Process control by distinguishing a white noise component of a process variance

• US 7,221,990 B2
• Filed: 04/06/2006
• Issued: 05/22/2007
• Est. Priority Date: 05/28/2004
• Status: Expired due to Fees

• Current Assignee: Applied Materials Incorporated
• Original Assignee: Applied Materials Incorporated
• Inventors: Paik, Young J.
• Primary Examiner(s): Picard; Leo
• Assistant Examiner(s): Garland; Steven R.
• Application Number:
  US11/398,673
• Publication Number:
  US 20060195214A1
• Time in Patent Office: 411 Days
• Field of Search: 700/121, 700/108, 700/109, 700/31, 700/28, 700/71, 702/84
• US Class Current: 700/109
• CPC Class Codes: G05B 13/041 in which a variable is automatically adjusted to optimise the performance

Abstract

A method, system and medium is provided for enabling improved control systems. An error, or deviation from a target result, is observed for example during manufacture of semiconductor chips. The error within standard deviation is caused by two components: a white noise component and a signal component (such as systematic errors). The white noise component is, e.g., random noise and therefore is relatively non-controllable. The systematic error component, in contrast, may be controlled by changing the control parameters. A ratio between the two components is calculated autoregressively. Based on the ratio and using the observed or measured error, the actual value of the error caused by the systematic component is calculated utilizing an autoregressive stochastic sequence. The actual value of the error is then used in determining when and how to change the control parameters. The autoregressive stochastic sequence addresses the issue of the effects of run-to-run deviations, and provides a mechanism that can extract the white noise component from the statistical process variance in real time. This results in an ability to provide tighter control, for example in feedback and feedforward variations of process control.

Claims

1. A system, implemented on at least one computer, for compensating for a measurement of error between a measured characteristic of at least one product produced by a process and a target result of the characteristic, the system comprising:

(a) means for receiving at least one observed value for at least one product, and receiving a target value for the at least one product;

(b) means for determining a first measurement of error between the at least one observed value of the at least one product and the target value of the at least one product;

(c) means for determining a white noise portion of the first measurement of error caused by white noise; and

(d) means for using the determined white noise portion to adjust the process to reduce a second measurement of error.

2. The system of claim 1, wherein the at least one product is a semi-conductor wafer, and the process is an automated semiconductor manufacturing process.

3. The system of claim 1, wherein the target value is derived from a specification, the specification being selected from at least one of a predetermined specification, and a real-time calculation taken from a plurality of prior observed values of products.

4. The system of claim 1, wherein the process is associated with at least one control parameter capable of being controlled, further comprising means for controlling the at least one control parameter during the process based on the white noise portion of the first measurement of error.

5. The system of claim 4, wherein the system receives a plurality of observed values for a plurality of products including the at least one product, and determines the first measurement of error between the plurality of observed values and the target value, the system further comprises:

means for utilizing the white noise portion of the first measurement of error as a threshold and determining whether or not to control the at least one control parameter when at least one of the plurality of observed values is outside the threshold.

6. The system of claim 4, wherein the process comprises a first device and a second device on which a plurality of products including the at least one product are processed, the at least one observed value being relative to the first device, the second device being associated with the at least one control parameter, wherein controlling the at least one control parameter comprises affecting at least the second device.

7. The system of claim 1, wherein the first measurement of error is a variance, and the white noise portion of the first measurement of error is calculated using an autoregressive stochastic sequence.

8. The system of claim 1, wherein the white noise portion of the first measurement of error is substantially determined by:

67_x=y/(1+z)

where δ

_x=a value representing a variation that a system can control

y=calculated value from previous values

z=estimated gain adjustment.

9. A system, implemented on at least one computer, for compensating for a measurement of error between a measured characteristic of at least one product produced by a process and a target result of the characteristic, the system comprising:

(a) a value receiving component for receiving at least one observed value for at least one product, and receiving a target value for the at least one product;

(b) a data repository for storing the observed value and the target value;

(c) an error determination component configured to determine a first measurement of error between the observed value of at least one product and the target value of the at least one product;

(d) a white noise estimator component configured to determine a white noise portion of the first measurement of error caused by white noise; and

(e) a process adjustment component configured to adjust the process to reduce a second measurement of error using the determined white noise portion.

10. The system of claim 9, wherein the at least one product is a semi-conductor wafer, and the process is an automated semiconductor manufacturing process.

11. The system of claim 9, wherein the target value is derived from a specification, the specification being selected from at least one of a predetermined specification, and a real-time calculation taken from a plurality of prior observed values of products.

12. The system of claim 9, wherein the process is associated with at least one control parameter capable of being controlled, further comprising a controller configured to control the at least one control parameter during the process based on the white noise portion of the first measurement of error.

13. The system of claim 12, wherein the value receiving component is configured to receive a plurality of observed values for a plurality of products including the at least one product, and the error determination component is configured to determine the first measurement of error between the plurality of observed values and the target value, the system further comprises:

control determination component configured to utilize the white noise portion of the first measurement of error as a threshold and determine whether or not to control the at least one control parameter when at least one of the plurality of observed values is outside the threshold.

