Collective Modes in the Luminescent Response of Si Nanodisk Chains with Embedded GeSi Quantum Dots

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Vladimir A. Zinovyev et al reported a research manuscript titled "Collective modes in the luminescent response of Si nanodisk chains with embedded GeSi quantum dots". Authors investigated the effect of chain period and Si nanodisk resonator radius on the emission pattern using micro-photoluminescence spectroscopy. Further authors found optimal parameters for nanodisk chains for the quantum dot emission. Authors also evaluated the change in far-field radiation pattern with increasing nanodisk chain length using theoretical calculations.  These studies will help in developing near-infrared light sources.

Minor: Reference 18: Nanolett should be in italic and year should be bold. Make sure all references are according to the standard format.

Author Response

We have corrected the reference 18.

Reviewer 2 Report

Comments and Suggestions for Authors

The paper entitled 'Collective modes in the luminescent response of Si nanodisk chains with embedded GeSi quantum dots' study the emission of GeSi QDs modulated by a chain of Si nanodisk supporting Mie resonance. The experimental results are profound, supported by solid theoretical analysis. Only several minor comments need to be addressed from me. 

1. In the manuscript, the name of the quantum dot is not consistent, sometime GeSi and sometime SiGe. The authors need to keep the naming in consistency in convenience to readers. 

2. The author may consider give more justification on why they chose to use GeSi quantum dot over other QDs of other materials. 

3. in Fig.8, I will recommend all the figures use the same colormap and normalize to the same maximum value, so that the Ey component will not show as strong as x/z component. To my understanding, Ey components in this resonance is much weaker. 

4. The author may want to provide more clarifications on why they consider the mode is a WGM, given the radius of the silicon pillar is even smaller than the wavelength. Can is be just a Mie resonance as previously reported? https://www.science.org/doi/10.1126/science.aaz3985

 

Comments on the Quality of English Language

Some grammar errors are found. The authors may want to carefully check their writing. Also, the use of 'it's' should be avoided. 

Author Response

Reply  to second referee

1. In the manuscript, the name of the quantum dot is not consistent, sometime GeSi and sometime SiGe. The authors need to keep the naming in consistency in convenience to readers.

We have replaced  "SiGe" by "GeSi" everywhere.

2. The author may consider give more justification on why they chose to use GeSi quantum dot over other QDs of other materials.

We have added some sentences about GeSi quantum dots in  Introduction.

3. in Fig.8, I will recommend all the figures use the same colormap and normalize to the same maximum value, so that the Ey component will not show as strong as x/z component. To my understanding, Ey components in this resonance is much weaker.

We have made alternative version of this Figure, now this is Figure 11.

4. The author may want to provide more clarifications on why they consider the mode is a WGM, given the radius of the silicon pillar is even smaller than the wavelength. Can is be just a Mie resonance as previously reported? https://www.science.org/doi/10.1126/science.aaz3985

You are right, indeed, the obtained modes can be considered as the Mie modes with a dominant magnetic octupole contribution. We have compared our modes  with analytical solution of the Mie problem for a sphere of same radius (385 nm)  and obtained that these modes have very similar near field distribution. Moreover, the numerical simulations of the eigenmodes for a sphere and a cylinder with the same radii give the  very close  spectral position of the resonances as well as similar  near-field distribution.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript entitled “Collective modes in the luminescent response of Si nanodisk chains with embedded GeSi quantum dots”  presents results of  experimental and theoretical investigations of the photoluminescence of GeSi quantum dots (QDs) embedded into a linear chain of silicon sub microresonators supporting whispering gallery modes (WGMs). The paper follows up recent papers on silicon microresonators based on Mie modes and WGMs and their tests for photoluminescence enhancement. The manuscript presents new experimental data and theoreticl results and there is merit to publish it. However, some issues have to be modified. They are:

11.       The manuscript uses the term nanodiscs. However, diameters and heights of the Si discs are above 100 nm, i.e., the discs can be classified rather as submicrometer discs than nanodiscs.

