Method and system for tuning advanced process control parameters

• US 8,229,588 B2
• Filed: 03/03/2009
• Issued: 07/24/2012
• Est. Priority Date: 03/03/2009
• Status: Active Grant

• Current Assignee: Taiwan Semiconductor Manufacturing Company Limited
• Original Assignee: Taiwan Semiconductor Manufacturing Company Limited
• Inventors: Tsen, Andy, Hsu, Chih-Wei, Fan, Ming-Yu, Fei, Wang Jo, Mou, Jong-I, Hung, Ming-Yeon
• Primary Examiner(s): Decady, Albert
• Assistant Examiner(s): WHITTINGTON, ANTHONY T
• Application Number:
  US12/396,996
• Publication Number:
  US 20100228370A1
• Time in Patent Office: 1,239 Days
• Field of Search: 700/103, 700/117, 700/121
• US Class Current: 700/121
• CPC Class Codes: G05B 13/044 not using a perturbation signal
• CPC Class Codes: H01L 21/67276 Production flow monitoring, e.g. for increasing throughput program-control systems per se G05B19/00, e.g. total factory control G05B19/418

Abstract

A method of advanced process control (APC) for semiconductor fabrication is provided. The method includes providing a present wafer to be processed by a semiconductor processing tool, providing first data of previous wafers that have been processed by the semiconductor processing tool, decoupling noise from the first data to generate second data, evaluating an APC performance based on proximity of the second data to a target data, determining a control parameter based on the APC performance, and controlling the semiconductor processing tool with the control parameter to process the present wafer.

Claims

1. A method of advanced process control (APC) for semiconductor fabrication, comprising:

providing a present wafer to be processed by a semiconductor processing tool;

providing first data of previous wafers that have been processed by the semiconductor processing tool;

decoupling noise from the first data to generate second data;

evaluating an APC performance based on proximity of the second data to a target data;

determining a control parameter based on the APC performance; and

controlling the semiconductor processing tool with the control parameter to process the present wafer,

wherein the control parameter includes a process time for a process, and

wherein decoupling noise from the first data includes;

identifying an APC impact on the first data to extract a chamber effect on the process, wherein the chamber effect includes a condition of the processing tool that can influence the process; and

removing a pre-stage effect on the process, wherein the pre-stage effect includes a pre-stage influencing the process.

2. The method of claim 1, wherein the semiconductor processing tool includes an etching tool, and wherein the control parameter includes a trim time for an etching process.

3. The method of claim 2, wherein

the process is an etching process,

wherein the chamber effect includes a condition of the etching tool that can influence the etching process;

and

wherein removing the pre-stage effect on the process includes removing a pre-stage effect on the etching process, wherein the pre-stage effect includes an after-development-inspection of a patterning process.

4. The method of claim 3, wherein an APC algorithm of the etching process is described by:

a model equation expressed as;

y=a×

x+b=R_trim×

T_trim+model_—

b

wherein y denotes a trimmed line width, a denotes a trim rate referred to as R_trim, X denotes a trim time referred to as T_trim, and b denotes a constant referred to as model_b.

a control equation expressed as;

u=T−

b/a

wherein u denotes a trim time to be applied to the etching process and T denotes a target trimmed line width; and

an exponentially weighed moving average (EWMA) equation expressed as;

b_n+1=ω

×

(y_n−

a×

x_n)+(1−

ω

)×

b_n

wherein ω

is between 0 and 1, and the EWMA equation is used to update b.

5. The method of claim 4, wherein identifying the APC impact includes:

decoupling the APC impact on the first data by setting a constant trim time, the constant trim time being an average trim time that was implemented to process the previous wafers;

determining a clean trim rate by a trim time different from the average trim time; and

determining an adjusted trimmed line width based on the clean trim rate and a difference between the average trim time and the different trim time.

6. The method of claim 1, wherein evaluating the APC performance includes:

determining a control profile distribution of an absolute difference between a measured trimmed line width of the first data and a decoupled trimmed line width of the second data;

determining a target profile distribution of an absolute difference between the measured trimmed line width of the first data and a target trimmed line width of the target data;

determining a first set of mean and standard deviation of the control profile distribution and a second set of mean and standard deviation of the target profile distribution;

determining a distance between the first set and the second set; and

setting a APC performance index based on the distance.

