Effect of Nitrogen Flow Ratio on Degradation Behaviors and Failure of Magnetron Sputter Deposited Tantalum Nitride

Round 1

Reviewer 1 Report

1) Some sentences (indicated in yellow in the attached document) must be rewritten with care, as English is really incomprehensible. Additionnal mispelling is also indicated.

2) Some questions and suggestions:

• Table 1: Columns 2, 4, 5 and 6 are unnecessary. These parameters which values do not change should be indicated in the text. Film thickness, roughness, and cristallite/grain sizes and adhesion force should be reported relative to gas flow rate in this table.
• The total gas pressure is kept constant (only the nitrogen flow rate varies). Why does the mean free path of particles decrease while the pressure remains constant?
• AFM topographical images (2D, not 3D)  should be shown with a common z-scale. We could thus see the grain shape.
• What do you mean with Ta hole in the film (page 6, line 187) ?
• It is not clear how you can state that "the conductivity type changes from electron conduction to hole conduction"? The sentence is not clear either.
• Finally, which is the good nitrogen flow rate? It is not stated in the conclusion.

Author Response

To Reviewer 1:

 

1.Q: Some sentences (indicated in yellow in the attached document) must be rewritten with care, as English is really incomprehensible. Additional misspelling is also indicated.

A: We revise these sentences and misspellings in the new manuscript with track changes. Please see the attachment. It is our new manuscript.

 

2.Q: Table 1: Columns 2, 4, 5 and 6 are unnecessary. These parameters which values do not change should be indicated in the text. Film thickness, roughness, and cristallite/grain sizes and adhesion force should be reported relative to gas flow rate in this table.

A: The new Table 1 is rewritten as below:

Table 1. Synthesis Conditions of TaN Films.

Experiment	Film thickness	Flow Ratio/%	Roughness/nm	Grain sizes/nm	Adhesion force/N
1	220.20	2.5	1.10	0.48	75.40
2	186.00	5	1.03	0.24	59.40
3	157.80	10	0.68	0.36	11.25
4	150.00	15	0.56	0.36	10.20
5	139.80	20	0.51	0.36	12.10
6	130.20	25	0.58	0.24	15.00

3.Q: The total gas pressure is kept constant (only the nitrogen flow rate varies). Why does the mean free path of particles decrease while the pressure remains constant?

A: In our opinion, when the nitrogen flow rate increases, there will be more nitrogen molecules between the target and the substrate. Thus, a Ta atom collides with more N atoms before reaching the substrate from the target. That is why we state that the mean free path of particles deceases.

 

4.Q: AFM topographical images (2D, not 3D) should be shown with a common z-scale. We could thus see the grain shape.

A: We regret that we cannot provide 2D images of AFM unless the samples could be re-tested, which is currently impossible to achieve.

 

5.Q: What do you mean with Ta hole in the film (page 6, line 187)?

A: Actually, we refer to the Ta vacancies in the films.

 

6.Q: It is not clear how you can state that "the conductivity type changes from electron conduction to hole conduction"? The sentence is not clear either.

A: With the increase of nitrogen rate, it is observed that the conduction of the films getting low. In our opinion, when the nitrogen rate is low, the films have higher Ta content, so they conduct in a way more like metals; but as the nitrogen rate gets high, Ta content in the films become less, and they conduct like semiconductors.

 

7.Q: Finally, which is the good nitrogen flow rate? It is not stated in the conclusion.

A: We think the good nitrogen flow rate is 2.5%, under which the conduction of the films is relatively higher, the adherence force larger, and deposition rate higher.

Author Response File: Author Response.docx

Reviewer 2 Report

It is recommended to provide deeper analysis. Detailed comments are given in bellow:

Effect of Nitrogen Flow Ratio on Degradation Behaviors and Failure of Magnetron Sputter Deposited Tantalum Nitride

from Li Zhigang, Zhang Yubao, Wang Yi, Li Jinfeng and Zhao Hongtao.