14. The system of claim 12, wherein the process comprises a first device and a second device on which a plurality of products including the at least one product are processed, the at least one observed value being relative to the first device, the second device being associated with the at least one control parameter, wherein controlling the at least one control parameter comprises affecting at least the second device.

15. The system of claim 9, wherein the first measurement of error is a variance, and the white noise portion of the first measurement of error is calculated using an autoregressive stochastic sequence.

16. The system of claim 9, wherein the white noise portion of the first measurement of error is substantially determined by:

δ

_x=y/(1z)

where δ

_x=a value representing a variation that a system can control

y=calculated value from previous values

z=estimated gain adjustment.

17. A computer-implemented method for compensating for a measurement of error between a measured characteristic of at least one product produced by a process and a target result of the characteristic, the method comprising:

(a) receiving at least one observed value for at least one product, and receiving a target value for the at least one product;

(b) determining a first measurement of error between the at least one observed value of the at least one product and the target value of the at least one product;

(c) determining a white noise portion of the first measurement of error caused by white noise; and

(d) using the determined white noise portion to adjust the process to reduce a second measurement of error.

18. The method of claim 17, wherein the at least one product is a semi-conductor wafer, and the process is an automated semiconductor manufacturing process.

19. The method of claim 17, wherein the target value is derived from a specification, the specification being selected from at least one of a predetermined specification, and a real-time calculation taken from a plurality of prior observed values of products.

20. The method of claim 17, wherein the process is associated with at least one control parameter capable of being controlled, the method further comprises controlling the at least one control parameter during the process based on the white noise portion of the first measurement of error.

21. The method of claim 20,

wherein receiving at least one observed value for at least one product comprises receiving a plurality of observed values for a plurality of products including the at least one product;

wherein determining a first measurement of error between the at least one observed value and the target value comprises determining a first measurement of error between the plurality of observed values and the target value,

the method further comprises utilizing the white noise portion of the first measurement of error as a threshold and determining whether or not to control the at least one control parameter when at least one of the plurality of observed values is outside the threshold.

22. The method of claim 20, wherein the process comprises a first device and a second device on which a plurality of products including the at least one product are processed, the at least one observed value being relative to the first device, the second device being associated with the at least one control parameter, wherein controlling the at least one control parameter comprises affecting at least the second device.

23. The method of claim 17, wherein the first measurement of error is a variance, and the white noise portion of the first measurement of error is calculated using an autoregressive stochastic sequence.

24. The method of claim 17, wherein the white noise portion of the first measurement of error is substantially determined by:

δ

_y=y/(1z)

where δ

_x=a value representing a variation that a system can control

y=calculated value from previous values

z=estimated gain adjustment.

25. A computer program product for use in compensating for a measurement of error between a measured characteristic of at least one product produced by a process and a target result of the characteristic, the computer program product comprising:

(a) at least one computer readable medium, readable by the manufacturing process;

(b) instructions, provided on the at least one computer readable medium, for receiving at least one observed value for at least one product, and receiving a target value for the at least one product;

(c) instructions, provided on the at least one computer readable medium, for determining a first measurement of error between the at least one observed value of the at least one product and the target value of the at least one product;

(d) instructions, provided on the at least one computer readable medium, for determining a white noise portion of the first measurement of error caused by white noise; and

(e) instructions, provided on the at least one computer readable medium, for using the determined white noise portion to adjust the process to reduce a second measurement of error.

26. The computer program product of claim 25, wherein the at least one product is a semi-conductor wafer, and the process is an automated semiconductor manufacturing process.

27. The computer program product of claim 25, wherein the target value is derived from a specification, the specification being selected from at least one of a predetermined specification, and a real-time calculation taken from a plurality of prior observed values of products.

28. The computer program product of claim 25, wherein the process is associated with at least one control parameter capable of being controlled, further comprising instructions, provided on the computer readable medium, for controlling the at least one control parameter during the process based on the white noise portion of the first measurement of error.

29. The computer program product of claim 28, further comprising instructions, provided on the computer readable medium, for receiving a plurality of observed values for a plurality of products including the at least one product, determining a white noise portion of a first measurement of error between the plurality of observed values and the target value caused by white noise and utilizing the white noise portion as a threshold, and determining whether or not to control the at least one control parameter when at least one of the plurality of observed values is outside the threshold.

30. The computer program product of claim 28, wherein the process comprises a first device and a second device on which a plurality of products including the at least one product are processed, the at least one observed value being relative to the first device, the second device being associated with the at least one control parameter, wherein the instructions for controlling the at least one control parameter comprises affecting at least the second device.

31. The computer program product of claim 25, wherein the first measurement of error is a variance, and the white noise portion of the first measurement of error is calculated using an autoregressive stochastic sequence.

32. The computer program product of claim 25, wherein the white noise portion of the first measurement of error is substantially determined by:

δ

_x=y/(1z)

where δ

_x=a value representing a variation that a system can control

y=calculated value from previous values

z=estimated gain adjustment.