7. The method of claim 6, wherein determining the control parameter includes determining the control parameter such that the distance between the first set and the second set approaches zero.

8. The method of claim 1, further comprising determining another control parameter based on the APC performance, the another control parameter including a sidewall angle (SWA) parameter.

9. The method of claim 8, wherein decoupling noise from the first data includes:

identifying an APC impact on the first data to extract a chamber effect on the etching process, wherein the chamber effect includes a condition of an etching tool that can influence an etching process;

removing a pre-stage effect on the etching process, wherein the pre-stage effect includes an after-development-inspection of a patterning process; and

removing a SWA effect on the etching process.

10. A method of advanced process control (APC) of an etching process, comprising:

providing data of previous wafers that have undergone the etching process;

removing an APC impact by a first modification of the data, the first modification isolating a chamber effect on the data, the chamber effect including a condition of an etching tool that can influence the etching process;

removing a pre-stage effect by a second modification of the data, the pre-stage effect including an after-development-inspection process;

determining an APC performance based on proximity of the modified data to a target data, the modified data including the first and second modifications of the data;

determining a control parameter for the etching process based the APC performance; and

controlling the etching process with the control parameter to process a present wafer,

wherein determining the APC performance includes;

determining a control profile distribution of an absolute difference between a measured trimmed line width of the unmodified data and a decoupled trimmed line width of the modified data;

determining a target profile distribution of an absolute difference between the measured trimmed line width of the unmodified data and a target trimmed line width of the target data;

determining a first set of mean and standard deviation of the control profile distribution and a second set of mean and standard deviation of the target profile distribution;

determining a distance between the first set and the second set; and

setting a APC performance index based on the distance.

11. The method of claim 10, wherein the data includes an after-development-inspection (ADI) critical dimension of a line width prior to the etching process and an after-etch-inspection (AEI) critical dimension of the line width after the etching process.

12. An advanced process control (APC) system for controlling a semiconductor processing tool, comprising:

memory operable to store first data of previous wafers that have been processed by the semiconductor processing tool; and

a controller operable to control the semiconductor processing tool based on instructions that;

decouple noise from the first data to generate a second data;

evaluate an APC performance based on proximity of the second data to a target data; and

determine a control parameter based on the APC performance; and

control the semiconductor processing tool with the control parameter to process a present wafer,

wherein the control parameter includes a trim time for an etching process,

wherein the instructions that decouple noise from the first data include instructions that;

identify an APC impact on the first data to extract a chamber effect on the etching process, wherein the chamber effect includes a condition of an etching tool that can influence the etching process; and

remove a pre-stage effect on the etching process, wherein the pre-stage effect includes an after-development-inspection of a patterning process.

13. The APC system of claim 12, wherein the instructions further include instructions that determine another control parameter based on the APC performance, the another control parameter including a sidewall angle (SWA) parameter;

and

wherein the instructions that decouple noise from the first data further includes instructions that remove a SWA effect on the etching process.

14. The APC system of claim 12, wherein an APC algorithm of the etching process is described by:

a model equation expressed as;

y=a×

x+b=R_trim×

T_trim+model_—

b

wherein y denotes a trimmed line width, a denotes a trim rate referred to as R_trim, X denotes a trim time referred to as T_trim, and b denotes a constant referred to as model_b;

a control equation expressed as;

15. The APC system of claim 14, wherein the instructions that identify the APC impact includes instructions that:

decouple the APC impact on the data by setting a constant trim time, the constant trim time being an average trim time that was implemented to process the previous wafers;

determine a clean trim rate by a trim time different from the average trim time; and

determine an adjusted trimmed line width based on the clean trim rate and a difference between the average trim time and the different trim time.

16. The APC system of claim 12, wherein the instructions that evaluate the APC performance include instructions that:

determine a control profile distribution of an absolute difference between a measured trimmed line width of the first data and a decoupled trimmed line width of the second data;

determine a target profile distribution of an absolute difference between the measured trimmed line width of the first data and a target trimmed line width of the target data;

determine a first set of mean and standard deviation of the control profile distribution and a second set of mean and standard deviation of the target profile distribution;

determine a distance between the first set and the second set; and

setting a APC performance index based on the distance.

17. The system of claim 16, wherein the instructions determine the control parameter includes instructions that determine the control parameter such that the distance between the first set and the second set approaches zero.