22.       More detailed explanation of Fig. 1a is necessary (e.g., the reason for the trapezoidal shapes in discs). The growing and etching rates would be useful.  In Fig. 1b, the disc diameter, period and chain distance should be added to the caption.  How did you prepare samples for SEM measurements? The photoluminescence excitation power would be added.

33.       Readers could be confused when comparing Figs. 2a and 2b because PL intensities are very different on these figures even for the same discs. Is there any reason why not use the same format as that in Fig. 1b in Fig. 2b?

44.       WGM resonances have been found for a disc period of 1 µm., i.e., the minimum period investigated. Could you decrease this period below 1 µm in order to find if the WGM resonances improve?

55.       It would be useful to see the dependence of PL intensity of the two WGM modes on the number of discs. There are many curves in Fig. 4a which make it a rather unclear from this point of view.

66.       What software did you use for the modeling? Specify it, please.

77.       Can you increase the experimental quality factors? What imperfections are responsible for a decrease of the quality factor of the transvers mode? Did you make any band resolution of measured PL spectra?

88.       Experimental quality factor values are lower than those obtained with submicrometer pillars employing Mie modes (ACS Photonics 2021, 8, 209-217). Discuss the difference, please?

99.       In Conclusion you write that” We demonstrate the high potential of the linear chains of Si disk resonators supporting WGM modes for development of effective light sources in near infrared range.” On my view, you have shown that specific linear chains of SI resonators embedding GeSi QDs can support two distinguished WGMs. A reader could hardly find results for effective light sources in the manuscript.  

110.   In Keywords GeSi quantum dots are missing

111.   In line 64 you use SiGe QDs instead of GeSi QDs. Correct it, please.

Author Response

Replay to third Referee

11. The manuscript uses the term nanodiscs. However, diameters and heights of the Si discs are above 100 nm, i.e., the discs can be classified rather as submicrometer discs than nanodiscs.

We have removed the prefix "nano".

22.  More detailed explanation of Fig. 1a is necessary (e.g., the reason for the trapezoidal shapes in discs). The growing and etching rates would be useful. In Fig. 1b, the disc diameter, period and chain distance should be added to the caption.  How did you prepare samples for SEM measurements? The photoluminescence excitation power would be added.

There's been a bit of a misunderstanding.  In Fig. 1a we show schematically  the quantum dots  embedded inside the resonators as having a trapezoidal shape. The resonators have a cylindrical shape.  We have added in the text the information about  growing and etching rates. Also we have added in the caption to Fig. 1b the disc diameter, period and chain distance. We performed SEM measurements immediately after etching the disks without additional preparations.

33. Readers could be confused when comparing Figs. 2a and 2b because PL intensities are very different on these figures even for the same discs. Is there any reason why not use the same format as that in Fig. 1b in Fig. 2b?

In all cases we present the data in arbitrary units. We shifted  the spectra along the y-axis on purpose so that there would be no crowding of spectra, because of which the readers can't see details. In new version of manuscript we present the data in Fig 2a and 2b in the same format.

44. WGM resonances have been found for a disc period of 1 µm., i.e., the minimum period investigated. Could you decrease this period below 1 µm in order to find if the WGM resonances improve?

We perform the additional  calculations  to  study how the quality of collective  mode depends on the period of chain.  Indeed, the decreasing of period results in improving of considered resonance. At period 950 nm the maximum quality value was obtained for "transverse" mode. However, for "longitudinal" mode the optimal period is 1020 nm. We add this result in Figure 7a in new version of manuscript.

55. It would be useful to see the dependence of PL intensity of the two WGM modes on the number of discs. There are many curves in Fig. 4a which make it a rather unclear from this point of view.

We processed the spectra and plotted the dependence of peak intensity on the number of resonators. We  have added this dependence in Supplementary Materials, Figure S2.

66. What software did you use for the modeling? Specify it, please.

We use commercial software COMSOL Multiphysics. We add this information in paragraph "Materials and Methods"

77.  Can you increase the experimental quality factors? What imperfections are responsible for a decrease of the quality factor of the transverse mode? Did you make any band resolution of measured PL spectra?

