Full spectrum endpoint detection

• US 6,969,619 B1
• Filed: 02/18/2003
• Issued: 11/29/2005
• Est. Priority Date: 02/18/2003
• Status: Active Grant

• Current Assignee: Novellus Systems Incorporated (Lam Research Corporation)
• Original Assignee: Novellus Systems Incorporated (Lam Research Corporation)
• Inventors: Winniczek, Jaroslaw W.
• Primary Examiner(s): Norton, Nadine G.
• Assistant Examiner(s): UMEZ ERONINI, LYNETTE T
• Application Number:
  US10/369,085
• Time in Patent Office: 1,015 Days
• Field of Search: 438/9, 438/14, 438/16, 438/17, 438/706, 438/710
• US Class Current: 438/9
• CPC Class Codes: G01N 21/73 using plasma burners or torches
• CPC Class Codes: H01J 37/32963 End-point detection
• CPC Class Codes: H01J 37/32972 Spectral analysis
• CPC Class Codes: H01L 21/3065 Plasma etching; Reactive-ion etching

Abstract

A method of endpoint detection during plasma processing of a semiconductor wafer comprises processing a semiconductor wafer using a plasma, detecting radiation emission from the plasma during the semiconductor processing, and tracking data points representing changes in spectra of the radiation as a function of time during the semiconductor processing. At any point prior to or during processing a plurality of profiles are provided, each profile representing a different processing condition affecting detection of the desired plasma processing endpoint of the semiconductor wafer. After selecting a desired profile, a first set of parameters are input, representing simplified values for determining when changes in spectra of the radiation indicate that plasma processing of the semiconductor wafer reaches a desired endpoint. The selected profile converts the input first set of parameters into a larger, second set of parameters, and then applies the second set of parameters to an algorithm that converts data points from the spectra of the radiation as a function of time into an endpoint curve. The method then uses the algorithm to track changes in spectra of the radiation as a function of time and determine when plasma processing of the semiconductor wafer reaches a desired endpoint.

Claims

1. A method of endpoint detection during plasma processing of a semiconductor wafer, comprising:

processing a semiconductor wafer using a plasma;

detecting radiation emission from the plasma during the semiconductor processing;

tracking data points representing changes in spectra of the radiation as a function of time during the semiconductor processing;

providing at any point prior to or during processing a plurality of profile programs, each profile program representing a different processing condition affecting detection of a plasma processing endpoint of the semiconductor wafer;

selecting a profile program;

inputting a first set of parameters into the selected profile program, the first set of parameters representing simplified values for determining when changes in spectra of the radiation indicate that plasma processing of the semiconductor wafer reaches an endpoint;

using the selected profile program, converting the input first set of parameters into a larger, second set of parameters;

applying the second set of parameters to an algorithm that converts data points from the spectra of the radiation as a function of time into an endpoint curve; and

using the algorithm to track changes in spectra of the radiation as a function of time and determine when plasma processing of the semiconductor wafer reaches an endpoint.

2. The method according to claim 1 wherein the processing conditions represented by the profile programs include signal-to-noise ratio and data collection rate.

3. The method according to claim 1 wherein the first set of parameters input into the profile program includes process time, and wherein the selected profile program converts the process time into sampling interval of the data points.

4. The method according to claim 1 wherein the first set of parameters input into the profile program is selected from the group consisting of endpoint threshold value, endpoint threshold crossing on peak rise, endpoint threshold crossing on peak top, endpoint threshold crossing on peak fall, maximum processing time, endpoint delay time, and relative detection gain setting.

5. The method according to claim 1 wherein the selected profile program converts the first set of parameters into one or more of the following parameters:

sampling interval, detector integration time, detector N average, filtering parameter, normalization period, amplitude or derivative, derivative smoothing filter, and threshold integrity period.

6. The method according to claim 1 wherein the algorithm comprises a single equation.

7. The method according to claim 6 wherein the algorithm is embedded in the selected profile program.

8. The method according to claim 1 wherein the algorithm comprises a full spectrum analysis of the spectra of the radiation emitted from the semiconductor wafer during plasma processing.

