Investigation of Morphology of Aluminum Co-Doped Scandium Stabilized Zirconia (ScAlSZ) Thin Films

Reviewer 1

Paper presents experimental and theoretically results on to the morphology of Aluminium co-doped Scandium stabilized Zirconia (ScAlSZ) thin films deposited by e-beam physical vapour deposition system EB-PVD on amorphous SiO₂. In the paper the influence of substrate temperature and deposition rates on the surface roughness and the phase structure of thin films are investigate by using of experimental and of mathematical model.

It is shown that surface roughness depends on deposition temperature and deposition rate is not monotonously.

The article is of interest of Coatings, but specific aspects mentioned in the following require the revision of the paper (major revision).

Some information are inconsistent or duplicated, the paper could be more synthesized

The duplicated information was corrected.

unfocused images was used

The high quality images are inserted in the manuscript: Fig. 1, Fig. 2, Fig. 5, and Fig. 6. It should be noted, that MSWord reduces the quality of images automatically. If Reviewers and Editor would like to get high quality and high resolution images, they could be sent separately.

The fig. 2 presents inconsistent results. It is reveal that the roughness of ScAlSZ thin films increases with the substrate temperature, but the explanation "There are two values (0.8nm/s and 1.6 nm/s for 600ºC) shifted from the dependence due to the effect of deposition rate", does not stand, as in 1.2 nm/s the same effect is not observed, or if it is so, the values obtained for the other deposition rates (including of 1.2 nm/s) are erroneous

Thank you for the comment. The plots were revised. We agree that those two points are shifted from the tendency. The AFM scans were done at least in three different places on the same sample, and average values of Rq were included in the plots. So, the Rq values should be reliable and confirm the presented results.

Related to the fig. 4, the deposition rate of 1.2 nm/s is presented as a particular breaking point where a sudden decrease of surface roughness appears, but the figure 4 show a sudden increase of the roughness. This effect is not explained in the paper and, correlated to the result presented in figure 2, it could be attributed to errors, so the experiment should be reconsidered / repeated

Thank you for the comment. The plots were revised. Everything depends on the viewpoint.  On the left-hand side, from 0.2 nm/s to 1.2 nm/s, the values of roughness increase, and on the right-hand side, from 1.2 nm/s to 1.6 nm/s decreases. That’s why 1.2 nm/s is a breaking point. Talking about the reliability of the results. The nature of the roughness dependences on the deposition rate is repeating (curves have the same shape) for all deposition temperatures. The highest deviation of roughness occurs when the substrate temperature is the highest, i.e. 600°C. That shows that the microstructure and growth tendency of thin films depend on substrate temperature as it is possible to see from SEM images (Fig. 5 and Fig. 6). Of course, there are small deviations. Moreover, the AFM scans were done at least in three different places on the same sample and repeated with several of them. The average values of Rq were included in the plots. So, we think that there is not necessary to repeat experiments.

Figure 7 caption should be completed (substrate temperature)

Title of horizontal axis of Figure 7 has been changed from “Temperature” to “Substrate Temperature”.

In order to be able to evaluate whether the model correlates with the experimental data, it should have been presented in figure 8 the results obtained by modelling at other deposition rates (including of 1.2 nm/s)

Additional plots of average grains size vs substrate temperature observed at the growth rates of 0.2, 1.2 and 1.6 nm/s have been to Figure 8.

Additional text describing Figure 8 has been added to Page 12.

This text and other parts added or changed are highlighted.

Similarly, for a more comprehensive evaluation of the model, the average grain size dependence on the surface roughness for other deposition rates should have been shown in Figure 9

Additional plots of surface roughness vs average grains size observed at the growth rates of 0.2, 1.2 and 1.6 nm/s have been to Figure 9.

Additional text describing Figure 9 has been added to Page 12.

This text and other parts added or changed are highlighted.

The graph should be plotting between points and presenting of the equations on the graph could be a plus (e.g. in the case of grows rate fig. 12)

The trendlines and their respective equations with R² values have been added to Figure 12.

Reviewer 2

The paper deals with the modeling of the growth and of the surface morphology of ScAlSZ thin films deposited by an e-beam technique. The results of the modeling are compared with those obtained by experiments. The topic is interesting, and the paper is well organized, so only minor changes are required prior to publication.

The quality of AFM images of Figures 1 and 3 is quite bad and should be improved.

The quality of AFM images was improved.

