PHOTORESIST DESIGN LAYOUT PATTERN PROXIMITY CORRECTION THROUGH FAST EDGE PLACEMENT ERROR PREDICTION VIA A PHYSICS-BASED ETCH PROFILE MODELING FRAMEWORK

• US 20170363950A1
• Filed: 06/21/2016
• Published: 12/21/2017
• Est. Priority Date: 06/21/2016
• Status: Active Grant

• Current Assignee: Lam Research Corporation
• Original Assignee: Lam Research Corporation
• Inventors: Sriraman, Saravanapriyan, Wise, Richard, Singh, Harmeet, Paterson, Alex, Bailey, III, Andrew D., Vahedi, Vahid, Gottscho, Richard A.
• Granted Patent: US 10,197,908 B2
• Time in Patent Office: Days
• CPC Class Codes: G03F 1/36 Masks having proximity correction features; Preparation thereof, e.g. optical proximity correction [OPC] design processes
• CPC Class Codes: G03F 1/70 Adapting basic layout or design of masks to lithographic process requirements, e.g., second iteration correction of mask patterns for imaging
• CPC Class Codes: G03F 1/80 Etching
• CPC Class Codes: G06F 2115/06 Structured ASICs
• CPC Class Codes: G06F 30/20 Design optimisation, verification or simulation optimisation, verification or simulation of circuit designs G06F30/30
• CPC Class Codes: G06F 30/392 Floor-planning or layout, e.g. partitioning or placement
• CPC Class Codes: G06F 30/398 Design verification or optimisation, e.g. using design rule check [DRC], layout versus schematics [LVS] or finite element methods [FEM] optical proximity correction [OPC] design processes G03F1/36

Abstract

Disclosed are methods of generating a proximity-corrected design layout for photoresist to be used in an etch operation. The methods may include identifying a feature in an initial design layout, and estimating one or more quantities characteristic of an in-feature plasma flux (IFPF) within the feature during the etch operation. The methods may further include estimating a quantity characteristic of an edge placement error (EPE) of the feature by comparing the one or more quantities characteristic of the IFPF to those in a look-up table (LUT, and/or through application of a multivariate model trained on the LUT, e.g., constructed through machine learning methods (MLM)) which associates values of the quantity characteristic of EPE with values of the one or more quantities characteristics of the IFPF. Thereafter, the initial design layout may be modified based on at the determined quantity characteristic of EPE.

Claims

1. A method of generating a proximity-corrected design layout for photoresist to be used in an etch operation, the method comprising:

(a) receiving an initial design layout;

(b) identifying a feature in the initial design layout, the feature'"'"'s pattern corresponding to a feature that would be etched into a material stack on a semiconductor substrate'"'"'s surface via a plasma-based etch process, performed in a processing chamber under a set of process conditions, when said stack is overlaid with a layer of photoresist pattern corresponding to the design layout;

(c) estimating one or more quantities characteristic of an in-feature plasma flux (IFPF) within the feature at a time t during such a plasma-based etch process;

(d) estimating a quantity characteristic of edge placement error (EPE) of the edge of the feature at time t by comparing the one or more quantities characteristic of the IFPF estimated in (c) to those in a look-up table (LUT) which associates values of the quantity characteristic of EPE at time t with values of the one or more quantities characteristics of the IFPF; and

(e) modifying the initial design layout based on at the quantity characteristic of EPE;

wherein the LUT was constructed by running a computerized etch profile model (EPM) under the set of process conditions at least to time t on a calibration pattern of photoresist overlaid on the material stack.

2. The method of claim 1, further comprising:

repeating (b) through (d) for one or more additional features whose patterns are in the initial design layout; and

wherein modifying the initial design in (e) is further based on the estimated quantity characteristic of EPE of these one or more additional features.

3. The method of claim 1, wherein in (c) the one or more quantities characteristic of the IFPF comprise:

a quantity characteristic of in-feature plasma ion flux (IFPIF); and

a quantity characteristic of in-feature plasma neutral flux (IFPNF).

4. The method of claim 3, wherein the quantity characteristic of IFPNF is a loaded plasma flux above the feature which accounts for the presence of the substrate in the processing chamber.

5. The method of claim 4, wherein the loaded plasma flux is estimated in (c) based on one or more quantities characteristic of far-field global plasma fluxes in the processing chamber.

6. The method of claim 5, wherein the one or more quantities characteristic of far-field global plasma fluxes are calculated with a computerized plasma model which accounts for processing chamber conditions but does not account for the presence of the substrate in the processing chamber.

7. The method of claim 3, wherein the quantity characteristic of the IFPIF is estimated in (c) based on a visibility kernel (VC) corresponding to the feature.

8. The method of claim 7, wherein the quantity characteristic of the IFPIF is calculated by a procedure comprising estimating the integral of the VC with the ion energy angular distribution function (IEADF) corresponding to one or more plasma ion fluxes (PIF) above the feature.

