Method and system for event detection in plasma processes

• US 20030223055A1
• Filed: 05/30/2002
• Published: 12/04/2003
• Est. Priority Date: 05/30/2002
• Status: Active Grant

• Current Assignee: Applied Materials Incorporated
• Original Assignee: Applied Materials Incorporated
• Inventors: Agarwal, Ashish, Lymberopoulos, Dimitris
• Granted Patent: US 7,006,205 B2
• Time in Patent Office: Days
• US Class Current: 356/72
• CPC Class Codes: G01N 21/73 using plasma burners or torches
• CPC Class Codes: H01J 37/32935 Monitoring and controlling tubes by information coming from the object and/or discharge
• CPC Class Codes: H01J 37/32972 Spectral analysis

Abstract

Plasma events are detected by analyzing the spectral emissions of a plasma process of a substrate. A plasma is monitored by a spectrometer which produces plasma emission data which includes an intensity value for each individual wavelength which is a large quantity of information. The plasma emission data is processed with an algorithm or combination of algorithms to reduce the quantity of the plasma emission data. A peak finding algorithm which identifies the wavelengths of light which are associated with a plasma process allowing the other wavelengths to be ignored. A data reduction algorithm provides a single value representative of the intensity of the emitted light from each peak. A noise reduction algorithm removes noise from the spectral signal by eliminating signals at wavelengths which do not exceed a threshold intensity and do not exceed a threshold wavelength span. The data may also be processed with principal component analysis to further reduce the optical emission data. By reducing the optical emission data, the neural network can provide faster data analysis. The neural network of the inventive system identifies the plasma event after being trained and can control the processing equipment when plasma events are detected.

Claims

1. A plasma processing system, comprising:

a plasma processing chamber capable of processing a substrate with a plasma;

a spectrometer for monitoring an optical emission of the plasma and producing a optical emission data;

an algorithm for reducing the quantity of optical emission data; and

a neural network for detecting a plasma processing event by analyzing the reduced optical emission data.

2. The plasma processing system of claim 1, wherein the algorithm is a peak finding algorithm which analyzes the spectral emission data from the spectrometer and identifies a plurality of optical wavelengths associated with the plasma emission as the wavelengths between a rapid rise in intensity followed by a rapid decline in intensity.

3. The plasma processing system of claim 2 wherein the rapid rise in intensity is defined as being an increase in emission intensity of more than 50 au over a 3 nm change in wavelength.

4. The plasma processing system of claim 2 wherein the algorithm is a data reduction algorithm which produces a cumulative value for the optical emission of the plurality of wavelengths between the rapid rise in intensity and the rapid decline in intensity.

5. The plasma processing system of claim 3 wherein the algorithm is a principal component analysis algorithm which reduces the quantity of optical emission data.

6. The plasma processing system of claim 1, wherein the algorithm is a noise filter which removes noise data from the optical emission data.

7. The plasma processing system of claim 6, wherein the noise filter removes spectral signals that have an intensity of less than about 50 au over a wavelength range of 3 nm.

8. The plasma processing system of claim 1, further comprising:

a controller for adjusting an operating condition of the plasma processing chamber in response to the plasma processing event being detected by the neural network.

9. A method of plasma processing a substrate, comprising the steps of:

exposing the substrate to a plasma in a plasma processing chamber;

monitoring the plasma with a spectrometer which produces plasma emission data;

processing the plasma emission data with an algorithm to obtain a reduced quantity of plasma emission data;

analyzing the reduced quantity of plasma emission data with a neural network to identifying a plasma processing event.

10. The method of plasma processing a substrate of claim 9 wherein the algorithm of the processing step is a peak finding algorithm which removes signals not associated with the plasma emission.

11. The method of plasma processing a substrate of claim 9 wherein the peak finding algorithm identifies a plurality of optical wavelengths associated with the plasma emission as the wavelengths between a rapid rise in intensity followed by a rapid decline in intensity.

12. The method of plasma processing a substrate of claim 11 wherein the rapid rise in intensity is defined as an increase in emission intensity of more than 50 au over a 3 nm change in wavelength and the rapid decline in intensity is defined as a decrease in emission intensity of more than 50 au over a 3 nm change in wavelength.

13. The method of plasma processing a substrate of claim 11 wherein the algorithm in the processing step includes a data reduction algorithm which produces a cumulative value for the optical emission of the plurality of wavelengths between the rapid rise in intensity and the rapid decline in intensity.

14. The method of plasma processing a substrate of claim 9 wherein the algorithm of the processing step is a principal component analysis algorithm.

15. The method of plasma processing a substrate of claim 9 further comprising:

adjusting an operating condition of the plasma processing chamber in response to identifying a plasma processing event.

16. A method for analyzing plasma emissions, comprising the steps of:

providing a plasma processing chamber containing a substrate;

exposing the substrate to a plasma;

monitoring the optical emission of the plasma with a spectrometer which produces plasma emission data;

processing the plasma emission data with an algorithm to obtain a reduced quantity of plasma emission data;

training a neural network to process the reduced quantity of plasma emission data; and

identifying a plasma processing event with the trained neural network from the reduced quantity of plasma emission data.

17. The method of claim 16 for analyzing plasma emissions, further comprising:

identifying a second plasma processing event with the neural network.

18. The method of claim 16 for analyzing plasma emissions wherein the training step includes an automatically training the neural network.

19. The method of claim 16 for analyzing plasma emissions wherein the training step includes an manually training the neural network.

20. The method of claim 16 for analyzing plasma emissions further comprising:

adjusting a plasma processing condition in response to the identification of the plasma processing event.