On the Arrangement of Pentagonal Columns in Tetragonal Tungsten Bronze-Type Nb18W16O93
Round 1
Reviewer 1 Report
The manuscript is a good contribution to its field and fits the journal well. Minor modifications are recommended:
1) Line 68: "could never been confirmed" (please fix English here)
2) Introduction, end: for easier reading and to facilitate understanding of this manuscript by the reader, it is recommended that the author add a few sentences that clearly indicate the novelty and originality of this manuscript. Just a few very brief phrases to specify the novelty and originality of this manuscript/study
Author Response
1) Line 68: "could never been confirmed" (please fix English here)
Changed to: “could not be confirmed”
2) Introduction, end: for easier reading and to facilitate understanding of this manuscript by the reader, it is recommended that the author add a few sentences that clearly indicate the novelty and originality of this manuscript. Just a few very brief phrases to specify the novelty and originality of this manuscript/study
While in the first paragraph of the introduction, the TTB structure is presented, the second one refers to the outstanding electrochemical properties of the sample Nb18W16O93 and its yet inappropriate structural characterization. I am sure that it is already clear from this that the structural investigation by a thorough analysis of HAADF-STEM images reported here sheds new light on this topic. However, I am stressing this point further by substituting the last sentence of the introduction by:
No evidence for the hitherto assumed structural model for Nb18W16O93 was found in HAADF-STEM images of a sample with this composition. Instead, the observed occupation of PTs points to a complex intergrowth of two phases as reported in this article.
Reviewer 2 Report
The manuscript describes a detailed interpretation of several lattice resolved electron micrographs recorded in the annular dark field (ADF) STEM mode of domains in a tungsten-bronze type oxide structure, which is quite interesting and certainly appropriate to this journal. The manuscript is well written and contains sets of beautifully assembled model structures next to the corresponding experimental images.
A few questions have arisen when reading the text and should be addressed in a revision, along with three minor grammatical corrections outlined at the end.
scientific points:
▪ lines 161-165: ‘The magnified area (inset in (a)) demonstrates the varying dot intensities with white arrows pointing to W-rich and yellow ones to Nb-rich positions, respectively. As the HAADF-STEM method generates images with the intensity I increasing with the atomic number Z (I ~ Z2; Z contrast imaging [23]), the metal positions appear as bright dots with their brightness increasing with the W content of the particular atomic column (ZW = 74 > ZNb = 41) [19,21].‘ This section contains two implicit assumptions.
i. The quoted square dependence is a large-angle approximation for Rutherford-type scattering of single atoms; this should be stated. In a crystal the intensity of an atomic column will have a power dependence on the atomic number with a power factor (exponent) somewhere between 1.2 and 1.8, depending on voltage and inner detector angle, and only converge towards 2 for very large angles as shown by Bleloch, Walther and others.
ii. ADF STEM intensity will also increase with the number of atoms, which is why fractional fill factors are so important, as shown by Pennycook, Browning and others.
▪ Lines 197/198: ‘the decreased O content that must be somehow adjusted by structural adaptions’ Does this stoichiometric deviation create strain within the structure and if so, could this be visualised as well by simultaneous STEM image acquisition at lower collection angles (BF, ABF or low angle ADF)?
technical points:
• lines 12/13: ‘…the O/∑M was determined to be 2.755 which is smaller than the value of Nb18W16O93 which is O/∑M = 2. 735.’ Surely the comparison should read ‘bigger than’?
• lines 84/85 and 88/89: The synthesis and the electron microscopy should be described briefly in one or two sentences each. For the data interpretation the specimen thicknesses, the electron beam convergence angle and the angular collection range of the ADF detector should be stated as these are important factors when it comes to considering the 2-dimensional projection of the 3-dimensional structure as mentioned above, in particular if the statement in lines 204/205 that ‘nanometer-sized domains [are] coherently intergrown with each other. ’ also applied along the electron beam direction.
• lines 234/235: I do not know how difficult sample preparation and ADF STEM imaging were in this particular case but would encourage the author to list the person(s) who did the experimental work as co-author(s) rather than giving the lab director an acknowledgement for ‘providing … images’.
grammar:
▪ line 68: ‘could never been confirmed’-> either ‘has never been confirmed’ or ‘could not be confirmed’
▪ lines 100 and 124: ‘in respect to’ -> ‘with respect to’
▪ line 212: ‘An example is the Nb6W8O39 structures’ -> use singular form: ‘structure’
Author Response
scientific points:
▪ lines 161-165: ‘The magnified area (inset in (a)) demonstrates the varying dot intensities with white arrows pointing to W-rich and yellow ones to Nb-rich positions, respectively. As the HAADF-STEM method generates images with the intensity I increasing with the atomic number Z (I ~ Z2; Z contrast imaging [23]), the metal positions appear as bright dots with their brightness increasing with the W content of the particular atomic column (ZW = 74 > ZNb = 41) [19,21].‘ This section contains two implicit assumptions.
i. The quoted square dependence is a large-angle approximation for Rutherford-type scattering of single atoms; this should be stated. In a crystal the intensity of an atomic column will have a power dependence on the atomic number with a power factor (exponent) somewhere between 1.2 and 1.8, depending on voltage and inner detector angle, and only converge towards 2 for very large angles as shown by Bleloch, Walther and others.
