Review Reports - Factor Design for the Oxide Etching Process to Reduce Edge Particle Contamination in Capacitively Coupled Plasma Etching Equipment

Round 1

Reviewer 1 Report

The paper presented several keep knobs to reduce the particles during a dry plasma etching process. The results are presented in a clear manner and the experiment design is systematic. The results would be interesting to the semiconductor industry as well as academia. It meets the criteria for publication without revision. But I do recommend some improvement in the future work.

The sample size the author presented is ok to show a general trend. But it would be more convincing if you can use 10 + samples for the same condition and use a t-test to conclude whether the improvement of particle performance is statistically significant. I understand those particle-grade wafers are expensive. It might not be feasible to use a 10+ wafer for each condition. But it would be good to have t-test results for the few keep knobs that the author is claiming to have the best particle reduction performance.
In figure 11, pressure. The author listed it as 0mT which is not achievable. Maybe the author should list the actual pressure getting from the manometer?
11, in the first particle map image, there seems to be a cluster of particles that might not be related to the process. It is more likely related to some debris peeling off either from the chamber or transfer module. Maybe the author should exclude samples demonstrating non-process-related particle maps.
There are spelling mistakes: Fig. 13, a, b. Inter-mitten should be spelled as intermittent. Line 285, high RF should be changed to Top RF to keep it consistent with the nomenclature in the figure.

Overall, the quality of this paper is there, and I find it is an insightful piece of work for the semiconductor industry. I recommend it for publication.

Author Response

Dear reviewer & editor:

We thank to editor and reviewer’s helpful comments. According to the comments of reviewer, the manuscript is revised. The responses to reviewer comments are attached as follows.

The sample size the author presented is ok to show a general trend. But it would be more convincing if you can use 10 + samples for the same condition and use a t-test to conclude whether the improvement of particle performance is statistically significant. I understand those particle-grade wafers are expensive. It might not be feasible to use a 10+ wafer for each condition. But it would be good to have t-test results for the few keep knobs that the author is claiming to have the best particle reduction performance.

Ans: Thanks to reviewer’s comment. Since the results of this research experiment have been applied to the etching equipment of the 12-inch wafer fab production line for practical verification, it has also received quite good results. We consider the sample size test results to be complete.

In figure 11, pressure. The author listed it as 0mT which is not achievable. Maybe the author should list the actual pressure getting from the manometer?

Ans: Thanks to reviewer’s comment, the content of the article has been corrected to the actual pressure of the manometer is 0.075mTorr.

11, in the first particle map image, there seems to be a cluster of particles that might not be related to the process. It is more likely related to some debris peeling off either from the chamber or transfer module. Maybe the author should exclude samples demonstrating non-process-related particle maps.

Ans: We consider the source of this particle swarm may still come from powder. During the wafer transfer process, the powder is lifted and blown onto the wafer surface by the flow field effect caused by the transfer drive (Ex. push-pin.. etc).

There are spelling mistakes: Fig. 13, a, b. Inter-mitten should be spelled as intermittent. Line 285, high RF should be changed to Top RF to keep it consistent with the nomenclature in the figure.

Ans: Thanks to reviewer’s comment, we revised the manuscript..

Reviewer 2 Report

The paper by Ku et al. presents an investigation of particle contamination during oxide etching in a CCP etch chamber. In this work, the sources of particle contamination within the chamber are identified and the factors affecting particle transport are investigated. Moreover, practical recommendations for reduction of particle contamination are provided. Overall the paper is well structured and clearly written. I think this work would be of interest for the plasma etch community.

The paper can be accepted for publication after the following questions/suggestions are addressed.

