Monolayer-Scale GaN/AlN Multiple Quantum Wells for High Power e-Beam Pumped UV-Emitters in the 240–270 nm Spectral Range

Round 1

Reviewer 1 Report

Dear Editor,

the manuscript entitled "Monolayer-scale GaN/AlN multiple quantum wells for high power e-beam pumped UV-emitters in the 240-270 nm spectral range" by Valentin Jmerik et al. reports about the production of multiple quantum wells with monolayer thickness precision for UV emissive cathodoluminescence application.
The submitted work is well structured and reports all the aspects regarding the production of the layered systems, the surface structures and their cathodoluminescence properties. All the presented experimental data are thoroughly discussed, thus making this manuscript scientifically sound.
As a consequence, I would recommend the Editor to publish this work, after the authors address the minor issues listed below:
1.page4, lines167-169: "In accordance with the classical mechanism of step-flow growth described by Burton-Cabrera and Frank (BCF) the radius of curvature (??) of these hillocks is determined by the expression ??~1 [??×??(?/?0)], [...] ". Where do the authors show that their experimental data agree with the theoretical predictions made by the BCF equation?
2.In paragraph 3.6.1 the authors discuss the blue shift of the CL peaks upon lowering the thickness of the GaN layers by dividing them into two subgroups: peaks 1-4 and peaks 5-7. Despite that, the discussion on the two subgroups is mostly analogous leading to the same conclusion that the narrower the GaN layer, the more energetic is the CL peak. Which is the reason why the authors chose to split the CL peaks in 2 subgroups?
3.The authors should briefly comment the shape of the CL peaks #2 and #3, featuring multiple peaks.

4.several typos can be found throughout the text. Acronyms should always be explicitly reported in their extended form the first time they are reported in the text. Italics letters (such as "n-" and "p-type") should be evenly reported in the whole work. All numbers should have their units.

Author Response

1.page4, lines167-169: "In accordance with the classical mechanism of step-flow growth described by Burton-Cabrera and Frank (BCF) the radius of curvature (??) of these hillocks is determined by the expression ??~1 [??×??(?/?0)], [...] ". Where do the authors show that their experimental data agree with the theoretical predictions made by the BCF equation?

In this study, the growth kinetics of the AlN buffer layer at ~800⁰C and the ML-GaN/AlN MQW structures at 690⁰C was not studied in detail. Meanwhile, we observed the difference in the density of growth hillocks on these surfaces. In the former layer, the higher growth temperatures lead to an increase in the characteristic radius of curvature ρс due to an exponential increase of equilibrium vapor pressure, while the structure grown at lower demonstrate a higher density of the growth hillocks (i.e. smaller radii). Figures 1and 2 clearly demonstrate this difference. Thus, the growth of these layers can be described using the basic formula of BCF theory with different supersaturation degrees (p/p0), which is a crucial parameter determining the radius curvature of the growth spirals and other parameters. Moreover, the MQW structures grown under different stoichiometric conditions also exhibit the various growth spiral density, as shown in Figures 2a and d. The structure grown under lower Al excess (i.e. with larger p in supersaturation formula) has relatively large growth spirals (Figure 2a), while the structure grown under relatively high Al excess demonstrate much denser growth spirals (Figure 2b). Here, I should apologize for the typo in the submitted manuscript, where instead of right Figure 2a and d we wrongly wrote Figure 1a and d. In addition, we add some information to improve the clearance of the text:

In general, the surface topology of the MQW structures is similar to the spiral hillock one observed during the PA MBE growth of AlN/с-Al₂O₃ templates, described above. The increased density of the spirals in the MQW structures can be associated in accordance with BCF theory with a lower growth temperature of 690°C leading to an increase in (p/p₀) pressure ratio (supersaturation degree), as well as with relatively high Al and Ga excesses accumulated during growth of barriers and wells in the MQW structures. This assumption is confirmed also by observation in Figure 2a,d of the different hillock densities of 4·10⁸ and 2·10⁹cm^-2 for the MQW structures grown using the different Al excess of 2.1 and 2.5 ML, respectively, which lead to the higher supersaturation in the latter.

2. In paragraph 3.6.1 the authors discuss the blue shift of the CL peaks upon lowering the thickness of the GaN layers by dividing them into two subgroups: peaks 1-4 and peaks 5-7. Despite that, the discussion on the two subgroups is mostly analogous leading to the same conclusion that the narrower the GaN layer, the more energetic is the CL peak. Which is the reason why the authors chose to split the CL peaks in 2 subgroups?