9. The method of claim 8, wherein the IEADF is estimated based on one or more quantities characteristic of far-field global plasma fluxes in the processing chamber which are calculated with a computerized plasma model which accounts for processing chamber conditions.

10. The method of claim 9, wherein the VC is estimated in (c) by assuming the feature has an opening corresponding to the initial design layout of photoresist and has substantially vertical sidewalls extending downward from the edges of the opening.

11. The method of claim 10, further comprising:

(c′

) re-estimating the VC of the feature based on the EPE estimated in (d);

(d′

) re-estimating the quantity characteristic of EPE at time t by comparing the value of the visibility kernel re-estimated in (c′

) to those in the LUT; and

wherein the initial design layout is modified in (e) further based on the re-estimated value of the quantity characteristic of EPE at time t from (d′

).

12. The method of claim 3, wherein the LUT comprises a list of entries, at least some of the entries comprising fields for the quantity characteristic of IFPIF, the quantity characteristic of IFPNF, and the corresponding quantity characteristic of EPE.

13. The method of claim 12, wherein at least some of the entries in the LUT further comprises one or more fields for etch time and/or feature depth.

14. The method of claim 12, wherein at least some of the entries in the LUT further comprises a field for in-feature passivant deposition flux (IFPDF).

15. The method of claim 12, wherein at least some of the entries in the LUT further comprise a field for edge shape indicator which corresponds to an edge shape present in the calibration pattern.

16. The method of claim 15, wherein the method further comprises determining an edge shape indicator for the feature to be etched by pattern matching the shape of said feature against the shapes of the features present in the calibration pattern, and using said determined edge shape indicator as a basis for searching the LUT.

17. The method of claim 16, wherein the LUT is searched first based on the feature'"'"'s determined edge shape indicator.

18. The method of claim 12:

wherein the LUT has been sorted based on one or more fields of the entries;

wherein the comparing in (d) comprises searching the LUT; and

wherein the estimating in (d) comprises interpolating between entries in the LUT after the searching.

19. The method of claim 18, wherein the interpolating comprises a polynomial-based interpolation scheme.

20. The method of claim 1, wherein the quantity characteristic of EPE is estimated in (d) using a trained machine learning model (MLM) which during operation:

compares one or more quantities characteristic of IFPF to those in the LUT; and

interpolates between values in the LUT;

wherein the MLM was trained on a dataset generated by running the computerized EPM, at least a subset of which was used to construct the LUT.

21. The method of claim 1,

wherein (c) and (d) are performed for t=t₁ to estimate a quantity characteristic of EPE at time t₁;

and

wherein the method further comprises performing (c) and (d) for t=t₂ (>

t₁), to estimate a quantity characteristic of EPE at time t₂;

wherein the initial design layout is modified in (e) based on the estimated values of the quantity characteristic of EPE at times t₁ and t₂; and

wherein the LUT was constructed by running the EPM at least to time t₂.

22. A method of generating a mask design, the method comprising:

generating a proximity-corrected design layout for photoresist using the method of claim 1;

generating a mask design based on the generated proximity-corrected photoresist design layout.

23. A method of etching a semiconductor substrate, the method comprising:

generating a mask design using the method of claim 22;

forming a mask based on the mask design;

performing a photolithography operation using the mask to transfer a layer of photoresist to the substrate substantially conforming to the proximity-corrected photoresist design layout; and

exposing the substrate to a plasma which etches the substrate.

24. The method of claim 1, wherein the computerized etch profile model (EPM) run on the calibration pattern of photoresist to construct the LUT used in (d) is a model which relates the etch profile of a feature on a semiconductor substrate to a set of independent input parameters, via the use of a plurality of model parameters, and which has been optimized by a method comprising:

(1) identifying a set of values for a selected set of the model parameters to be optimized;

(2) identifying multiple sets of values for a selected set of independent input parameters to optimize over;

(3) for each set of values specified in (2), receiving an experimental reflectance spectra generated from an optical measurement of an experimental etch process performed using the set of values specified in (2);

(4) for each set of values specified in (2), generating a computed reflectance spectra from the model using the set of values specified in (1) and (2); and

(5) modifying one or more values specified in (1) for the selected set of model parameters and repeating (4) with the modified set of values so as to reduce a metric indicative of the differences between the experimental reflectance spectra received in (3) and corresponding computed reflectance spectra generated in (4) with respect to one or more sets of values for the selected independent input parameters specified in (2);

wherein calculating the metric in (5) comprises an operation of;

calculating the differences between the computed and corresponding experimental reflectance spectra and projecting the differences onto a reduced-dimensional subspace; and

/or

projecting the computed and corresponding experimental reflectance spectra onto a reduced-dimensional subspace and calculating the difference between the reflectance spectra as projected onto the subspace.