I would rename Section 2 (e.g. Basic Principles of Particle transport in plasmas would better reflect the content of the section).
In Section 2.1 it would be appropriate to mention that in RF CCP discharges electrostatic field is localized in the sheath while the plasma bulk is nearly equipotential.
Section 2.3.2. What is the linearized Debye length? I don't understand why Eq (3) only applies when the particle size is smaller than the linearized Debye length. What is the applied electric field E? In CCP, only RF fields are applied.
Section 2.3.2. What does the neutral point mean (line 119)? In the low pressure conditions of the current study, the mean free path of neutral species is between 70 and 6000 micron. Therefore, the neutral drag can not be calculated from hydrodynamic considerations. In this section, correct expressions of molecular drag force should be used.
Section 3. Equation (8) holds when g<<d. In case g>>d, the term k_dg should not be neglected.
Experimental sequence in Section 3 should be better explained, experimental conditions at each step should be clearly reported. I would suggest to use a table with all relevant parameters instead of the text. I didn't understand whether the RF power was applied in N2 or Ar. Is the chamber equipped with only the bottom RF source or both electrodes are independently powered? Fig 2 should be updated to reflect the actual experimental configuration.
Section 4.3. What is the purpose of Eq 9? Lines from 233 to 248 make no sense to me. This part should be rewritten in a more clear way.
Figure 9. The second measurement seems to always yield lower particle count. What is the reason for the overall poor reproducibility between the 1st and the 2nd measurement throughout the paper?
Section 4.4. Eq 10 does not describe the gas viscosity but rather the drag force (see comment 4 above).
Figure 10. What is the source of contamination at 0 sccm 0 mTorr?
Section 4.6, line 292 requires editing.
Figure 15. The particle count in this experiment is much lower than in previous tests. Does this mean that there was no additional powder in the chamber as compared to Sections 4.2 - 4.7? What was the condition of the chamber walls in this experiment? What is the difference in experimental conditions between Figure 15 and Figure 16?
Sections 4.8 - 4.9. A full set of experimental conditions should be presented and not only the factors that are being investigated. Again, what were conditions of the chamber walls? Was there any surface pretreatment? Without the full information the results of the different tests can not be compared to each other.
If would be desirable to present an optimized set of parameters that allows significant reduction of the particle count at given chamber conditions. For example, can we compare Figure 1 and Figure 21 and claim that an improvement of particle count has been achieved by optimizing the process conditions?
The image quality/readability of Figure 4 and Figure 13 should be improved.

Author Response

Dear reviewer & editor:

We thank to editor and reviewer’s helpful comments. According to the comments of reviewer, the manuscript is revised. The responses to reviewer comments are attached as follows.

I would rename Section 2 (e.g. Basic Principles of Particle transport in plasmas would better reflect the content of the section).

Ans: Thanks to reviewer’s comment, we have changed the title of Section 2 to Basic principles of particle transport in plasmas

In Section 2.1 it would be appropriate to mention that in RF CCP discharges electrostatic field is localized in the sheath while the plasma bulk is nearly equipotential.

Ans: Thanks to reviewer’s comment, we revised the manuscript to describe RF CCP discharges electrostatic field in Section 2.

Section 2.3.2. What is the linearized Debye length? I don't understand why Eq (3) only applies when the particle size is smaller than the linearized Debye length. What is the applied electric field E? In CCP, only RF fields are applied.

Ans: Thanks to reviewer’s comment, we revised the manuscript as the Section 2.3.2.

The dust particles in the plasma are negatively charged. The electric force acting on such particles, which causes the trapping of the dust particles in the plasma, the expression is

where is the dust particle charge, and is the applied electric field. The electric force is directed inward (i.e., toward the plasma glow) and traps dust particles within the plasma. The electric field mainly refers to radio frequency plasma and electrostatic field, including radio frequency power and ESC voltage output.

Section 2.3.2. What does the neutral point mean (line 119)? In the low pressure conditions of the current study, the mean free path of neutral species is between 70 and 6000 micron. Therefore, the neutral drag can not be calculated from hydrodynamic considerations. In this section, correct expressions of molecular drag force should be used.

Ans: Thanks to reviewer’s comment, we revised the manuscript as the Section 2.3.3

Gas viscous force causes particles to suspend when they continuously collide against gas molecules. Resuspended particles are accelerated in the direction in which the gas is flowing. The gas viscous force is expressed by the following equation

where is the thermal velocity of a gas molecule, is the mass of the gas molecule, is the relative velocity between a particle and the gas, and is the radius of the particle. Gas-induced particle resuspension is therefore not only due to shock waves, but also to gas viscous forces.

Section 3. Equation (8) holds when g<<d. In case g>>d, the term kdg should not be neglected.

Ans: Thanks to reviewer’s comment, we revised the manuscript.

Experimental sequence in Section 3 should be better explained, experimental conditions at each step should be clearly reported. I would suggest to use a table with all relevant parameters instead of the text. I didn't understand whether the RF power was applied in N2 or Ar. Is the chamber equipped with only the bottom RF source or both electrodes are independently powered? Fig 2 should be updated to reflect the actual experimental configuration.

