Quantitative Modeling of Near-Field Interactions in Terahertz Near-Field Microscopy

Round 1

Reviewer 1 Report

The manuscript describes a numerical modeling of near-field interaction between the tip and a sample’s surface in the THz-SNOM spectroscopy. Conceptually this work is similar to Ref. 16 (Mooshammer et al., ACS Photonics 7, 344 (2020)). The only difference is that the present study carries out a simplified modeling in 2D, whereas Ref. 16 used a 3D model. The authors claim that the simulation with the 2D model is much faster than with the 3D model, but all results are basically the same. In principle, this is a useful finding that would justify the publication of the manuscript. However, in our opinion, the current version does not provide a sufficient evidence for the authors’ claims and is not written clearly enough to be accepted in Applied Sciences. Below we list the main problems that have to be fixed:

1      1) The comparison between the 2D and the 3D models. The manuscript gives no direct comparison. Instead a set of plots with the near-field distribution is shown and it is claimed that it agrees with Ref. 16. This is far from sufficient. Of course, the near-field distribution simulated by any model will look qualitatively similar. However, this does not prove the correctness of the model. One should show the simulated scattering amplitude, the width of the field distribution and the field enhancement factor as a function of the tip-sample distance like in Ref. 16 and to compare them with the 3D model results.

       2) The simulation conditions are not clearly described. What is the frequency of the incident in the simulations shown in Figs. 3, 6 and 7? Also these figures should have a scale bar in order to give an idea about the length scale of the field distribution.

3      3) Section 3.2 discusses the edge effect and Fig. 4 shows the schematic diagram of the scattering signal. However, no results of the model are presented. Figs. 6 and 7 depict the field distributions for few tip positions, but there is no plots of the measureable quantity such as the scattering as the function of the tip position.

4      4) The advantage of the 2D vs 3D modeling should be quantified. How much faster the same scattering geometry can be simulated using the 2D model?

In conclusion, potentially this work can be published, but not in the present form.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This research presents a physical model and numerical simulation method to simulate the tip-sample interaction in THz-sSNOM. The result could be helpful for improving imaging quality.

My comments:

1. Page 2 Line 85. Please specify which software was used in this work.

2. Page 5 Line 152. What does "first portion" refer to?

3. Page 5 Line 177. Please include the references of "other's 3D models".

4. Please describe the simulation conditions in detail for Figure 3, 6, and 7, including all geometry parameters and EM wave parameters.

5. Please check and polish the language, such as Page 5 Line 158, Page 8 Line 243.

6. Page 10 Line 303. The authors mentioned a "subsequent experiment", but it was not presented in this manuscript. Do you mean a "follow-up study"?

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The revised version provides some additional details and it looks already better compared to the initially submitted manuscript. In my opinion, few more things have to be done before the publication:

11) The new version gives the difference in the element numbers for the 2D and 3D cases. Somewhat surprisingly, the 3D modelling requires only about 3 times more elements, which should not increase the computation time dramatically. Still it would be good to perform a direct comparison of the computation time and the results (for example, the approach curves) for 2D and 3D models. The authors claim that their RAM is not sufficient for the 3D calculation. However, the number of elements does not seem so huge and the COMSOL error is related to a bad convergence. It should be feasible to run the 3D COMSOL model on a reasonable PC. Then the comparison would be straightforward.

22) The new Fig. 3 with approach curves is too small and it is almost impossible to read any text there. It has to be scaled properly. Also it would be good to estimate the decay lengths for these curves as it was done in Ref. 16.

33) Fig. 5 probably has an error in the horizontal scale. I guess that the position units should be in micrometers, not nanometers.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

It is a bit disappointing that the authors have not managed to perform a direct comparison between the 2D and 3D models. However, I do appreciate their effort. Regarding the decay length for the approach curves. As far as I understand, it just the length at which the electric field amplitude drops by the factor of e=2.718 (see the inset in Fig. 3b of Ref. 16). According to Fig. 3 of the present manuscript, this length should be around 10 nm. Thus, I do not understand the problem with its estimation. Finally, I recommend to improve the quality of Fig. 9 that shows the edge effect, since it is one of the main results of the manuscript. The authors have placed the scale bar on the color plots. However, the plots with electric field distribution does not have any units for the horizontal position axis. Moreover, these graphs are rather pale and small. I recommend stretching them, increasing the font size and the line thickness. Then the manuscript can be accepted for publication.

Author Response

Please see the attachment.

Author Response File: Author Response.docx