Abstract:

• one the main noticed functional properties to optimize is R_sh which is not material characteristic, please use resistivity;
• do not interchange present and past indefinite time (valid for the whole manuscript);

Introduction:

• line 36: why magnetron sputtering helps “to overcome the disadvantages of the high melting point of ceramic films”? there are many other methods for deposition of oxide films (MOCVD, ALD, PLD etc.)
• line 43: “… different the nature.”, better without article
• line 46: “There are still many issues…” – what are these issues? Why the adhesion of TiN on oxide substrate is so important?
• line 48: why exactly Al₂O₃ substrate, and not TiO₂, quartz SiO₂ or even glass?

Materials and Methods:

-line 58: very few details are given for the Al₂O₃ substrate, is it polycrystalline? is it polished? What is the roughness?

-line 60: what was the purity of the gas mixture? How the N₂ content was set up?

-line 65: which XRD system was used? Which measurement conditions were applied?

-line 66: no details for the AFM device and measurement conditions.

Results:

- line 71 and Fig.1: XRD does not give a spectrum! It is a pattern;

- line 72: how one ordinary scientist “can see” that we deal with the fcc-TaN? Which standard can prove it?

- line 73: do not note any difference in texture if you do not calculate texture coefficients;

- line 75: “week” -> “weak”;

- line 78-79: what does the sentence “and the position of the layer is caused by the development of compressive stress” mean? The existence of compressive stresses should be proven but not just stated because of the peak shift.

- Fig.1: where are the reflexes from the Al₂O₃ substrate? Why only (200)-TaN reflexes shift with changing N₂-content in a sputtering gas?

- part 3.2_AFM: very many speculations based on very small differing roughness data. Roughness does also depend on film thickness which decreases almost by a half with rising N₂-flow (Fig. 3 plus same deposition time).

- part 3.4_adhesion: this part is also rather speculative, because the evidence of stresses in the films was not given and authors do not compare energy of metal and nitrogen species in relation to the plasma conditions. It is recommended to revise this part and extend with variation of e.g. bias voltage.

- Fig. 5: As noticed above, one may not compare R_sh if film thickness differs so much.

- It is recommended to present resistivity values and give some comments and/or measurements of band-gap if exists and concentration / mobility of electrons.

Author Response

To Reviewer 2:

1.Q: one the main noticed functional properties to optimize is R_sh which is not material characteristic, please use resistivity.

A: The thickness of the TaN films varies from 130 to 220 nm. They are more suitable to be characterized by sheet resistance.

 

2.Q: Do not interchange present and past indefinite time (valid for the whole manuscript);

A: We agree with this suggestion, so we check the whole manuscript and revise it. Please see the attachment. It is the new manuscript.

 

3.Q: line 36: why magnetron sputtering helps “to overcome the disadvantages of the high melting point of ceramic films”? there are many other methods for deposition of oxide films (MOCVD, ALD, PLD etc.)

A: By reactive sputtering, TaN films are successfully fabricated. Besides magnetron sputtering, MOCVD, ALD, PLD are suitable methods for deposition of TaN films. This article continues our previous work, focusing on the influence of working pressure on the properties of TaN films.

 

4.Q: line 43: “… different the nature.”, better without article

A: We alter “nature” to “properties”.

 

5.Q: line 46: “There are still many issues…” – what are these issues? Why the adhesion of TaN on oxide substrate is so important?

A: Here, “issues” refer to the content discussed in this article, such as the influence of film microstructure, roughness, deposition rate, film-base bonding force, and electrical properties. The bonding force between the TaN film and the Al₂O₃ substrate determines the failure of the film and the device based on the film. TaN deposited on Al₂O₃ substrate can be used as a high temperature resistant substrate material for practical applications, which has an important application background. In addition, we believe that the binding force is related to the deposition rate of the film.

 

6.Q: line 48: why exactly Al₂O₃ substrate, and not TiO₂, quartz SiO₂ or even glass?

A: The use of Al₂O₃ substrate is a continuation of our previous work and applied to specific high-temperature devices.

 

7.Q: line 58: very few details are given for the Al₂O₃ substrate, is it polycrystalline? is it polished? What is the roughness?

A: The Al₂O₃ substrate was glazed after high-temperature sintering, without further polishing, and no roughness test was done.

 

8.Q: line 60: what was the purity of the gas mixture? How the N₂ content was set up?