Ans: Thanks to reviewer’s comment. The relevant parameter table has been completed. As for the RF power, it can be applied to N2 and Ar gases, depending on the requirements of the process recipe. For example, in EUV process, N2 gas is mostly used in wafer chucking step; Argon gas is mostly used in wafer process step. As for the chamber part, there are mainly two bottom RF sources of 40.68MHZ and 12.88MHZ, and the upper electrode is configured with a DC bias system for power supply. We have updated the drawing configuration as shown in Figure 2 and 4.

Section 4.3. What is the purpose of Eq 9? Lines from 233 to 248 make no sense to me. This part should be rewritten in a more clear way.

Ans: Thanks to reviewer’s comment, we revised the manuscript as Section 4.3, the detail as below:

To verify that the ESC voltage caused a large number of particles to be generated on the wafer surface, we sprayed powder on the surface of the edge ring around the ESC to simulate worst-case particle deposition conditions. In the experimental process, the ESC voltage was set to four conditions of 2500 V, 2000 V, 1700 V, and 1500 V to reduce the chucking force of the wafer to analyze and verify the number of particle contamination, as shown in Figure 8. The experimental results are shown in Figure 9. As the setting of the ESC voltage output parameter decreases, the particle contamination on the wafer also reduced. Therefore, the electric field, which varies with the ESC voltage output, is closely related to the number of suspended particles. Therefore, not only the electric field generated by the radio frequency discharge will cause the particles to be suspended. In the case of ESC output voltage, particles also cause a suspension mechanism. On the other hand, in order to verify whether the number of particles is also related to the helium pressure of the ESC voltage, we performed experiments at the same ESC voltage of 2500 V using helium pressures of 0, 15 and 30 Torr. From the particle map results in Figure 10, it is shown that the size of the helium pressure does not affect the number of particles. Therefore, using a lower ESC voltage (i.e. lower chucking force) during wafer processing can reduce particle deposition on the wafer.

Figure 9. The second measurement seems to always yield lower particle count. What is the reason for the overall poor reproducibility between the 1st and the 2nd measurement throughout the paper?

Ans: Thanks to the reviewers for their comments. The main reason is that as the number of experiments increases, the powder may decrease, so the overall particle count will also show fewer particles than the first time.

Section 4.4. Eq 10 does not describe the gas viscosity but rather the drag force (see comment 4 above).

Ans: Thanks to reviewer’s comment, we revised the manuscript.

Figure 10. What is the source of contamination at 0 sccm 0 mTorr?

Ans: Thanks to reviewer’s comment, the main source of contamination of this issue comes from the powder placed around the ESC, through the influence of the moving parts and gas pressure difference during wafer transfer, resulting in the change of the gas flow field, and falling on the wafer.

Section 4.6, line 292 requires editing.

Ans: Thanks to reviewer’s comment, we revised the manuscript.

Figure 15. The particle count in this experiment is much lower than in previous tests. Does this mean that there was no additional powder in the chamber as compared to Sections 4.2 - 4.7? What was the condition of the chamber walls in this experiment? What is the difference in experimental conditions between Figure 15 and Figure 16?

Ans: Thanks to reviewer’s comment. The powder in the chamber decreases as the number of experiments increases. In the experiment of Figure 15, there appears to be no additional powder in the chamber. As for the whole experiment process, the chamber wall was deposited and attached by by-products. The difference between the experimental conditions in Figure 16 and Figure 15 is mainly to clarify the particle contamination caused by the rise and fall of the ESC temperature for analysis and verification.

Sections 4.8 - 4.9. A full set of experimental conditions should be presented and not only the factors that are being investigated. Again, what were conditions of the chamber walls? Was there any surface pretreatment? Without the full information the results of the different tests can not be compared to each other.

Ans: The experimental condition table has been established. There is deposition on the chamber wall during the experiment, and no any surface pretreatment such as Pre-season or dry clean on whole experiment condition.

If would be desirable to present an optimized set of parameters that allows significant reduction of the particle count at given chamber conditions. For example, can we compare Figure 1 and Figure 21 and claim that an improvement of particle count has been achieved by optimizing the process conditions?

Ans: Thanks to reviewer’s comment, we revised the manuscript.

We compared particle map results at 225 RF hours chamber condition from the previous settings and optimized settings, as shown in Figure 23. Particle contamination can be improved after optimizing recipe settings.

The image quality/readability of Figure 4 and Figure 13 should be improved.

Ans: Thanks to reviewer’s comment, we re-inserted the original uncompressed images in Figures.

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript presents interesting results in a logical manner and deserves to be published in its current form.