Epitaxial Lateral Overgrowth of {11-22} InGaN Layers Using Patterned InGaN Template and Improvement of Optical Properties from Multiple Quantum Wells

Round 1

Reviewer 1 Report

The authors studied the relaxed semipolar (11-22) InGaN template on ELOG-GaN and ELOG-InGaN. The topic itself might be interesting to researchers in this field, but the writing and presentation is difficult to the readers. Spell and grammatical errors are too much to list one by one. The editing before submission is careless. For example: line 133 to 137 is a replica of line 89 to line 92; line 169 to 172 is also a replica of line 89 to line 92. How many times do the readers need to refer to the figure 1?

Besides, for a study focusing on relaxed InGaN template, how the composition and degree of relaxation is determined via RSM is critical information. However, the reviewer didn't see any RSM data neither in the manuscript  nor supplemental. Because the manuscript is seriously flawed from many aspects, the reviewer suggested a rejection. 

Author Response

Thank you very much for the detailed peer review. We have carefully revised the manuscript to address your comments. We have also reviewed the manuscript for grammatical and vocabulary errors. Below are the responses to each reviewer's comment.

The authors studied the relaxed semipolar (11-22) InGaN template on ELOG-GaN and ELOG-InGaN. The topic itself might be interesting to researchers in this field, but the writing and presentation is difficult to the readers. Spell and grammatical errors are too much to list one by one. The editing before submission is careless. For example: line 133 to 137 is a replica of line 89 to line 92; line 169 to 172 is also a replica of line 89 to line 92. How many times do the readers need to refer to the figure 1?

Besides, for a study focusing on relaxed InGaN template, how the composition and degree of relaxation is determined via RSM is critical information. However, the reviewer didn't see any RSM data neither in the manuscript nor supplemental. Because the manuscript is seriously flawed from many aspects, the reviewer suggested a rejection. 

Response:

We apologize for the errors. We have removed the repeated lines from sections 3.2 and 3.3.

We did not include the relaxation mechanism originally because we used references to previous studies to explain the relaxation mechanism. As per your suggestion, we have now added RSM data on the relaxed InGaN template in the Supplementary file, and the method of investigating the relaxation mechanism is added on page 3, line 104 of the revised manuscript.

The In composition can be obtained from the Qy value in RSM, but in this study, we determined the composition from the XRD result using Vegard’s law because all the samples were completely relaxed. This information is given on page 4, line 137 of the revised manuscript.

In addition, spelling and grammatical errors have been checked and corrected by native English speakers.

Reviewer 2 Report

Dear Editor, Dear Authors,

The manuscript presents an interesting proposition of improvement of InGaN template preparation in non-polar direction. The idea is very promising and the results are convincing. I would recommend the paper for publication after corrections mentioned below.

Corrections and comments:
1) Please provide structure of the diodes, and quantum wells. It is important to know whether there was any buffer. What were widths and composition of wells and barriers?

2) Line 85: PL spectra were excited with laser and acquired with some spectrometer not with a laser.

3) Lines 89 –90, grammatical error:
Instead of
"First, the InGaN layers grown on the {11-22} GaN templates with different In contents were characterized at growth temperatures from 700 to 850°C. "
It is better to write:
"First, the InGaN layers grown on the {11-22} GaN templates with different In contents obtained at temperatures from 700 to 850°C were characterized."
The same sentence is copy – pasted to lines 133 – 134 and 169 – 170 with the same error.

4) Growth temperature should be mentioned for samples in figure 6.

5) What was the growth time for overgrowth of layers on GaN and InGaN substrates? Was it the same?

6) Peak at 470 nm in figure 7 is obviously due to some defects. What is your hypothesis for it?

7) Line 246:
Instead of:
"MQWs fabricated on GaN and ELO-InGaN templates at RT."
Should be:
"MQWs fabricated on GaN and ELO-InGaN templates measured at RT."

8) It is very nice to see red LEDs made of InGaN but up to now such LEDs have very low efficiency. What is efficiency of your devices compared to commercially available AlGaInAs diodes?

Author Response

Thank you very much for the detailed peer review. We have carefully revised the manuscript to address your comments. We have also reviewed the manuscript for grammatical and vocabulary errors. Below are the responses to each reviewer's comment.

Dear Editor, Dear Authors,

The manuscript presents an interesting proposition of improvement of InGaN template preparation in non-polar direction. The idea is very promising and the results are convincing. I would recommend the paper for publication after corrections mentioned below.

