Factors Controlling the Gallium Preference in High-Al Chromitites
Round 1
Reviewer 1 Report
The authors have analysed chromites/chromitites from a number of occurrences for major and trace elements. They find a positive correlation between the Ga content and Al and a negative one between Ga and Cr/(Cr+Al). The slope of correlation lines is different for the different occurrences. It would have been nice to comment on possible origins.
Instead, the authors use literature data on crystal structure refinements of spinels to argue on the crystal site preferences of Ga (and Al) in their chromites. Although their conclusion is entirely possible (and plausible) – as it is known that MgGa2O4 is an inverse spinel – their own data are not (and can not be) used with confidence to support this claim. I therefore wonder what the purpose of the manuscript in its present form is.
I suggest the authors concentrate on their own results and attempt to interpret these in terms of the T,P,X dependence of Ga partitioning in chromite. There must be reasons for the different correlation lines in their Fig. 2a, which deserve to be exploited. Also, the title of the paper calls for such an attempt.
I have made a number of corrections and comments in the manuscript that, I hope, will be of help for the authors.
Comments for author File: 
Comments.pdf
Author Response
Reply to reviewer 1
Thank you very much for the very constructive comments and the linguistic improvement of this work.
General comment: The authors have analyzed chromites/chromitites from a number of occurrences for major and trace elements. They find a positive correlation between the Ga content and Al and a negative one between Ga and Cr/(Cr+Al). The slope of correlation lines is different for the different occurrences. It would have been nice to comment on possible origins.
Instead, the authors use literature data on crystal structure refinements of spinels to argue on the crystal site preferences of Ga (and Al) in their chromites. Although their conclusion is entirely possible (and plausible) – as it is known that MgGa2O4 is an inverse spinel – their own data are not (and can not be) used with confidence to support this claim. I therefore wonder what the purpose of the manuscript in its present form is.
I suggest the authors concentrate on their own results and attempt to interpret these in terms of the T,P,X dependence of Ga partitioning in chromite. There must be reasons for the different correlation lines in their Fig. 2a, which deserve to be exploited. Also, the title of the paper calls for such an attempt.
Reply: Very constructive comments. We hope the revised manuscript to be substantially improved.
Comment: I have made a number of corrections and comments in the manuscript that, I hope, will be of help for the authors.
Reply: Thank you very much. All corrections noted on the text were made.
Your help is greatly appreciated.
Specific comments
Line 9, comment: convincing explanation?
Reply: It was corrected.
Line 12, comment: R-squared is not the correlation coefficient R. Also, the square of a number cannot be negative!
Reply: It was corrected.
Line 26, comment: "electronics industry" is not an application, but electronic devices may be.
Reply: It was corrected.
Line 38, You should note that these radii are for octahedral coordination. As you argue that Al3+ and Ga3+ may also occupy tetrahedral positions, you might also report radii for tetrahedral coordination.
Is [13] the source of your radii? Newer compilations give slightly different values, for example:
D. Shannon (1976) Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallographica A32, 771–773 OR use the radii given in [4].Reply: This paragraph was re-written and the ionic radii values reported by R. D. Shannon (1976) were used.
Line 42, comment: Partition coefficients are usually dependent upon T, P and X and not a constant.
Reply: It was re-written and corrected.
Line 49, comment: Unfortunately, u is explained only much later.
Reply: This comment was deleted as unnecessary.
Line 83, comment: Are you sure that chromite can be dissolved this way completely?? I have never been able to dissolve Al-Cr spinel from peridotite xenoliths by acids, except, maybe, H2SO4. Maybe a pressure vessel was used by ACME in which case you might mention it.
Reply: As far as we are aware a pressure vessel was not used by ACME. However, a batch of same samples of chromitites were analyzed by SGS Global – Minerals Division Geochemistry Services Analytical Laboratories Ltd., and were dissolved using Sodium peroxide fusion, combined ICP-AES and ICP-MS; the obtained results are comparable.
Line 79, comment: In line 413 you state that this research is based on an evaluation of available data from the literature. If this is true for trace element, including PGE, analyses, reporting analytical details here is misleading. You should simply point the readers to the relevant literature. These references can be placed as a footnote under Table 2.
Reply: It was corrected.
Line 91, comment: Make sure that the spelling is consistent throughout the manuscript (also used are Kempersai, Kemprsai).
Reply: it was corrected.
Line 109, comment: incompatible for what - minerals from peridotite?
Reply: It was clarified and rewritten.
Line 120, comment: Make sure that the spelling is consistent throughout the manuscript (also used are Kempersai, Kemprsai).
Reply: It was corrected.
Line 137, comment: why not report the high and low values in both, Cr2O3 and Al2O3? - Otherwise, report the high and low values of Cr2O3 only.
Reply: It was re-written and corrected.
Table 1, comment: It would be helpful in a paper like this one, to provide a column "cations based on 4 (or 32) oxygens".
Reply: They are provided now. The Fe3+/(Al+Cr+Fe3+) atomic ratio was plotted versus Ga and this diagram is discussed.
Table 1. Comment: Move up "Urals, Kempersai" one line. Otherwise it seems to belong to the Rhodope massif.