25. The method of claim 1, wherein the computerized etch profile model (EPM) run on the calibration pattern of photoresist to construct the LUT used in (d) is a model which relates the etch profile of a feature on a semiconductor substrate to a set of independent input parameters, via the use of a plurality of model parameters, and which has been optimized by a method comprising:

(1) identifying a set of values for a selected set of the model parameters to be optimized;

(2) identifying multiple sets of values for a selected set of independent input parameters to optimize over;

(3) for each set of values specified in (2), receiving an experimental etch profile resulting from an experimental etch process performed using the set of values specified in (2);

(4) for each set of values specified in (2), generating a computed etch profile from the model using the set of values specified in (1) and (2); and

(5) modifying one or more values specified in (1) for the selected set of model parameters and repeating (4) with the modified set of values so as to reduce a metric indicative of the combined differences between the experimental etch profiles received in (3) and corresponding computed etch profiles generated in (4) over all the sets of values for the selected independent input parameters specified in (2);

wherein calculating the metrics in (5) comprises;

calculating the differences between the computed and corresponding experimental etch profiles and projecting the differences onto a reduced-dimensional subspace; and

/or

projecting the computed and corresponding experimental etch profiles onto a reduced-dimensional subspace and calculating the difference between the etch profiles as projected onto the subspace.

26. A computer system for generating a proximity-corrected design layout for photoresist to be used in an etch operation, the system comprising:

a processor, and

a memory, the memory storing a look-up table (LUT) and computer-readable instructions for execution on the processor, including instructions for;

(a) receiving an initial design layout;

(b) identifying a feature in the initial design layout, the feature'"'"'s pattern corresponding to a feature that would be etched into a material stack on a semiconductor substrate'"'"'s surface via a plasma-based etch process, performed in a processing chamber under a set of process conditions, when said stack is overlaid with a layer of photoresist pattern corresponding to the design layout;

(c) estimating one or more quantities characteristic of an in-feature plasma flux (IFPF) within the feature at a time t during such a plasma-based etch process;

(d) estimating a quantity characteristic of edge placement error (EPE) of the edge of the feature at time t by comparing the one or more quantities characteristic of the IFPF estimated in (c) to those in the LUT which associates values of the quantity characteristic of EPE at time t with values of the one or more quantities characteristics of the IFPF; and

(e) modifying the initial design layout based on at the quantity characteristic of EPE;

wherein the LUT was constructed by running a computerized etch profile model (EPM) under the set of process conditions at least to time t on a calibration pattern of photoresist overlaid on the material stack.

27. The computer system of claim 26:

wherein the initial design layout received in (a) is read from a computer-readable medium; and

wherein the computer-readable instructions stored in the memory for execution on the processor further include instructions for;

(f) writing the proximity-corrected design layout to a computer-readable medium.

29. A system for etching semiconductor substrates, the system comprising:

the computer system of claim 26;

a photolithography module configured to;

receive a proximity-corrected design layout for photoresist from the computer system;

form a mask from the proximity-corrected design layout; and

perform a photolithography operation using the mask to transfer a layer of photoresist to a semiconductor substrate substantially conforming to the proximity-corrected photoresist design layout; and

a plasma-etcher configured to generate a plasma which contacts the semiconductor substrate and etches those portions of the substrate surface not covered with photoresist transferred by the photolithography module.

28. One or more computer-readable media having a look-up table (LUT) and computer-readable instructions stored thereon, including instructions for:

(a) receiving an initial design layout;

(b) identifying a feature in the initial design layout, the feature'"'"'s pattern corresponding to a feature that would be etched into a material stack on a semiconductor substrate'"'"'s surface via a plasma-based etch process, performed in a processing chamber under a set of process conditions, when said stack is overlaid with a layer of photoresist pattern corresponding to the design layout;

(c) estimating one or more quantities characteristic of an in-feature plasma flux (IFPF) within the feature at a time t during such a plasma-based etch process;

(d) estimating a quantity characteristic of edge placement error (EPE) of the edge of the feature at time t by comparing the one or more quantities characteristic of the IFPF estimated in (c) to those in the LUT which associates values of the quantity characteristic of EPE at time t with values of the one or more quantities characteristics of the IFPF; and

(e) modifying the initial design layout based on at the quantity characteristic of EPE;

wherein the LUT was constructed by running a computerized etch profile model (EPM) under the set of process conditions at least to time t on a calibration pattern of photoresist overlaid on the material stack.

30. A method of estimating a quantity characteristic of an edge placement error (EPE) of an edge of a feature on a semiconductor substrate having a design layout of photoresist overlaid on a material stack, the feature to be etched in a real or simulated plasma-based etch process performed in a correspondingly real or simulated processing chamber under a set of process conditions, the method comprising:

(a) estimating one or more quantities characteristic of an in-feature plasma flux (IFPF) at time t during the etch; and

(b) estimating the quantity characteristic of EPE at time t by comparing the one or more quantities characteristic of the IFPF estimated in (a) to those in a LUT which associates values of EPE at time t with one or more quantities characteristic of the IFPF;

wherein the LUT was constructed by running a computerized etch profile model (EPM) under the set of process conditions at least to time t on a calibration pattern of photoresist overlaid on the material stack.