High-Selectivity Growth of GaN Nanorod Arrays by Liquid-Target Magnetron Sputter Epitaxy

Round 1

Reviewer 1 Report

In this paper, the authors developed an array of GaN (NR) nanorods grown on Si substrates without a catalyst, using liquid-target reactive magnetron sputter epitaxy.

The work presented remains original since, in terms of plagiarism assessment, only about 22% of this document consists of text more or less similar to the content of 62 sources considered most relevant. The largest section with similarities contains 42 words and has a similarity index of 99% with its main source. The similarity index presented above does not in any way indicate the proven presence of plagiarism or a lack of rigour in the document. There may be perfectly legitimate reasons for passages in the analyzed document to be found in the identified sources.

The text is written in a good and comprehensive English. I believe this is an interesting contribution and I recommend for publication in Coating Journal in present form except for this remark: When the authors have converted their manuscript from "word" to "PDF", the phrase "Error! Reference source not found" appears several times in the text. The message typically occurs because a cross-referenced bookmark or heading has been deleted.  The authors have to fix this in their text in order to avoid its appearance again.

 

Comments for author File: Comments.pdf

Author Response

The authors thank the reviewer for the positive comments. We have converted all cross-reference to plain text to prevent the error message from happening.

Reviewer 2 Report

In this study, the application of liquid-target magnetron sputter epitaxy was used to high-selectivity growth of GaN nanorod (NR) arrays. By changing the ratio of N₂ and Ar, the NR shows a strong correlation with P_N2 on the selectivity, coalescence, and growth rate of NR in both radial and axial directions. When the P_N2 ratio between 80-90%, the growth rate of NRs was found to render both high growth rate and high selectivity. After a careful review, I recommend the paper is accepted with some major changes.

 

1. Why does changing the ratio of N₂ and Ar cause these differences? If changing other gas can get similar results or not.
2. Is the thickness of the growing nanotube uniform? the author maybe can describe in a new draft.
3. In the figure 5, the author changes different temperature and various pitches. However, the reaction time are the same or not? If using different reaction time maybe can get different results.

 

Minor typos are spread around the text:

Page 3. line 117. “Error! Reference source not found. shows (a) top-view and (b) 45° bird-eye view SEM”

Page 4. Line 131. “existence of multiple NRs inside a single nanohole, indicated by the red circles in Error! Reference source not found.,”

Page 5. Line 160. “Error! Reference source not found. shows the effect of PN2 on the average diameter of the NRs.”

Page 5. Line 166. “individual NRs within a single opening, as shown in Error! Reference source not found. Increasing”

Page 5. Line 175. “in Error! Reference source not found., complete coalescence within the nanoholes is achieved,”

Page 6. Line 187. “at various pitch are shown in Error! Reference source not found.”

Page 6. Line 195. “grown at two different temperature is given in Error! Reference source not found..”

Page 7. Line 204. “as depicted in Error! Reference source not found.”

Page 7. Line 217. “as shown in Error! Reference source not found.”

Author Response

The authors thank the reviewer for the positive comments. We have converted all cross-reference to plain text to prevent the error message from happening.

1. As discussed in the main text of the manuscript, changing the value of P_N2 and P_Ar changes the Ga/N ratio on substrate surface, resulting in the different growth behaviour. This is described in line 140 in the discussion:

“The effect of P_N2 and P_Ar on the SAG selectivity can be explained in terms of the Ga/N ratio on the substrate surface. Changing P_N2 and P_Ar directly modifies the Ga/N ratio within the chamber, affecting the availability of Ga adatoms and l_Ga [43,44]”

The N₂ gas performs target sputtering while also providing the active nitrogen species, and is necessary for the formation of GaN. The Ar gas has higher mass and performs more effective target sputtering, resulting in more Ga yield while also preventing the formation of nitride layer on the liquid Ga target. Changing P_N2 and P_Ar balances out these two effects. Although this is discussed in detail in reference 37, we add more discussions staring from line 142 to emphasize this effect:

“Increasing P_N2 results in more active nitrogen species, lowering the Ga/N ratio. On the other hand, increasing P_Ar increases the sputtering yield due to the higher atomic mass of Ar and the removal of nitride layer on the target, effectively increasing the Ga/N ratio.”

In principle, Ar can be replaced with other noble gases that does not participate in the chemical reaction. However, currently Ar is the most cost-effective noble gas in terms of price and sputtering yield, which is why most sputtering processes uses Ar in the mixture. As this discussion is beyond the scope of this paper, we decided not to include it in the main text.

 

2. The thickness of the nanorods, referred to as the diameter in the manuscript, is not uniform. The diameter is calculated statistically by using ImageJ image analysis software. We have added the following on line 112.

“To account for the size distribution of the NRs, the diameter was calculated statistically by analyzing the plan-view SEM images with ImageJ software.”

The size distribution is represented by the error bar in figure 3, 4, and 6.