The cross sections of Figure 5 are very similar and do not seem to indicate a different growth mechanism.

With all respect to your opinion, we think there is a difference, especially between Fig. 5 a) and Fig. 5 c). So, we would not like to change the figure.

In line 158, the authors speak about the characterization of ScSZ films. Those films are never discussed in the whole paper.

Thank you for the remark. It was simple mistyping, and it was removed.

Reviewer 3

The paper entitled “Investigation of Morphology of Aluminum Co-doped Scandium Stabilized Zirconia (ScAlSZ) Thin Films” focuses on investigating the influence of the deposition parameters of an e-beam physical vapor deposition process on the formation of thin films and their microstructure. The influence of substrate temperature and growth rate on the phase structure and surface roughness of biphasic thin films is also reported. The dependences of surface roughness on substrate temperature and growth rate as well as the surface roughness of grown films and average nanoparticle size have been determined. After the deposition, the coating structure and composition were evaluated by analyzing the cross-sectional morphology, elemental composition, and surface roughness. In my opinion, the paper should be interesting from a scientific and practical point of view.

I would like to recommend the publication of the paper after addressing the following issues:   

Although the experimental part is consistently revealed and explained, the authors could add more details about the experimental conditions and/or parameters.

The experimental part was supplemented with experimental parameters and conditions.

Why do the authors use the Rq surface roughness parameter instead of the common Ra

We prefer to use Rq due to a few reasons:

·         Rq is the square root of the distribution of surface height and is more sensitive than the average roughness for large deviations from the mean line.

·         Ra and Rq is usually used as interchangeable.

·         Other authors also use Rq for the evaluation of thin roughness https://link.springer.com/article/10.1007/s10856-016-5707-4

It is not clear from the text how many topographic images (10 μm x 10 μm) have been taken to determine the roughness values. The results seem statistically unreliable. The same for the average grain size and the thickness of the thin films.

The AFM scans were done at least in three different places on the same sample, and average values of Rq were included in the plots. So, the Rq values should be reliable. The grain size was also calculated from at least three different scans randomly measuring 10 grains in each topographic image.

Why have the authors used in all their graphs curved (smoothed) lines between the data points instead of fitting them with straight lines? These smoothed lines distort the data and they may be improper especially when we have no information about the values between the measured points. For plotting measured data, the only valid connecting curve between points is a straight line.

The line shapes were changed in Fig. 2 and Fig. 4 as were recommended. Thank you.

Where could the readers see the results from the elemental analysis performed by using energy-dispersive X-ray spectroscopy?

The additional information about Sc, Al, Zr, and O concentrations (EDS measurements) is added and can be found in supplementary materials (Table S1).

To compare the phase composition of the examined coatings with the calculated surface map, it would be better for the authors to present XDR diffraction patterns of the coatings and to explain which phase is A and which one is B.

XRD results and crystal phase distribution in the formed thin films are already published: Sriubas, M.; Kainbayev, N.; Virbukas, D.; Bočkute, K.; Rutkuniene, Ž.; Laukaitis, G. Structure and Conductivity Studies of Scandia and Alumina Doped Zirconia Thin Films. Coatings 2019, Vol. 9, Page 317 2019, 9, 317, doi:10.3390/COATINGS9050317. The additional explanation with the reference is added to the text.

Where applicable, the obtained results can be compared with other similar studies.

Roughness of the surface are expected to influence charge transfer at the electrolyte/electrode interface, which would impact the electrochemical impedance of SOFCs. The obtained results are additionally compared in the article using these references:

1. Talebi, T.; Haji, M.; Raissi, B. Effect of sintering temperature on the microstructure, roughness and electrochemical impedance of electrophoretically deposited YSZ electrolyte for SOFCs. Int. J. Hydrogen Energy 2010, 35, 9420–9426, doi:10.1016/J.IJHYDENE.2010.05.079.

2. Singh, A. V.; Ferri, M.; Tamplenizza, M.; Borghi, F.; Divitini, G.; Ducati, C.; Lenardi, C.; Piazzoni, C.; Merlini, M.; Podestà, A.; et al. Bottom-up engineering of the surface roughness of nanostructured cubic zirconia to control cell adhesion. Nanotechnology 2012, 23, 475101, doi:10.1088/0957-4484/23/47/475101.

3. Heiroth, S.; Lippert, T.; Wokaun, A.; Döbeli, M. Microstructure and electrical conductivity of YSZ thin films prepared by pulsed laser deposition. Appl. Phys. A 2008 933 2008, 93, 639–643, doi:10.1007/S00339-008-4689-6.