13. The method according to claim 6 wherein the algorithm is embedded in the selected profile program.

9. A method of endpoint detection during plasma processing of a semiconductor wafer, comprising:

processing a semiconductor wafer using a plasma;

detecting radiation emission from the plasma during the semiconductor processing;

tracking data points representing changes in spectra of the radiation as a function of time during the semiconductor processing;

providing a plurality of profile programs, each profile program representing a different processing condition affecting detection of a plasma processing endpoint of the semiconductor wafer;

selecting a profile program;

inputting a first set of parameters, including process time, into the selected profile program, the first set of parameters representing simplified values for determining when changes in spectra of the radiation indicate that plasma processing of the semiconductor wafer reaches an endpoint;

using the selected profile program, converting the input first set of parameters into a larger, second set of parameters, including converting the process time into sampling interval of the data points;

applying the second set of parameters to an algorithm comprising a full spectrum analysis of the spectra of the radiation emitted from the semiconductor wafer during plasma processing, the algorithm converting data points from the spectra of the radiation as a function of time into an endpoint curve; and

using the algorithm to track changes in spectra of the radiation as a function of time and determine when plasma processing of the semiconductor wafer reaches an endpoint.

10. The method according to claim 9 wherein the processing conditions represented by the profile programs include signal-to-noise ratio and data collection rate.

11. The method according to claim 9 wherein the first set of parameters input into the profile program further includes parameters selected from the group consisting of endpoint threshold value, endpoint threshold crossing on peak rise, endpoint threshold crossing on peak top, endpoint threshold crossing on peak fall, maximum processing time, endpoint delay time, and relative detection gain setting.

12. The method according to claim 9 wherein the selected profile program converts the first set of parameters into one or more of the following parameters:

sampling interval, detector integration time, detector N average, filtering parameter, normalization period, amplitude or derivative, derivative smoothing filter, and threshold integrity period.

14. A method of endpoint detection during plasma processing of a semiconductor wafer, comprising:

processing a semiconductor wafer using a plasma;

detecting radiation emission from the plasma during the semiconductor processing;

tracking data points representing changes in spectra of the radiation as a function of time during the semiconductor processing;

providing a plurality of profiles, each profile representing a different processing condition affecting detection of a plasma processing endpoint of the semiconductor wafer;

selecting a profile;

inputting a first set of parameters, including relative detection gain setting, into the selected profile, the first set of parameters representing simplified values for determining when changes in spectra of the radiation indicate that plasma processing of the semiconductor wafer reaches an endpoint;

using the selected profile, converting the input first set of parameters into a larger, second set of parameters, including converting the relative detection gain setting into integration time for the data points;

applying the second set of parameters to an algorithm comprising a full spectrum analysis of the spectra of the radiation emitted from the semiconductor wafer during plasma processing, the algorithm converting data points from the spectra of the radiation as a function of time into an endpoint curve; and

using the algorithm to track changes in spectra of the radiation as a function of time and determine when plasma processing of the semiconductor wafer reaches an end point.

15. The method according to claim 14 wherein the processing conditions represented by the profiles include signal-to-noise ratio and data collection rate.

16. The method according to claim 14 wherein the first set of parameters input into the profile further includes parameters selected from the group consisting of endpoint threshold value, endpoint threshold crossing on peak rise, endpoint threshold crossing on peak top, endpoint threshold crossing on peak fall, maximum processing time, endpoint delay time, and relative detection gain setting.

17. The method according to claim 14 wherein the selected profile converts the first set of parameters into one or more of the following parameters:

sampling interval, detector integration time, detector N average, filtering parameter, normalization period, amplitude or derivative, derivative smoothing filter, and threshold integrity period.

18. The method according to claim 14 wherein the algorithm is embedded in the selected profile.

19. A method of endpoint detection during plasma processing of a semiconductor wafer, comprising:

processing a semiconductor wafer using a plasma;

detecting radiation emission from the plasma during the semiconductor processing;

tracking data points representing changes in spectra of the radiation as a function of time during the semiconductor processing;

providing at any point prior to or during processing a plurality of profiles, each profile representing a different processing condition affecting detection of a plasma processing endpoint of the semiconductor wafer;

selecting a profile;

inputting a first set of parameters into the selected profile, the first set of parameters representing simplified values for determining when changes in spectra of the radiation indicate that plasma processing of the semiconductor wafer reaches an endpoint, and including relative detection gain setting;

using the selected profile, converting the input first set of parameters into a larger, second set of parameters, including converting the relative detection gain setting into integration time for the data points;

applying the second set of parameters to an algorithm that converts data points from the spectra of the radiation as a function of time into an endpoint curve; and

using the algorithm to track changes in spectra of the radiation as a function of time and determine when plasma processing of the semiconductor wafer reaches an endpoint.