Method for transferring process control models between plasma processing chambers
View Patent ↗A method of transferring a multi-variate process control model from one RF-powered plasma processing chamber (the reference chamber) to another nominally identical RF-powered plasma processing chamber (the target chamber) comprises first determining a multi-variate process control model on the reference tool based upon sensor data, 100 . Then, a set of sensor data is taken for the reference chamber by running a designed experiment, 102 , and a corresponding set of sensor data is taken for the target chamber using the same designed experiment, 104 . The relationship between the reference and target chambers is determined by comparing the results of the designed experiments and calculating a transform matrix representing differences between the two chambers, 106 . Finally, the process control model is transformed from the reference chamber to the target chamber using the transform matrix thus obtained, 108.
1. A method of transferring a multi-variate process control model from one plasma processing chamber (the reference chamber) to another nominally identical plasma processing chamber (the target chamber), comprising the steps of:
(a) determining a multi-variate process control model on the reference tool based upon sensor data,
(b) taking a fingerprint of the reference chamber by running a designed experiment,
(c) taking a fingerprint of the target chamber by running the same designed experiment,
(d) determining the relationship between the reference and target chambers by comparing the results of the designed experiments and calculating a transform matrix representing differences between the two chambers, and
(e) transforming the process control model from the reference chamber to the target chamber using the transform matrix obtained in step (d).
2. The method of claim 1 , wherein step (a) is performed using a designed experiment and steps (b) and (c) are performed on a simpler designed experiment than that used for step (a).
3. The method of claim 2 , wherein the designed experiment performed in steps (b) and (c) is a subset of the designed experiment performed in step (a).
4. The method of claim 2 , wherein the designed experiment performed in steps (b) and (c) is performed without a plasma in the chamber.
5. The method claimed in claim 1 , wherein the plasma processing chamber is an RF-powered plasma processing chamber.
6. The method of claim 5 , wherein the fingerprint in steps (b) and (c) is recorded using at least an RF impedance sensor.
7. The method of claim 5 , wherein the fingerprint in steps (b) and (c) is recorded using at least one sensor other than an RF impedance sensor.
8. The method of claim 1 , wherein the fingerprint in steps (b) and (c) is recorded using more than one sensor.
9. The method of claim 5 , wherein the sensor data comprises a plurality of Fourier components of the RF power source driving the chamber.
10. The method of claim 1 , wherein the process control model is a fault library.
11. The method of claim 1 , wherein the process control model is a yield-prediction model.
12. The method of claim 1 , wherein the reference and target chamber are one and the same chamber and the transformation method is used to update the process control model on the chamber.