Corrections and comments:
1) Please provide structure of the diodes, and quantum wells. It is important to know whether there was any buffer. What were widths and composition of wells and barriers?
Response:

Thank you for your comments on our manuscript. A buffer layer is important, but this LED did not have any buffer layer, such as the InGaN/GaN superlattice. The thicknesses of the GaN barrier and InGaN well layers were 16 nm and 3 nm, respectively. We calculated the In content of the InGaN well layer using the peaks present in each PL spectrum. We employed the modified Vegard’s law, including the linear interpolation and quadratic term based on a bowing parameter, obtained from previous studies. The In composition of the InGaN well layers in the InGaN/GaN MQWs on the GaN template and InGaN layers with In contents of 2% and 6% were estimated to be 21.9%, 23.0%, and 25.0%, respectively. We have added the following sentences on page 3, line 83 and page 9, line 269 of the revised manuscript:

→page 3, line 83

For the growth of the MQWs, the thicknesses of the GaN (or InGaN) barrier layer and InGaN well layer were 16 nm and 3 nm, respectively. A buffer layer, such as the InGaN/GaN superlattice, is important to improve the LED performance, but there was no buffer layer beneath the MQWs.

→page 9, line 269

We also calculated the In content of the InGaN well layer based on the peaks present in each PL spectrum. We employed the modified Vegard’s law, including the linear interpolation and quadratic term based on a bowing parameter, obtained from previous studies.^47,48 The In contents of the InGaN well layers in the InGaN/GaN MQWs on the GaN template and InGaN layers with In contents of 2% and 6% were estimated to be 21.9%, 23.0%, and 25.0%, respectively.

2) Line 85: PL spectra were excited with laser and acquired with some spectrometer not with a laser.
Response:

Thank you for pointing this out. We have revised the text on page 3, line 89 as follows:

Photoluminescence (PL) spectra were excited using a He-Cd laser at room temperature (RT), and the spectra were recorded by a spectrometer.

3) Lines 89 –90, grammatical error:
Instead of
"First, the InGaN layers grown on the {11-22} GaN templates with different In contents were characterized at growth temperatures from 700 to 850 °C. "
It is better to write:
"First, the InGaN layers grown on the {11-22} GaN templates with different In contents obtained at temperatures from 700 to 850 °C were characterized."
The same sentence is copy – pasted to lines 133 – 134 and 169 – 170 with the same error.
Response:

Thank you for pointing this out. As per your suggestion, we have revised the sentence on page 3, line 93 of the revised manuscript. Following is the revised sentence:

First, we characterized the InGaN layers grown on the {11-22} GaN templates, which were obtained at growth temperatures from 700 to 850 °C, with different In contents.

We apologize for the error and have removed the repeated lines from sections 3.2 and 3.3.

4) Growth temperature should be mentioned for samples in figure 6.
Response:

The ELO sample was grown at a temperature of 800 °C. We have added the value of growth temperature in the revised manuscript on page 7, line 191 and captions of figures 4 and 6, as shown below.

A growth temperature of 800 °C was used for the ELO of the InGaN layer. The growth conditions, including the growth time, were identical to those of the trench-patterned {11-22} GaN template.

Figure 4. Cross-sectional and plan-view SEM images of the {11-22} InGaN layers grown on the GaN template at a growth temperature of 800 °C, at the initial stage of growth (a) and (b), and after coalescence (c) and (d).

Figure 6. Cross-sectional and plan-view SEM images of the {11-22} InGaN layers grown on the InGaN template at a growth temperature of 800 °C, at the initial stage of growth (a) and (b), and after coalescence (c) and (d).

5) What was the growth time for overgrowth of layers on GaN and InGaN substrates? Was it the same?
Response: The growth conditions of ELO-InGaN were similar to those of the GaN and InGaN templates. The following sentence has been added in the revised manuscript on page 7, line 192:

The growth conditions, including the growth time, were identical to those of the trench-patterned {11-22} GaN template.

6) Peak at 470 nm in figure 7 is obviously due to some defects. What is your hypothesis for it?
Response:

The origin of this peak is not completely known. We observed that by changing the composition of the InGaN template, the main deep-level emission was observed at a wavelength corresponding to the gap between the near-band edge and yellow luminescence of GaN, but the deep emission from ELO was observed at a much shorter wavelength. Investigation of the origin of these peaks is a subject for future work. We believe that this emission is not due to (11-22) but could be caused by a plane that appears at ELO, which is neither the c-plane nor (11-22), because it is more prominent at ELO. We have provided the data for deep emissions from InGaN and GaN in the supplementary file. The following sentence has been added in the revised manuscript on page 9, line 249:

The PL spectrum of the ELO-InGaN layer was composed of two peaks at 415 and 470 nm. The peak at 470 nm is considered to be a deep level emission; however, the 470 nm peak is present at a smaller wavelength than the equivalent yellow luminescence observed in GaN. We believe that this emission is not due to (11-22) but could be caused by a plane that appears at ELO, which is neither the c-plane nor (11-22), because it is more prominent at ELO. The mechanism has not yet been fully understood and further investigation should be conducted.