In this study, we divided our MQW structures into two groups comparing their CL intensities, which demonstrate a nonmonotonic dependence on the QW thickness with the maximum intensity in the structure with a QW thickness of about 2ML (structure 4). At the current level of understanding of the processes in such ultrathin QWs, we cannot explain the continuous increase in the position of the CL peaks with thinning of the wells and the observed nonmonotonic behavior of the CL intensity. Indeed, both aforementioned factors leading to the observed blue shift of the CL should also lead to its enhancement that contradicts to the experiment data. Further work is needed to clarify this issue. In this paragraph (from 360 to 372 lines), we briefly discuss the plan of such studies, taking into account the possible effects of localization of charge carriers etc.

3. The authors should briefly comment the shape of the CL peaks #2 and #3, featuring multiple peaks.

I must apologize for the typo in the description of the CL spectra on line 382, where we incorrectly typed “spectra 6,7 in figure 7a” instead of the correct “spectra 2,3 in Figure 7a”. Therefore, all of the above discussion describes the feature of spectra 2,3.

4. Several typos can be found throughout the text. Acronyms should always be explicitly reported in their extended form the first time they are reported in the text. Italics letters (such as "n-" and "p-type") should be evenly reported in the whole work. All numbers should have their units.

We corrected the typos and added missing units. They are highlighted in green.

# English language and style are fine/minor spell check required

We have followed this recommendation and the changes are highlighted in green.

Reviewer 2 Report

In the article "Monolayer-scale GaN/AlN multiple quantum wells for high power e-beam pumped UV-emitters in the 240-270nm spectral range" the authors realized a monolayer-scale GaN/AlN multiple quantum well for electron beam pumped UV emitters. The article is well written and the results seem to be correct. I recommend it for the publication.

Author Response

Thank you for your review and high evaluation of our article.

# English language and style are fine/minor spell check required

We have followed this recommendation and the changes are highlighted in green.

Reviewer 3 Report

The authors prepared monolayer-scale GaN/AlN multiple quantum wells for high-power e-beam pumped UV emitters. These structures show quite flat stepped surfaces. Also, the stress-free growth in this manuscript is advantageous for UV emitters and those UV emitters show quite good characteristics. So, there are some questions in the manuscript.

1. Please show the schematic diagram for understanding the growth mechanism.
2. What is the difference between the general x-ray diffraction pattern and figure 5? Please show the x-ray diffraction pattern.
3. In the manuscript on page 5, Figure 1a, d of the different hillock densities~ are written, but there is no figure 1d in the manuscript. Please check it.
4. How calculate the ratio of nitrogen? Please show any evidence of element ratios such as EDX or XPS.
5. Also, high-resolution images of the surface are recommended.
6. What are the different characteristics between small radii and larger radii up to 1μm?

Author Response

1. Please show the schematic diagram for understanding the growth mechanism.

We add in Figure 1 the schematic diagram of the spiral step-flow growth discussed in the article

2. What is the difference between the general x-ray diffraction pattern and figure 5? Please show the x-ray diffraction pattern.

Figure 5a shows the general reciprocal space mapping (RSM) of around the (11-24) reflection with three MQW satellites from 1^st to 3^rd reflexes, while Figure 5b shows a narrow RSM region near 0^th one. The latter RSM was plotted using a larger scale to determine the average Al-content and stress in the MQW structure.

3. In the manuscript on page 5, Figure 1a, d of the different hillock densities~ are written, but there is no figure 1d in the manuscript. Please check it.

This is a typo and Figure 2a and d should be written instead of the incorrect Figure 1a and d.

4. How calculate the ratio of nitrogen? Please show any evidence of element ratios such as EDX or XPS.

We have described in detail the calibration of the growth fluxes in the growth rate units in our previous articles, which are summarized in the Chapter on PA MBE growth (Ref.34). Briefly, measurements of the growth rates of III-N layers grown using PA MBE under nitrogen- and group-III enriched stoichiometric conditions are used to calibrate the group-III and Nitrogen fluxes, respectively. This approach developed twenty years ago and most of the technological groups working in PA MBE technology used ML/s units for flux calibration. In the Chapter, we also describe the possible accumulation of excess metal in the form of the surface microdroplets, while the binary III-N layers have unit stoichiometry with extremely small deviation in the form of small metallic nanoclusters or point defects. In this work, we used metal-modulated epitaxy and growth interruption under activated nitrogen flux (see Ref.29) to exclude the appearance of excess metal in the AlN buffer layers and the MQW structure.

5. Also, high-resolution images of the surface are recommended.

Thank you for the recommendation. At present, we are writing a new article about the growth mechanisms of the ML-GaN/AlN structures using the advanced AFM study. It will include the study of the surface with a high resolution.

6. What are the different characteristics between small radii and larger radii up to 1μm?

The growth spirals with a large radius (up to 1µm) exhibited the AlN buffer layer grown at the higher growth temperature (up to 850⁰C), while the smaller radii of the growth spirals were revealed for the MQW structures grown at the much lower growth temperature (690⁰C). This difference completely corresponds to the spiral growth mechanism in accordance with BCF theory.