Of course, the reviewer is right that the relation of the intensity to the atomic number is only approximated by the exponent of two. Nonetheless, this relation is widely used for the qualitative explanation of the HAADF-STEM imaging contrast ([20-22] and new [24]). The validity of this approach for Nb-W oxides was proven by the comprehensive study of Kirkland and Saxton on Nb8W9O47 that revealed an inhomogeneous distribution of Nb and W on certain positions [20] An advantage of HAADF-STEM imaging is its small dependence of the contrast on sample thickness and defocus so that direct image interpretation is possible (see [23] and new [24]).
ii. ADF STEM intensity will also increase with the number of atoms, which is why fractional fill factors are so important, as shown by Pennycook, Browning and others.
Thickness variations are not present here (stressed by the inclusion: …in regions of constant thickness..., line 169): the continuous brightness in the rather large area shown in Figure 5a confirms a rather homogeneous thickness of the whole observed area. This exclusion of a major influence of varying thickness allows the rough correlation of the intensity of single spots to the occupancy by Nb and W as described in the manuscript.
▪ Lines 197/198: ‘the decreased O content that must be somehow adjusted by structural adaptions’ Does this deviation create strain within the structure and if so, could this be visualised as well by simultaneous STEM image acquisition at lower collection angles (BF, ABF or low angle ADF)?
As the underlying TTB substructure is perfect, only the occupation of PCs is affected by the non-stoichiometry. This leads to some local distortions of the TTB structure as filled PTs are more regular pentagons as empty ones. Even in more disordered Nb-W oxides with incorporation of planar defects in the TTB structure, strain is not observed, see e.g. our recent article: Structural developments during the low-temperature oxidation of Nb7W10O47. J. Solid State Chem. 2021 https://doi.org/10.1016/j.jssc.2021.122430
technical points:
- lines 12/13: ‘…the O/∑M was determined to be 2.755 which is smaller than the value of Nb18W16O93 which is O/∑M = 2. 735.’ Surely the comparison should read ‘bigger than’?
Of course. Corrected.
lines 84/85 and 88/89: The synthesis and the electron microscopy should be described briefly in one or two sentences each. For the data interpretation the specimen thicknesses, the electron beam convergence angle and the angular collection range of the ADF detector should be stated as these are important factors when it comes to considering the 2-dimensional projection of the 3-dimensional structure as mentioned above, in particular if the statement in lines 204/205 that ‘nanometer-sized domains [are] coherently intergrown with each other. ’ also applied along the electron beam direction.
For the experimental details of synthesis and electron microscopy, I can only provide the description given in reference [12]. I asked the authors for the same values as the reviewer some time ago but got no answer.
As shown and discussed in previous papers (e.g. [20-22]), the structure of Nb-W oxides is often disordered in the ab plane but always well-ordered along the short crystallographic c axis, i.e. along the electron beam direction. Thus, the 2D projection allows to construct reliable 3D structural models.
lines 234/235: I do not know how difficult sample preparation and ADF STEM imaging were in this particular case but would encourage the author to list the person(s) who did the experimental work as co-author(s) rather than giving the lab director an acknowledgement for ‘providing … images’.
Being an electron microscopist myself, I understand the objections of the reviewer and share his view that the STEM work is a scientific contribution to this article. Unfortunately, it is impossible to get any response from the research group which performed the original study (reference [12]) After I had received the STEM images with the permission to use those, I tried to stay in contact with them but never got any reply to several emails. Firstly, I informed the group about the insights retrievable from these images and asked them whether they were interested in a cooperation on this field, later sent them some of my interpretations for discussion and finally updated them that I was going to publish an article using these data. All emails remained unanswered.
grammar:
▪ line 68: ‘could never been confirmed’-> either ‘has never been confirmed’ or ‘could not be confirmed’
Changed to: “could not be confirmed”
▪ lines 100 and 124: ‘in respect to’ -> ‘with respect to’
Changed
▪ line 212: ‘An example is the Nb6W8O39 structures’ -> use singular form: ‘structure’
Changed