7) Line 246:
Instead of:
"MQWs fabricated on GaN and ELO-InGaN templates at RT."
Should be:
"MQWs fabricated on GaN and ELO-InGaN templates measured at RT."
Response:

As per your suggestion, we have revised the sentence on page 9, line 261 of the revised manuscript.

Figure 8 shows the RT-PL spectra of the InGaN/GaN MQWs grown on the GaN and ELO-InGaN templates.

8) It is very nice to see red LEDs made of InGaN but up to now such LEDs have very low efficiency. What is efficiency of your devices compared to commercially available AlGaInAs diodes?

Response: Thank you for the comment. We have not yet fabricated the LED, so cannot provide the value of efficiency for comparison with AlGaInAs diodes. Instead, we have discussed the properties of nitride-based LEDs in the introduction section. The highest reported efficiency of nitride-based red LEDs is approximately 5%. In the introduction section, I have added recent data and references related to the EQEs of red LEDs. The following text has been added on page 2, line 52:

The EQE of nitride-based red LEDs is significantly lower than those of the blue and green LEDs; the highest EQE of red LED is 4.5%.^{1, 32, 33} Previously, EQEs of 1% and 3% were obtained by Eu doping and using InGaN underlayers, respectively.^34,35 The efficiency of red LED is gradually increasing but further improvement is still needed. 

Reviewer 3 Report

The paper report the growth and characterization of thick, completely relaxed {11-22} oriented InGaN layers using epitaxial lateral overgrowth (ELO). Although it was difficult to grow ELO- InGaN layers on patterned GaN templates, we succeeded in growing ELO-InGaN layers on a patterned InGaN template. The full width at half maximum of the X-ray rocking curve of ELO-InGaN on the InGaN templates was less than that of non-ELO InGaN. The photoluminescence intensity of InGaN/GaN multiple quantum wells on ELO-InGaN was approximately five times stronger than that on the {11-22} GaN template. I recommend to accept this manuscript after major revisions. The points to be addressed are as follows:

1.  First, the InGaN layers grown on the {11-22} GaN templates with different In contents were characterized at growth temperatures from 700 to 850 °C. Figure 1 shows the images of the epilayer surfaces obtained by differential interference contrast microscopy (Nomarski microscopy). When the growth tem" is repeated three times in this manuscript and should be examined carefully.

Author Response

Thank you very much for the detailed peer review. We have carefully revised the manuscript to address your comments. We have also reviewed the manuscript for grammatical and vocabulary errors. Below are the responses to each reviewer's comment.

The paper report the growth and characterization of thick, completely relaxed {11-22} oriented InGaN layers using epitaxial lateral overgrowth (ELO). Although it was difficult to grow ELO- InGaN layers on patterned GaN templates, we succeeded in growing ELO-InGaN layers on a patterned InGaN template. The full width at half maximum of the X-ray rocking curve of ELO-InGaN on the InGaN templates was less than that of non-ELO InGaN. The photoluminescence intensity of InGaN/GaN multiple quantum wells on ELO-InGaN was approximately five times stronger than that on the {11-22} GaN template. I recommend to accept this manuscript after major revisions. The points to be addressed are as follows:

1. First, the InGaN layers grown on the {11-22} GaN templates with different In contents were characterized at growth temperatures from 700 to 850 °C. Figure 1 shows the images of the epilayer surfaces obtained by differential interference contrast microscopy (Nomarski microscopy). When the growth tem" is repeated three times in this manuscript and should be examined carefully.

Response: We apologize for the oversight. We have removed the repeated lines from sections 3.2 and 3.3.

In addition, spelling and grammatical errors have been checked and corrected by native English speakers.

Round 2

Reviewer 1 Report

See attached file.

Comments for author File: Comments.docx

Author Response

Dear reviewer:

We wish to re-submit the manuscript titled “Epitaxial lateral overgrowth of {11-22} InGaN layers using patterned InGaN template and improvement of optical properties from multiple quantum wells.”