3. The reaction time is the same for each temperature growth. The growth was performed two times at different temperatures to see the effect of growth temperature on the growth behavior. The arrays with various pitches were all patterned on the same substrate, meaning that for a particular growth temperature T_g, all the growth conditions are identical. To clarify, we have added the following in line 106.

“A 5×5 µm² nanohole arrays with a pitch of 100, 200, 300, and 400 nm was patterned using FIBL on a single substrate to ensure identical growth conditions for all pitches”

Reviewer 3 Report

In this work, the authors examine the growth of GaN nanorods on patterned substrate using magnetron sputter epitaxy (MSE). In particular, they correlate the nanorod ensemble morphology on growth parameters and comprehensively explain the results by taking into account adatoms diffusion on the substrate surface.

 

The novelty of this work resides in the use of growth models and concepts previously developed for molecular beam epitaxy (MBE) and adapted here to the case of selective-area growth by MSE. The agreement between the model and the experimental results in remarkable and worth publication. Yet, in my opinion, the authors have missed one important feature in their conceptual analysis: the incubation time. This time describes the delay observed before effective nucleation of GaN on Si while Ga and N radicals are both provided on the Si surface [Fernandez-Garrido et al, Nanolett., 15, 3 (2015)]. Its dependence on the growth temperature could easily account for the observations of the authors in Fig. 6a.

 

 

Consequently, I recommend the publication of this manuscript after MAJOR modifications.

 

General questions/remarks:

- the incubation time of GaN on foreign substrate (Si, graphene) and its effect on selective-area growth has been several time reported in the literature and should be considered as well in this study [Musolino et al, Nanotechnology, 26, 8 (2015), Morassi et al, Cryst. Growth Des. 20, 2 (2020)]. In particular, the late nucleation of GaN compared to the moment at which the atomic fluxes are provided prevents a direct estimation of the vertical growth rate (V) of the nanorods solely based on their final length (L) and of the growth time (t_growth). Indeed, V = L/(t_growth – t_incubation), with t_incubation the growth time. The latter was show to exponentially increase with temperature for self-assembled growth on Si. Thus, the length difference reported in this work in Fig. 6 between the nanorods grown at 970 and 980 degC could be simply taken into account by the existence of an incubation time, as already reported in the case of SAG on Si substrates by MBE. What is the opinion of the authors on this?

- details on the growth of the TiNx mask are missing, also in reference [34] where the authors refer to. The nitridation step of the Ti mask was seen as critical for SAG by MBE should be detailed.

- the size of the holes in the mask are not mentioned as well.

Author Response

The authors thank the reviewer for the valuable feedback.

1. We agree that the increased incubation time of the nanorod may also contribute to the reduced nanorod growth rate. From our previous work, we see that the NRs have a finite incubation (Serban, et al. Rep. 2017, 7, 12701.). Before vertical growth occurs, incubation consists of the formation of wetting layer, 3D nuclei formation, and coalescence of GaN islands.

However, we believe that the main mechanism affecting the growth rate is due to the growth occuring within the desorption-dominated process regime, which we have observed in our previous work (Serban, et al. Thin Solid Films 2018, 660, 950–955.). Furthermore, the increased saturation pitch Ps value at lower temperature indicates longer Ga diffusion length due to longer adatom residence time. While incubation will be delayed, the subsequent vertical growth rate will also be severely affected due to the increased desorption of Ga adatom on the surface. To properly quantify the incubation time, a time-series growth is required for both temperature, which is beyond the scope of this current work. We have added the following paragraph in line 247 to acknowledge the effect of increased incubation time.

“Another consequence of increasing T_G is an increased NR incubation time [17,52]. Prior to vertical growth, SAG NRs follow several incubation steps namely the formation of wetting layer, formation of 3D nuclei, and the coalescence of GaN islands [14]. By increasing the incubation time and delaying the onset of vertical growth, the overall NR growth rate is virtually reduced, which may contribute to the smaller NR dimension at increased temperature.

However, the reduced growth rate is mainly caused by less adatom incorporation, as evidenced by the presence of P_s at 980 °C. The decrease of l_Ga results in less incorporation of Ga adatoms and consequently lower radial growth. The effect of increased desorption rate also reflects to a shorter residence time of Ga adatoms directly impinging on NR top, leading to a shorter NR growth.”

 

2. Unlike in MBE-based SAG growth where the TiNx mask is obtained by exposing a Ti layer to nitrogen plasma, we directly deposited a TiNx mask layer using reactive sputtering followed by patterning using FIB. We have added the following line in the experimental section (line 83) for clarity:

“TiN_x was deposited through reactive sputtering of a Ti target under pure nitrogen ambient at room temperature and 20 mTorr pressure”

3. The nanoholes are 30 nm in diameter. We have added the following in the experimental section (line 87) for clarity:

“A low milling current of 2 pA and short milling time of 5 s were used for to create an opening which consists of a 8×6 nanohole array with 30 nm hole diameter”

Round 2

Reviewer 2 Report

The revised draft is an improvement on that submitted initially and is now acceptable for publication