Phases and Interfaces in the Cr–Fe–Si Ternary System: X-ray Diffraction and Electron Microscopy Study

Round 1

Reviewer 1 Report

This work experimentally investigated the phase composition in as-grown and annealed ternary Fe–Si–Cr system. Analytical scanning and transmission electron microscopy were used to study the mutual spatial distribution, shape and size of the formed phase grains, and determine the crystallographic orientation relations at the interphase boundaries. The conclusions can be basically supported by the experimental data and analysis. Here are some issues that need to be addressed before considering the publication of this work:

1. The phase diagram and thermodynamic assessment of the Fe-Cr-Si ternary system have been widely reported. The author did not make a detailed investigation on the relevant research. In the Introduction section, the authors should elaborate the research status of this ternary system, including phase equilibrium, thermodynamic properties and microstructure morphology, etc., so as to highlight the research purpose of this work.

2. The references about the crystallographic data for each phase in Table I should be given.

3. The authors should give the binary and ternary phase diagrams and mark the positions of the experimental component points.

4. The sentence " Quantitative composition of phases in the samples in accordance with Rietveld re-119 finement is estimated at 50% CrSi2hex-43% -FeSi2 tetr-7% -FeSi cub for the as grown 120 sample and 50% CrSi2 hex: 47% -FeSi2 orth-1.5% -FeSi cub-1.5% Si. " in section 3.1 is puzzling. It is recommended to reorganize the language and refine the symbols.

5. The Figures are presented poorly, with many inappropriate overlaps and mistypesetting. It is recommended to re-edit the Figures.

6. Thermal expansion coefficient is related to temperature, and the relationship between thermal expansion coefficient and temperature should be given when calculating the lattice mismatch.

Author Response

1. The phase diagram and thermodynamic assessment of the Fe-Cr-Si ternary system have been widely reported. The author did not make a detailed investigation on the relevant research. In the Introduction section, the authors should elaborate the research status of this ternary system, including phase equilibrium, thermodynamic properties and microstructure morphology, etc., so as to highlight the research purpose of this work.

Answer: Precisely because the phase diagrams of the CrFeSi ternary system have been sufficiently well studied and presented in the literature, we do not see the need to reconsider them. We used in this work the comprehensive data from the references [13, 14] to focus our attention on the microstructure an alloy of nominal composition CrFeSi4, namely: identification of the phases present, their spatial distribution and mutual orientation relationships, grain sizes, mismatch of crystal lattices at the interfaces and the occurrence of strains between phases. In the Introduction, the purpose of the study was already defined in this direction. The main advantage of electron microscopy study was not only to identify the presence of specific phases and their composition, including the minor phases, but also to access to the mutual spatial orientation of adjacent phases. From our point of view, not enough studies have been conducted so far to provide the data that can lead to the understanding of the brittleness of such alloys. We added some comments and some references on TEM study of Cr and Fe silicides to Introduction.

2. The references about the crystallographic data for each phase in Table I should be given.

Answer: Added references to crystal phases that we might expect could form.

3. The authors should give the binary and ternary phase diagrams and mark the positions of the experimental component points.

Answer: The positions of the experimental component points of the ternary system with nominal composition CrFeSi4 is at 17:17:66 (at.%) and is introduced in the paragraph Results, 3.1. Powder X-ray diffraction study.

4. The sentence " Quantitative composition of phases in the samples in accordance with Rietveld refinement is estimated at 50% CrSi2hex-43% -FeSi2 tetr-7% -FeSi cub for the as grown 120 sample and 50% CrSi2 hex: 47% -FeSi2 orth-1.5% -FeSi cub-1.5% Si. " in section 3.1 is puzzling. It is recommended to reorganize the language and refine the symbols.

Answer: The  new Table 2, which listed all the phases, is presented in paragraph 3.1. Powder X-ray diffraction study.

5. The Figures are presented poorly, with many inappropriate overlaps and mistypesetting. It is recommended to re-edit the Figures.

Answer: The Figure 2, 3 and 4 are revised and reshaped. We believe that it is clearer now.

6. Thermal expansion coefficient is related to temperature, and the relationship between thermal expansion coefficient and temperature should be given when calculating the lattice mismatch.

Answer: For sake of simplicity, it seems reasonable enough to limit ourselves to using average thermal expansion coefficients to estimate the contribution to the formation of possible strains, because such a contribution is negligible compared to the contribution made by the geometric mismatch of crystal lattices. The corresponding references (the sources of the data) are given.

Author Response File: Author Response.pdf

Reviewer 2 Report

From reviewing your work, I found it interesting, however some points should be improved, and others clarified so that the work can be accepted in this famous and prestigious journal.