4. Eftekhari, L.; Raoufi, D. Crystallography characteristics of tetragonal nano-zirconia films under various oxygen partial pressure. https://doi.org/10.1080/02670844.2018.1555913 2018, 35, 618–626, doi:10.1080/02670844.2018.1555913.

Reviewer’s comments

Actions done

Reviewer 1

In Figure 2 the explanation still does not stand. Unlike all other curves, in the case of deposition rate of 1.2 nm/s, the curve starts from higher values (at 50°C). What is the explanation? If there is no clear explanation, the experiment should be repeated for this value (as well as for values close to it to explain the phenomenon).

The samples formed using 1.2 nm/s deposition rate were remeasured additionally in another laboratory. The adjusted information is presented in Figure 2 and Figure 4.

Regarding to figure 4, how did you come to the conclusion that exactly growth rate of 1.2 nm/s is the breaking point? Do you have a mathematical relationship through which you could prove this? (e.g. the derivative of the function changes its sign at that point). It seems that you do not have enough data for this. A supplement with experiments performed for growth rate of 1 nm/s and 1.4 nm/s could clarify this aspect.

The breaking point becomes more expressed with increasing deposition temperature. The same dependence is visible at all temperatures even after the remeasurement of samples (1.2 nm/s series), that evidently shows an existence of the breaking point. In addition, the paragraph was changed, and the following comments were added: “nonlinear at higher deposition temperatures” and “our experiments showed”. The similar results and roughness dependence on the deposition rate behavior were obtained with samarium doped ceria thin films [Sriubas, M.; Laukaitis, G. The influence of the technological parameters on the ionic conductivity of samarium doped ceria thin films. Mater. Sci.  2015, 21, 105–110.].

In the case of figure 12 (and other graphs), it is enough to present the trendline, without joining the points.

Lines used to connect the points in figures 9 and 12 have been removed where a trendline for a respective plot is present.

Reviewer 3

The authors have carefully addressed all the review’s recommendations and the manuscript has been substantially improved. Just one more comment, please also use straight lines for connecting the points in figures 7, 8, 9, 11 and 12.

Straight lines have been added to connect the points in figures 7, 8, 9, 11 and 12 where a trendline for a respective plot is absent. No lines are now used to connect the points in those figures where a trendline for a respective plot is present.

Reviewer’s comments

Actions done

Reviewer 1

In Figure 2 the explanation still does not stand. Unlike all other curves, in the case of deposition rate of 1.2 nm/s, the curve starts from higher values (at 50°C). What is the explanation? If there is no clear explanation, the experiment should be repeated for this value (as well as for values close to it to explain the phenomenon).

The samples formed using 1.2 nm/s deposition rate were remeasured additionally in another laboratory. The adjusted information is presented in Figure 2 and Figure 4.

Regarding to figure 4, how did you come to the conclusion that exactly growth rate of 1.2 nm/s is the breaking point? Do you have a mathematical relationship through which you could prove this? (e.g. the derivative of the function changes its sign at that point). It seems that you do not have enough data for this. A supplement with experiments performed for growth rate of 1 nm/s and 1.4 nm/s could clarify this aspect.

The breaking point becomes more expressed with increasing deposition temperature. The same dependence is visible at all temperatures even after the remeasurement of samples (1.2 nm/s series), that evidently shows an existence of the breaking point. In addition, the paragraph was changed, and the following comments were added: “nonlinear at higher deposition temperatures” and “our experiments showed”. The similar results and roughness dependence on the deposition rate behavior were obtained with samarium doped ceria thin films [Sriubas, M.; Laukaitis, G. The influence of the technological parameters on the ionic conductivity of samarium doped ceria thin films. Mater. Sci.  2015, 21, 105–110.].

In the case of figure 12 (and other graphs), it is enough to present the trendline, without joining the points.

Lines used to connect the points in figures 9 and 12 have been removed where a trendline for a respective plot is present.

Reviewer 3

The authors have carefully addressed all the review’s recommendations and the manuscript has been substantially improved. Just one more comment, please also use straight lines for connecting the points in figures 7, 8, 9, 11 and 12.

Straight lines have been added to connect the points in figures 7, 8, 9, 11 and 12 where a trendline for a respective plot is absent. No lines are now used to connect the points in those figures where a trendline for a respective plot is present.