A: The purity of the gas mixture is 99.99%. The N₂ content was set up by a rotameter. The relevant description is added to the paper.

 

9.Q: line 65: which XRD system was used? Which measurement conditions were applied?

A: The instrument information and measurement conditions are now list in the paper.

 

10.Q: line 66: no details for the AFM device and measurement conditions.

A: The instrument information and measurement conditions are now list in the paper.

 

11.Q: line 71 and Fig.1: XRD does not give a spectrum! It is a pattern.

A: Both “Spectrum”s are altered to “pattern”.

 

12.Q: line 73: do not note any difference in texture if you do not calculate texture coefficients.

A: Accept. We'll omit the statements.

13.Q: line 75: “week” -> “weak”

A: We re-correct this misspelling.

 

14:Q: line 78-79: what does the sentence “and the position of the layer is caused by the development of compressive stress” mean? The existence of compressive stresses should be proven but not just stated because of the peak shift.

A: We meant that the compressive stress arising caused the peak shift according to Reference [3]. 

 

15.Q: Fig.1: where are the reflexes from the Al₂O₃ substrate? Why only (200)-TaN reflexes shift with changing N₂-content in a sputtering gas?

A: The Al₂O₃ substrate was glazed after high-temperature sintering. Due to the formation of an amorphous state, there is no XRD peak.

Actually, it is the (111) shifts to lower degree with the increasing of nitrogen flow ratio, and the descend of peak strength suggests that the films are losing their crystalline structure and becoming more amorphous

 

 

16.Q: part 3.2_AFM: very many speculations based on very small differing roughness data. Roughness does also depend on film thickness which decreases almost by a half with rising N2-flow (Fig. 3 plus same deposition time).

A: In this paper, AFM image was used to demonstrate the surface state of TaN film but not very focus on the roughness.

17.Q: part 3.4_adhesion: this part is also rather speculative, because the evidence of stresses in the films was not given and authors do not compare energy of metal and nitrogen species in relation to the plasma conditions. It is recommended to revise this part and extend with variation of e.g. bias voltage.

 

A: We take the suggestion into account, but in our experiment we did not change the bias voltage.

18.Q: Fig. 5: As noticed above, one may not compare R_sh if film thickness differs so much.

A: Yes. we have made a revision about this part.

19.Q: It is recommended to present resistivity values and give some comments and/or measurements of band-gap if exists and concentration / mobility of electrons.

A: Thank you for the recommendation. For the TaN film in this paper, we pay more attentation on the square resistance, so we had only measured it. For band-gap and concentration/mobility of electrons, it’s not the key point in our research, we will do more research in details about the electric properties in the coming one. 

 

 

Author Response File: Author Response.docx

Reviewer 3 Report

The authors report about the effect of Nitrogen flow ratio on several properties of Tantalum Nitride films deposited by Magnetron sputtering.

The paper, based on experimental data obtained from several techniques, is interesting as it provides evidences of the role of Nitrogen in shaping the properties of Tantalum Nitride Films. However, several issues should be addressed before publication.

 

• The paper is entirely based on experimental data. Therefore, sufficient technical details should be provided about the different instruments used (XRD, AFM…) and their characteristics (brand and model, operating mode, etc.)
• The presentation of the different values of surface roughness, resulting from AFM data (section 3.2) should be given as a table or as a graph. The present form (6 lines of text detailing the correspondence between nitrogen flow rate and roughness) is not very effective.
• First sentence in section 3.4. The importance of the bonding strength is presented as the most important property of the film, yet this aspect clearly depends on the applications envisioned for this kind of film. Either the author should provide a brief motivation of the sentence, or they should point out that the relevance of a particular property depends on the target application.
• Labels of Fig 1. The labels concerning TaN(111) and TaN(200) should be centered on the respective lines. The 10% nitrogen label should be reported in bold, as done for all the other % values.
• The authors report for three times target “poisoning” as the cause for some experimental observation. A clear explanation of this phenomenon should be provided to help the reader better understand this point.
• English should be polished throughout the manuscript. A few sentences are difficult to read. The manuscript should be checked for typos (e.g. Tatio -> Ratio).