1-      The language of the manuscript needs to be revised.

2-      The authors used Rietveld analysis for confirmation the identification of phases but did not stated or listed any fitted parameters as the weighted pattern (R_wp), expected values (R_exp), or the fit goodness index (χ²) which should be stated.

3-      The statement in line 121 “….for the as grown sample and 50 % CrSi2 hex: 47% β-FeSi2 orth - 1.5 % ε-FeSi cub - 1.5% Si.” not completed.

4-     Can the authors confirm if they used ZAF correction with EDXS measurements or not?

5-      The section (3.2. SEM and EDXS analysis) lacks more references.

6- The section (3.4. Crystallographic orientation relationships …) lacks more references.

Author Response

1-      The language of the manuscript needs to be revised.

Answer: Revision was done and correction are introduced.

2-      The authors used Rietveld analysis for confirmation the identification of phases but did not stated or listed any fitted parameters as the weighted pattern (R_wp), expected values (R_exp), or the fit goodness index (χ²) which should be stated.

Answer: The quantitative ratio of the identified phases was estimated in the results of XRD full-profile refinement by the Rietveld method using PANalytical X'Pert HighScore Plus software. In the initial stages the overall zero error, phase scale factors, background, profile parameters and lattice parameters of identified phases were refined. The peak shapes were approximated with pseudo-Voight profile function. Since the CrSi2 and FeSi2 phases exhibited the different pronounced orientations the preferred texture parameters were also refined in the final stages. The Rietveld refinements resulted in the following quantitative phase compositions: CrSi2 (50.5(4)%) – FeSi2tetr (41.9(4)%) – FeSi (7.5(2)%) with final agreement indices Rwp = 13.4%, GOF = 6.4 for as grown sample and CrSi2 (50.7(3) %) – FeSi2orth (46.8(3)%) – FeSi (1.3(1)%) – Si (1.2(1)%) with final agreement indices Rwp = 11.5 %, GOF = 5.2 for the annealed sample.

The revision was done and corrections are introduced

3-      The statement in line 121 “….for the as grown sample and 50 % CrSi2 hex: 47% β-FeSi2 orth - 1.5 % ε-FeSi cub - 1.5% Si.” not completed.

Answer: The new Table 2 was introduced.

4-     Can the authors confirm if they used ZAF correction with EDXS measurements or not?

Answer: ZAF is not a “correction” but a translation of the number of X-Ray photons emitted into the mass % or atomic % concentration in EDXS analysis on bulk samples. By default, it is included in all EDXS quantification software unless another option is chosen for special difficult cases.  We used ZAF  for EDXS element quantification on SEM and Cliff-Lorimer with thickness correction  on TEM, respectively.  The text is updated into the 2. Materials and Methods paragraph.

5-      The section (3.2. SEM and EDXS analysis) lacks more references.

Answer: The references on element quantification in SEM are introduced (Robinson [55], Lloyd [56], Goldstein [57])

6- The section (3.4. Crystallographic orientation relationships …) lacks more references.

Answer: The references on crystallographic orientation relationships, lattice mismatch, thermal expansion coefficient mismatch are introduced ([62]Howe, J. M.;  Aaronson, H. I.; Hirth, J.; [67] Kladko, V.; Kuchuk, et al.; [68] Liu, H.; Cheng, X.; Valanoor, N.)

Author Response File: Author Response.pdf

Reviewer 3 Report

This manuscript focuses on the phases and interfaces in the Cr–Fe–Si ternary system by X-ray and electron diffraction study. The article is well written and provides an in-depth structural analysis of the ternary phases (microstructure, phase composition, and crystallinity). The overall design, including methodology and results, is reasonable and would be of great interest to the readership of Inorganics. I would recommend this article for publication in Inorganics after minor modifications.

1. In the introduction, the authors pointed out that "Cr-Fe-Si" can be utilized as anode materials for Li-ion batteries. I am not entirely convinced by their claim; a number of papers published on the subject emphasize that FeSi2 does not contribute to capacity because it is largely unreactive or inactive in terms of storing Li-ions. The results reported here highlight that they have a significant amount of FeSi2 in their material system. Any comments on this?

DOI: 10.1002/smll.201906812

2. Why did you choose the 1037 K temperature and 72-hour annealing time? Did you optimize the time and temperature for the ternary system?

3. Oxygen contamination is very important, especially when the annealing is performed at high temperatures. Did you see any evidence of oxygen contamination in your material, especially from the EDX results? I can sense from Fig. 4 that there could be a slightly smaller oxygen peak. This should be checked thoroughly.