The experimental values of quality were obtained on the basis of PL  spectrum fitting by a set of Lorentz functions. The number of these functions in the expansion was chosen to be the same when analyzing several spectra in a series.  We have added the   PL  spectrum fitting in Supplementary Materials, Figure S1.

Concerning the reason of experimental decreasing quality factor as compared with theoretical value.  Since we are dealing with quality values obtained  from luminescence measurements, then the absorption on free charge carriers should play a major role. At given sizes of disks the excitation already at low pump powers should lead to a significant concentration of charge carriers in the disk, since the "diffusive escaping" may not exist here. Unfortunately, the dependence of the quality factor on the excitation power was not studied. The power was chosen based on the criterion of signal observation and absence of overheating. The overheating was controlled  by avoiding the presence of displacement of peaks to the long-wavelength region at high pumping.

Of course, there is also imperfection of experimental structures, because of which the quality value drops too, but our structures are no worse than those in Rutskaya's work. By the way, the volume of disks in Rutskaya's work is 6 times less, so the losses on free carriers are less and the difference between experimental quality value (~500)  and theoretical value (~800) is smaller.

88. Experimental quality factor values are lower than those obtained with submicrometer pillars employing Mie modes (ACS Photonics 2021, 8, 209-217). Discuss the difference, please?

One reason is the different losses on free carriers discussed in the answer on previous question 77. The second reason is the different nature of the modes responsible for luminescent response in our case and in the case of work (ACS Photonics 2021, 8, 209-217). Rutskaya et al. deal with the band edge mode, while we study the modes above the light line.

99. In Conclusion you write that” We demonstrate the high potential of the linear chains of Si disk resonators supporting WGM modes for development of effective light sources in near infrared range.” On my view, you have shown that specific linear chains of SI resonators embedding GeSi QDs can support two distinguished WGMs. A reader could hardly find results for effective light sources in the manuscript.

OK, we have rewrote Conclusions.

110. In Keywords GeSi quantum dots are missing

We have added  GeSi quantum dots in Keywords

111. In line 64 you use SiGe QDs instead of GeSi QDs. Correct it, please.

We have corrected this misprint.

 

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

In this manuscript, Viadimir et al reported a collective modes in the luminescent response of Si nanodisk chains with embedded GeSi QDs. The paper is well organized and can be accepted after the following issue were concerned.

 

1. The actual bonding between Ge,Si and Si should be provided.

 

2. What is the mechanism to enhance the luminescent response of Si nanodisk through GeSi QDs? Were GeSi unique or it can also be other 2D quantum dots?

 

3. The outstanding properties of this GeSi QDs should be compared with others' work.

 

4. Progress of this field can be added, especially in those two years.

Author Response

Replay to fourth Referee

1. The actual bonding between Ge,Si and Si should be provided.

We have added the corresponding sentences to Introduction.

2. What is the mechanism to enhance the luminescent response of Si nanodisk through GeSi QDs? Were GeSi unique or it can also be other 2D quantum dots?

The enhancement of luminescence response of GeSi QDs can be achieved due to an increase in the probability of optical transitions, when QDs are placed at the maximum of the electric field inside the resonator (Purcell effect). GeSi quantum dots are unique in that there are already well-developed spatial ordering methods for them and it is well compatible with Si technology, which is not the case for other quantum dots.

3. The outstanding properties of this GeSi QDs should be compared with others' work.

The outstanding properties of GeSi quantum dots are that they can be embedded naturally into a Si disk resonator during the MBE process, or grown directly on its surface as demonstrated, for example, in the work [S. Wang et al.  Nanotechnology 2018 29 345606]. Moreover, with the correct selection of growth conditions and resonator parameters, the nucleation of quantum dots will occur in places that correspond to the maxima of the field distribution of the resonator eigenmodes, which will subsequently can lead to an enhanced luminescent response of QDs. 

4. Progress of this field can be added, especially in those two years.

We have added the discussion of the progress in this field in Introduction.

 

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

I recommend to publish this manuscript without changes.