Author Response

To Reviewer 3:

1.Q: The paper is entirely based on experimental data. Therefore, sufficient technical details should be provided about the different instruments used (XRD, AFM…) and their characteristics (brand and model, operating mode, etc.)

A: In the new manuscript we give more instruments information and measurement conditions. Please see the attachment. It is our new manuscript.

 

2.Q: The presentation of the different values of surface roughness, resulting from AFM data (section 3.2) should be given as a table or as a graph. The present form (6 lines of text detailing the correspondence between nitrogen flow rate and roughness) is not very effective.

A: In the new manuscript we list Film thickness, roughness, grain sizes and adhesion force, while delete columns 2, 4, 5 and 6.

 

3.Q: First sentence in section 3.4. The importance of the bonding strength is presented as the most important property of the film, yet this aspect clearly depends on the applications envisioned for this kind of film. Either the author should provide a brief motivation of the sentence, or they should point out that the relevance of a particular property depends on the target application.

A: Yes, we have add a sentence to explain the applications background.

 

4.Q: Labels of Fig 1. The labels concerning TaN(111) and TaN(200) should be centered on the respective lines. The 10% nitrogen label should be reported in bold, as done for all the other % values.

A: We re-correct these mistakes.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Several sentences (indicated in yellow in the attached document) must be rewritten with care. Spelling errors are also indicated. This was already mentioned in my first report and it has not yet been fully taken into account. In addition, there are a few questions in the attached file, which should be treated with care. 

In the abstract, the authors mention the influence of nitrogen flux on the TCR values of the films. But no TCR value appears in the text. The correct nitrogen flow is still not indicated in the conclusion. These points must be considered with care. 

Comments for author File: Comments.pdf

Author Response

To Reviewer 1:

Thank you for all suggestions. We upload a  new manuscript. Please see the attachment. 

Revision details are list below:

Abstract:

Line 10, 11, and 13:

We do not provide any data concerning TCR, so it is not necessary to retain the words or phrases.

 

Line 11:

Add a comma between “resistance” and “TaN films”.

 

Line 16:

We change “film-based bond strength” to “adhesion force”.

 

Line 17:

We change “While” to “And”.

 

Introduction:

Line 48~50:

We correct all “film” and “structure” in their plural form.

 

Materials and Methods:

Line 54:

We delete “as the substrate”.

 

Line 58:

We correct “structure” in their plural form.

 

Line 63:

We correct the last sentence.

 

Table1.

We put units in Column 5 and 6.

 

Line 71~74:

We simplify the description of AFM instrument.

 

Line 85~87:

We explain the reason for the shifts of diffraction peaks of higher N₂ flow rate.

 

Line 97:

We correct “ratio” in its plural form.

 

Line 114:

We rewrite the sentence.

 

Line 117:

We change the “membrane” to “thin films”.

 

Fig. 2.:

We rewrite the caption in which N₂ flow ratios are indicated.

 

Line 139~142:

We rewrite this sentence.

 

Line 152:

We add a space between “Figure 4” and “shows”.

 

Line 153:

We correct “film” in its plural form.

 

Line 156:

We add “increases” after “slightly”.

 

Line 167~168:

We rewrite the sentence.

 

Line 178:

We delete the repetitive words.

 

Figure 4:

We replace “and” with “on” in the caption.

 

Line 184:

We delete “and the substrate”.

 

Figure 5:

We replace “currency” with “ratio”.

Author Response File: Author Response.docx

Reviewer 2 Report

See the attached file.

Comments for author File: Comments.docx

Author Response

To Reviewer 2:

Thank you for all the suggestions. We upload a new manuscript. 

1. We replace all “square” with “sheet”. We give the size of substrates and films in Methods.

 

2. We add labels and ticks of the axis of intensity in the XRD pattern (Fig. 1).

 

3. Line 86~88:

We rewrite the sentence.

 

4. We add a new table (Table 2) to list XRD peak positions. Not only TaN (111) peak, but also other peaks shifts to lower degree with the N₂ flow rate ratio increase. We try to calculate stress approximately by sin²Ψ method. The residual stress is compressive when N₂ flow ratio is 2.5%, and turn to tensile since the ratio increase to 5% and larger. The details is shown in the coverletter.

Author Response File: Author Response.docx