-collection procedure has been completed for the ([k^th]_ij) processed data-collection substrate (S_ijk);

l) obtaining additional data for the ([k^th]_ij) processed data-collection substrate (S_ijk), when the data-collection procedure has not been completed; and

m) calculating Pooled Polymer De-protection Variance (PPDV, σ

_m) data for the ([k^th]_ij) processed data-collection substrate (S_ijk), when the data-collection procedure has been completed.

17. The method of claim 16, wherein obtaining the additional data for the ([k^th]_ij) processed data-collection substrate (S_ijk) further comprises:

determining if a new evaluation zone is required for the ([k^th]_ij) processed data-collection substrate (S_ijk);

selecting an [(n+1)^th]_ijklm evaluation zone; and

repeating steps i)-m).

18. The method of claim 16, wherein obtaining the additional data for the ([k^th]_ij) processed data-collection substrate (S_ijk) further comprises:

determining if a new simulation parameter is required for the ([k^th]_ij) processed data-collection substrate (S_ijk);

selecting an [(m+1)^th]_ijkl simulation parameter; and

repeating steps g)-m).

19. The method of claim 16, wherein obtaining the additional data for the ([k^th]_ij) processed data-collection substrate (S_ijk) further comprises:

determining if a new analysis segment is required for the ([k^th]_ij) processed data-collection substrate (S_ijk);

selecting an [(l+1)^th]_ijk analysis segment; and

repeating steps f)-m).

20. The method of claim 16, wherein the modified photoresist model includes at least one methacrylate polymer and at least one Triphenyl Sulfonium (TPS)-nonaflate Photo Acid Generator (PAG).