End point detection method for plasma etching of semiconductor wafers with low exposed area

• US 20060157446A1
• Filed: 01/18/2006
• Published: 07/20/2006
• Est. Priority Date: 01/19/2005
• Status: Active Grant

• Current Assignee: Pivotal Systems Corporation
• Original Assignee: Pivotal Systems Corporation
• Inventors: Gorin, Georges J., Johel, Sumer, Lane, Barton, Kieffel, Herve C., Spruytte, Sylvia G. J. P.
• Granted Patent: US 7,871,830 B2
• Time in Patent Office: Days
• US Class Current: 216/59
• CPC Class Codes: H01J 37/32082 Radio frequency generated discharge H01J37/32357, H01J37/32366, H01J37/32394 and H01J37/32403 take precedence
• CPC Class Codes: H01J 37/32623 Mechanical discharge control means
• CPC Class Codes: H01J 37/32935 Monitoring and controlling tubes by information coming from the object and/or discharge
• CPC Class Codes: H01L 22/26 Acting in response to an ongoing measurement without interruption of processing, e.g. endpoint detection, in-situ thickness measurement endpoint detection arrangements in CMP apparatus B24B37/013, in discharge apparatus H01J37/32

Abstract

A method for controlling the plasma etching of semiconductor wafers determines the impedance of a plasma chamber using values representing voltage, current, and the phase angle between them, as provided by a sensor. All or less than all of the data during a first time period may be used to calculate a model. During a second time period, real time data is used to calculate a version of the instant impedance of the chamber. This version of impendence is compared to a time-projected version of the model. The method determines that etching should be stopped when the received data deviates from the extrapolated model by a certain amount. In some embodiments a rolling average is used in the second time period, the rolling average compared to the model to determine the end point condition.

Claims

1. A method for controlling a plasma processing chamber, wherein the plasma processing chamber includes an RF generator, the RF generator providing a drive signal to an electrode to generate a gas plasma, and an RF sensor for sensing operating characteristics of the drive signal, the method comprising:

(1) commanding the plasma processing chamber to begin processing;

(2) receiving a plurality of data sets from the RF sensor, each data set corresponding to an operating characteristic of the drive signal during a first time period, and wherein each data set corresponds to a unique time point;

(3) saving the data sets received from the RF sensor and their corresponding time points;

(4) calculating a mathematical model of the data sets received from the RF sensor during the first time period;

(5) receiving an additional data set from the RF sensor during a second time period;

(6) saving the additional data set received from the RF sensor and its corresponding time point;

(7) calculating a version of the mathematical model for a time point corresponding to the time point of the additional data set;

(8) calculating a residual between the additional data set and the version of the mathematical model for a time point corresponding to the time point of the additional data set;

(9) determining if at least one of the following conditions is true;

Condition 1;

the residual is greater than a predetermined amount;

Condition 2;

the time point of the additional data set is later than a predetermined maximum time;

(10) if at least one of the Conditions 1 and 2 is true, commanding the plasma processing chamber to stop processing; and

(11) if none of the Conditions 1 and 2 is true, then repeating the process steps beginning with the operation (5) above.

2. The method according to claim 1, wherein less than all of the data sets received during the first time period are used to calculate the mathematical model.

3. The method according to claim 1, wherein each data set comprises a representation of RMS voltage provided by the RF generator to the electrode of the plasma processing chamber.

4. The method according to claim 1, wherein each data set comprises a representation of RMS current provided by the RF generator to the electrode of the plasma processing chamber.

5. The method according to claim 1, wherein each data set comprises a representation of phase angle between the RMS voltage and the RMS current provided by the RF generator to the electrode of the plasma processing chamber.

6. The method according to claim 1, wherein each data set comprises a representation of electrical reactance of the plasma processing chamber, wherein the representation of the electrical reactance is received from the RF sensor.

7. The method according to claim 1, wherein a characteristic of the plasma processing chamber is calculated from each data set, wherein the calculated characteristic of a data set corresponds to the time point of the data set.

8. The method according to claim 7, wherein the calculated characteristic corresponds to the instant electrical reactance of the plasma processing chamber.

9. The method according to claim 8, wherein the instant electrical reactance of the plasma processing chamber is calculated by the formula of Reactance=(V*SIN(Theta)/I), wherein V, I, and Theta respectively represent the RMS voltage, RMS current, and the angle between the RMS voltage and the RMS current, provided to the electrode of the plasma processing chamber by the RF generator.

10. The method according to claim 1 wherein the mathematical model is calculated by fitting the data sets of the first time period to an exponential curve.

11. The method according to claim 10, wherein the exponential curve is of the form X(t_i)=A+Be(exp (−

λ

*t_i)), wherein X(t_i) is instant reactance of the plasma processing chamber, A and B and λ

are empirically derived constants, ant t_i is the time of the instant data set.

12. The method according to claim 10, wherein the mathematical model is modified to reduce the effect of noise during the first time period.

13. The method according to claim 1, wherein the Condition 1 predetermined amount is compared to a standard deviation of the residual.

14. The method according to claim 13, wherein the standard deviation is approximately two.

15. The method according to claim 1, wherein the Condition 1 predetermined amount is a sliding ten-point average of a standard deviation of the residual.

16. The method according to claim 15, wherein the sliding ten-point average of the standard deviation is approximately two.