Method to detect wafer arcing in semiconductor manufacturing equipment
Methods and systems for accurate arc detection in semiconductor manufacturing tools are disclosed. Such methods and systems provide real-time arc detection and near real-time notification for corrective actions during a semiconductor manufacturing process. Such methods and systems utilize data with high sample rate and wavelet analysis to provide for more accurate arc detection, which leads to more effective and cost efficient semiconductor manufacturing operations.
1. A method comprising:
sampling, by a sensor, data of an electrical parameter associated with an arc event within a semiconductor processing chamber during a semiconductor process;
filtering, using discrete wavelet transforms over time for different time windows and frequency bands, the sampled data to generate a time domain arc signal;
selecting a predetermined alarm value as a time domain spike threshold for the electrical parameter;
comparing the time domain arc signal with the predetermined alarm value;
determining whether an arc event occurred based on the comparing; and
halting the semiconductor process if it is determined that the arc event occurred.
2. The method according to claim 1 , wherein the electrical parameter is a voltage.
3. The method according to claim 1 , wherein the electrical parameter is a bias voltage of an electrode.
4. The method according to claim 1 , wherein the sampling samples the electrical parameter at a sampling rate greater than 200 kHz.
5. The method according to claim 1 , wherein the sampling samples the electrical parameter at a sampling rate greater than 500 kHz.
6. The method according to claim 1 , wherein the predetermined alarm value comprises a plurality of predetermined alarm values.
7. The method according to claim 1 , wherein the predetermined alarm value is different for a different semiconductor process.
8. The method according to claim 1 , wherein determining the predetermined alarm value includes:
determining a specific process during which the sampled data is sampled; and
looking up the predetermined alarm value for the specific process from a database.
9. A process monitoring system comprising:
a process parameter sensor for sampling data of an electrical parameter associated with an arc event within a semiconductor processing chamber; and
a processor coupled with the process parameter sensor, wherein the processor is configured to:
receive sampled data from the process parameter sensor during a process;
filter, using discrete wavelet transforms over time for different time windows and frequency bands, the sampled data to generate a time domain arc signal;
select a predetermined alarm value as a time domain spike threshold for the electrical parameter;
compare the time domain arc signal with the predetermined alarm value;
determine whether an arc event occurred based on the comparing; and
halt the process if it is determined that the arc event occurred.
10. The process monitoring system according to claim 9 , wherein the process parameter sensor comprises a data acquisition unit that samples the data at a rate greater than 200 kHz.
11. The process monitoring system according to claim 9 , wherein the process parameter sensor comprises a data acquisition unit that samples the data at a rate greater than 500 kHz.
12. The process monitoring system according to claim 9 , wherein the electrical parameter comprises a voltage.
13. The process monitoring system according to claim 9 , wherein the electrical parameter is a bias voltage of an electrode.
14. The process monitoring system according to claim 9 , wherein the predetermined alarm value comprises a plurality of predetermined alarm values.
15. The process monitoring system according to claim 9 , wherein the predetermined alarm value is different for a different process.
16. The system of according to claim 9 , further comprising:
a database coupled with the processor;
wherein the processor is configured to determine the predetermined alarm value by:
determining a specific process during which the sampled data is sampled; and
looking up the predetermined alarm value for the specific process from the database.