Quantification of High Resolution Pulsed RF GDOES Depth Profiles for Mo/B4C/Si Nano-Multilayers
, Songyou Lian 2,3,†, Patrick Chapon 4, Yibing Song 1, Jiangyong Wang 2,* and Congkang Xu 2,*Round 1
Reviewer 1 Report
Authors present an improvement of model for correct depth profile resolving capability of RF glow discharge OES. I think that the presented manuscript is written very good. It contains an important insights in a particular problem on literally subnanometer scales and discussing an interesting issue with experimental method. The math behind it has more applications in other depth profiling techniques and can be useful for broader community.
Though I do not have any questions to language, it took me awhile until I fully understood the manuscript (most probably because i am not directly involved with depth profiling sputtering techniques). In my opinion few moments could be improved to increase its readability and consistency.
1) Page 2, line 51. Though from a context it could be guessed, it is better to define A and D notation of layers. Also, later on authors use BC layer and B element, it is slightly confusing in my opinion.
2) Page 2-3, lines 83-85. If I understood correct, it is border conditions, right? If yes, maybe it is better to reformulate sentences to make it clear.
3) In my opinion (might be wrong), Figures 1, 2, and 3 are not necessary. all Maybe their number can be reduced, or added a table with shifts of maximum position as this is the most crucial inrormation. At least, I felt that in discussion on Page 7, lines 148-162 a tabular data would better support the discussion similar to Fig. 4.
4) I wonder why not to add an experiemntal comparison within modeling section with 3 sandwitch structures, or why you decide to present it separately?
5) As far as I understand, Authors do not have a sample production capability, though in Section 4 about application of model on real sample, in my opinion zoom in Figure 5 (for a 2-3 sandwitch A/BC/D layers) and shorter sputtering time (or zoom) in Figure 6 would be better.
6) Page 8, line 175 - Better to rewrite sentence - that 7 nm is a total sadwitch layer thickness.
Author Response
1) Page 2, line 51. Though from a context it could be guessed, it is better to define A and D notation of layers. Also, later on authors use BC layer and B element, it is slightly confusing in my opinion.
In the simulated sandwich structure, the first layer is pure element A (A layer), the second layer is the mixture of elements B and C (BC layer) and the third layer is pure element D (D layer). To clarify, the text is modified as “…on the depth profiles of elements A and D….”.
2) Page 2-3, lines 83-85. If I understood correct, it is border conditions, right? If yes, maybe it is better to reformulate sentences to make it clear.
The contribution of λ to the depth resolution function in the MRI model is originated from the detected signals by analyzer. Because of low sputtering power in GDOES depth profiling, only the atoms from the topmost layer (surface) of sample are sputtered out and detected. The atoms underneath of the surface have almost no effect on the detected signals. Therefore, the value of λ could be set as zero (taking λ= 0.01 nm for simulation). The sentences in lines 83-85 is reformulated as “Because of low sputtering power in GDOES depth profiling, only the atoms of the topmost layer….. “
3) In my opinion (might be wrong), Figures 1, 2, and 3 are not necessary. all Maybe their number can be reduced, or added a table with shifts of maximum position as this is the most crucial information. At least, I felt that in discussion on Page 7, lines 148-162 a tabular data would better support the discussion similar to Fig. 4.
Thanks for the suggestion. Table 1 is added and Figs. 1-3 are modified and the inset numbers are reduced.
4) I wonder why not to add an experiemntal comparison within modeling section with 3 sandwitch structures, or why you decide to present it separately?
It will be more clear to see the comparison between the fit and the experimental data if the fit is presented separately.
5) As far as I understand, Authors do not have a sample production capability, though in Section 4 about application of model on real sample, in my opinion zoom in Figure 5 (for a 2-3 sandwitch A/BC/D layers) and shorter sputtering time (or zoom) in Figure 6 would be better.
Yes, the experimental data were taken from the reference. If we take a small part of data to fit, it would be sure that the fitting and the comparison will look better. However, it is a great challenge to fit the data as more as possible in order to verify the extended MRI model. Certainly, we have to take into account the view effect in the published paper. Therefore, we take the data in the sputtering time range of 15-35s to fit.
6) Page 8, line 175 - Better to rewrite sentence - that 7 nm is a total sadwitch layer thickness.
Yes, the total thickness of the sandwitch layer is 7nm. The text in line 175 is modified as “The total thickness of Mo/B4C/Si multilayer and the individual layer thickness values of Mo, Si and B4C layers in each period are determined as 7.00 ± 0.05 nm, 3.0 ± 0.2 nm, 3.7 ± 0.2 nm and 0.3 nm, respectively, by HR-TEM and GIXRD techniques.”
Reviewer 2 Report
In this manuscript, the Authors investigate the extended "mixing-roughness-information" (MRI) model in the context of correctly identifying the depth profiles of trilayers with very thin delta layers, where parameters such as the sputtering rates or layers thickness are varied. The manuscript is clear and well written and the conclusions supported by the presented data. The main outcome is probably that the common FWHM model of measured time-concentration profiles is unreliable. I do not have significant suggestions to improve the manuscript, except to take care to prepare all the panels of multi-panel figures (i.e. fig 1, 2, 3, 7) with the same width.
Author Response
Thanks for the suggestion. It has been done.
Reviewer 3 Report
The manuscript “Quantification of High Resolution Pulsed RF GDOES Depth Profiles for Mo/B4C/Si Nano-Multilayers” by Yang et al. demonstrates the validity of extended mixing-roughness-information (MRI) depth model to estimate thickness of individual layers of nano-multilayer and multi-element films when measured using pulsed-radio frequency glow discharge optical emission spectrometry (RF GDOES). It also covers the influence of thickness and sputter rate of various ultrathin layers.
It is an interesting paper, and especially useful is suggested model for the technique in question. The posted question was addressed, and the usefulness of the proposed model when compared to the conventional FWHM was fully proved. Furthermore, MRI simulation results were compared with those of experimental data, and were found to be in very good agreement: layer thickness estimation, sputter rate of individual layers, as well as each element.
Manuscript is well written and easy to read, conclusions and discussion are supported by data.
The work can be accepted in present form.
Author Response
Thanks for the positive remarks.