We thank you and the reviewers for your thoughtful suggestions and insights. The manuscript has benefited from these insightful suggestions. I look forward to working with you and the reviewers to move this manuscript closer to publication in the Crystals.

The manuscript has been rechecked and the necessary changes have been made in accordance with the reviewers’ suggestions. We have also reviewed and corrected grammatical and vocabulary errors. The responses to all comments have been prepared and attached herewith.

Thank you for your consideration. I look forward to hearing from you.

Sincerely,

Narihito Okada

Yamaguchi University

2-16-1 Tokiwadai Ube, Yamaguchi 755-8611, Japan

Tel: +81-836-85-9411

Email: nokada@yamaguchi-u.ac.jp

Dear editorial office of MDPI Crystal,

The authors demonstrated epitaxial lateral over-growth (ELOG) on semipolar (11-22) patterned GaN and InGaN. The improvement is supported by the X-ray rocking curve FWHM, SEM observation, and PL intensity of InGaN quantum wells on different InGaN templates. The difference and mechanism of regrowth on GaN or InGaN is properly addressed. The manuscript quality is much improved after an extensive revision. Below are few questions that the author shall still address before the manuscript is ready to publish.

1. The crystal quality of regrown InGaN is described by the FWHM’s of symmetric XRC along different orientation, but the quality of underlayer GaN is described by the threading dislocation density and stacking fault density. The author shall provide one common index to all layers so readers can compare the crystal quality between layers. The reviewer suggests to list the corresponding FWHMs of semipolar GaN in the experimental section.

Reply:

We thank the reviewer for the helpful suggestion. The XRC-FWHM of the GaN template was previously mentioned in the text; we have now added it to the experimental section.

Moreover, we have revised the manuscript in accordance with the reviewer’s suggestion as follows:

➞Page2 Line 73,

The full widths at half maxima (FWHMs) of the X-ray rocking curves (XRCs) for the symmetric 11-22 reflection with the azimuth of the X-ray parallel to the <10-10> and <11-23> directions of the GaN template were 1679 and 1116 arcsec, respectively.

➞Page7 Line210

The XRC-FWHMs for the symmetric 11-22 reflection with the azimuth of the X-ray parallel to the <10-10> and <11-23> directions of the sample were 1328 and 796 arcsec, respectively. These values are lower than those of the materials grown on the GaN template (1679 and 1116 arcsec, respectively).

2. In page 7, line 192-193, the author revealed the InGaN regrowth temperature is 800 ^o Is the growth temperature of the trenched base layer the same? What’s the thickness of final coalesced InGaN template?

Reply:

We thank the reviewer for this question. Yes, the growth conditions of the regrown-InGaN and base layers were the same. We have accordingly revised the manuscript to include these specifics.

➞Page7 Line194

For the regrown InGaN layer, the film thicknesses were 3.6 μm and 2.5 μm from the trench top and bottom to the surface, respectively.

➞Page7 Line210

3. Figure 8 showed PL of the same quantum well on different ELO-InGaN. Do the two ELO-InGaN with different composition have the same regrowth thickness? Are they regrown on the same patterned InGaN base or different ones? Please specify.

Reply:

We thank the reviewer for the questions. The two ELO-InGaN layers were grown at different temperatures while the other parameters were identical. The growth temperatures of these InGaN layers were the same for the base and regrowth layers; for the ELO-InGaN layers with 2% and 6% In composition, the growth temperatures were 850 °C and 800 °C, respectively. We have revised the manuscript to include these details.

Both the trench-patterned InGaN base layers had the same structure as in Fig. 6. The thicknesses of both the regrown ELO-InGaN layers were almost the same. In other words, the two InGaN templates are considered to be equivalent in performance, except for the In composition.

➞Page9 Line264

4. The In composition dependence on growth temperature in figure 2 and in supplemental materials is inconsistent. Besides, there’s no correspondence mark in the manuscript of the supplemental materials. The purpose of the supplemental material seems to be unclear.

Reply:

We thank the reviewer for this comment. The reason for the difference in composition is that the sample in the supplemental material and the sample used in this study were fabricated at different times. We had fabricated many InGaN templates in the past and have confirmed full relaxation in all of them. Therefore, we did not perform reciprocal lattice space mapping in this measurement. In this study, we used data obtained from other experiments as a proof that the InGaN layers are fully relaxed. No revisions have been made to the manuscript in this case.

5. Page 2, line 51, it shall be ScAlMgO₄.

Reply:

Thank you for pointing that out. We have revised the word to ScAlMgO₄.

➞Page2 Line51