4. Section 3.3, Fig. 3: Comparison between annealed and unannealed: The authors pointed out that the CrSi2 grains in the non-annealed sample showed Fe impurities, while they were almost negligible after annealing. Could you plot the close-up view (6-7 KeV range) of both spectra in one graph for the comparison?

5. Some diffraction patterns and images are not of high-quality. In my experience, the TEM sample was too thick, which is quite evident from the diffraction patterns as well as features in the BF-images.

Author Response

Comments and Suggestions for Authors

This manuscript focuses on the phases and interfaces in the Cr–Fe–Si ternary system by X-ray and electron diffraction study. The article is well written and provides an in-depth structural analysis of the ternary phases (microstructure, phase composition, and crystallinity). The overall design, including methodology and results, is reasonable and would be of great interest to the readership of Inorganics. I would recommend this article for publication in Inorganics after minor modifications.

1. In the introduction, the authors pointed out that "Cr-Fe-Si" can be utilized as anode materials for Li-ion batteries. I am not entirely convinced by their claim; a number of papers published on the subject emphasize that FeSi2 does not contribute to capacity because it is largely unreactive or inactive in terms of storing Li-ions. The results reported here highlight that they have a significant amount of FeSi2 in their material system. Any comments on this?

DOI: 10.1002/smll.201906812

Answer: The paper what we referred (https://doi.org/10.1016/j.jpowsour.2015.01.122) is dedicated to the investigation of the SiFeCr (Si-rich) composite material in which the nanoscale Si particles are finely dispersed in matrices, Silicide phases. The authors believed that “this structure can improve the cycle performance because the highly elastic silicide phases (FeSi2, CrSi2) could buffer the volume change of the nano-sized active Si during lithiation/delithiation (cycling)”. It was concluded that “the SiFeCr nanocomposite could be a promising candidate for use as an anode material because of its outstanding capacity retention”. The similar conclusion was done in the paper what was proposed by Reviewer 3 (DOI: 10.1002/smll.201906812): “Possibly, the low capacity c-FeSi2 crystallites could be solid electrolyte interphase forming obstacles and therefore, contribute to impede the consumption of the active channels as well as to protect the core Si phase”. Thus, as far as can be understood, metal silicides are chosen to protect the Si anode and thus indirectly improve the characteristics of Li-ion batteries. We decided necessary to mention the cited paper (https://doi.org/10.1016/j.jpowsour.2015.01.122) in the Introduction since the composition was ternary SiFeCr (Si rich) and important for the for large-scale production.

We are thankful to the reviewer for the mention of the very interesting paper (10.1002/smll.201906812)

2.  Why did you choose the 1037 K temperature and 72-hour annealing time? Did you optimize the time and temperature for the ternary system?

Answer: The time and temperature were chosen as optimal for completing the phase transformation from alpha-FeSi2 (tetragonal) to beta-FeSi2 (orthorhombic) phase for further use in thermoelectric design.

3. Oxygen contamination is very important, especially when the annealing is performed at high temperatures. Did you see any evidence of oxygen contamination in your material, especially from the EDX results? I can sense from Fig. 4 that there could be a slightly smaller oxygen peak. This should be checked thoroughly.

Answer: Melting and solidification was done in the furnace with vacuum of about 0.1 Pa. The obtained samples were annealed in air. The EDXS analysis results on CrSi2 grains (see appendix below) show that the annealed samples contain more oxygen than the unannealed samples. As can be seen, the deconvolution of the Cr-L and O-K lines is questionable, but the annealed contains about 3 to 6 at% oxygen while the unannealed contains less than 1%, if any. These concentrations look still  acceptable for crystal and device manufacturers.

The more details of conditions of materials preparation and characterization are added in the paragraph 2. Materials and Methods.

4. Section 3.3, Fig. 3: Comparison between annealed and unannealed: The authors pointed out that the CrSi2 grains in the non-annealed sample showed Fe impurities, while they were almost negligible after annealing. Could you plot the close-up view (6-7 KeV range) of both spectra in one graph for the comparison?

Answer: It is done. 

1. Some diffraction patterns and images are not of high-quality. In my experience, the TEM sample was too thick, which is quite evident from the diffraction patterns as well as features in the BF-images.

Answer: The thickness of samples was in the range from 30 nm to 150 nm where diffraction patterns were obtained. The simulation of diffraction patterns and comparison with the experimental confirmed this thickness. We do not consider that the quality of SAED patterns was not of high-quality. Also we have to recognize that from a technical point of view, a significant effort must be made to increase the observable surface and to reduce the thickness of the thin samples used in transmissive electron microscopy, a challenge on a particularly fragile raw material to be won in order to have sufficient ranges to correlate the information between neighboring phases and to reinforce the statistical value of the results.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The current version is acceptable.