Reply: It was corrected.
Table 1, Comment on the sample from Gomati: For this sample I calculate Cr/(Cr+Al) = 0.44!
Reply: Thank you, It was corrected.
Table 1, comment: Do these high SiO2 concentrations make any sense??
Reply: SiO2 was deleted from all analyses as unnecessary and the analyses recalculated without this oxide.
Table 1, comment: How did you distinguish between Fe2+ and Fe3+ – by stoichiometry?
Reply: Contents of Fe2O3 and FeO were calculated on the basis of the spinel stoichiometry (Section 2 . Materials and Methods).
Lines 156, 159, 160, comment on R2.
Reply: They were all corrected.
Table 2, comment: A note on typical analytical errors, in particular for Ga, would be in place here, for example ±5%. I guess ACME provides such estimates.
Reply: This information is provided in the Chapter 2 now.
Table 2, Comment: I understand that PGE analyses and the other trace element analyses have been acquired in different projects. Is it certain that both analyses have been carried out on the very same samples? You should comment on this, possibly in chapter 2.
Reply: It is clarified now (Chapter 2) that although the provided PGE data (Table 2) have been published in previous studies, and that a portion from the same samples was used for the presented trace element analyses.
Table 2, comment: ppb, I suppose
Reply: Yes, it was corrected.
Figure 2a, Comment: For Kempirsai you have essentially a two-point regression line (2 points closely together at the high end and two at the low end). Also note that the significance of a correlation coefficient is a function of the sample number. For 4 samples you need a much higher value than for eight samples for the correlation to be statistically significant at a given level, for example 99%.
Reply: You are right, but we would like to give this trend because the composition was homogeneous, as throughout many large chromite deposits. For this reason, in the next figure (2c,d), the samples from Vourinos and Kempirsai were plotted as a group.
Figure 2, comment: Can you change the colours of the correlation lines to match those of the symbols?
Again, correct R^2.
Reply: They were corrected.
Figure 2b, comment: what about the 2nd figure?
Reply: It was corrected.
Line 174, Comment: If this has been established, then the chromitites ARE derived from boninitic magmas...
Otherwise: It has been suggested... It has been argued...
Reply: It was corrected.
Reply: It was corrected.
Line 179, Comment: I am not convinced that you can use Pd/Ir as an indicator of magma fractionation to match Ga or Cr/(Cr+Al) of spinel. PGE are chalcophile and/or siderophile elements whereas Al, Cr and Ga are lithophile. – On page 5 you write that Pd/Ir is independent of major element composition!
Reply: This paragraph was re-written.
Line 183, comment: Are the chromitites affected by contaminated magmas or do contaminated magmas crystallize (a variety of) chromitites?
Reply: It was re-written
Line 204, comment This is already implied by the first sentence of this paragraph.
Reply: it was deleted
Line 206, comment Why do you call this (and M-O) a site? The site is T (or M). T-O should denote a bond length.
Remember that you have explained the spinel structure as a cubic close packing of oxygens with tetrahedral and octahedral vacancies partially filled by cations.
Reply: It was corrected.
Line 207, comment You mean ions (or cations)?
Reply: It was clarified: cations
Lines 226, comment at least for the end-member composition MgGa2O4
Reply: It was corrected
Line 228, comment Again, reporting a single value for D is an oversimplification. Either report a range of provide information on T, P, X for D = 4.6.
Reply: This sentence was deleted. This is the topic at the first part of the discussion.
Line 234, comment spell out or define as you use this also in Fig. 3
Reply: In was deleted
Line 236, comment: Martignago et al. [50] performed crystal structure refinements on three natural spinels.
Reply: It was corrected
Line 245, comment: I do not understand the meaning of this sentence.
Reply: It was deleted.
Line 246. Comment spell out or define as you use this also in Fig. 3.
Reply: It was defined.
Line 272,comment: It is not clear to me why the correlation between Ga and Al would during re-equilibration. Also, what do you understand by re-equilibration? Do you mean sub-solidus exchange with other minerals (which is sluggish) or reaction with melts/fluids?
Reply: The term re-equilibration used with the mean of the sum-solidus exchange.
Line 278, comment: Which minerals co-existing with spinel might take up the Ga?
Given that Ga is a highly compatible element when co-existing with silicate melt, you would need minerals with even higher D values.
Reply: These conclusions were re-written.
Thank you
Author Response File: 
Author Response.docx
Reviewer 2 Report
Review comments on the manuscript Minerals-578542: Controlling factors on the gallium preference in high-Al chromitites by I.-P. D. Eliopoulos & G. D. Eliopoulos
The manuscript aims to address the incorporation of gallium (Ga) into chromitite by examining chromitite samples from a few localities. Scanning electron microscope (SEM) with energy dispersive spectrometer (EDS) and inductively coupled plasma mass spectrometer (ICP-MS) were employed to analyze the samples. The mineral chemistry and trace element geochemistry of the bulk rock were then used to discuss the preference of gallium in chromitites.
The major concerns of the manuscript are related to the approach and analytical techniques being used here. Samples from several geological localities were analyzed, but contained little detailed petrographic observation and detailed sample chemistry. SEM-EDS is known for its capability of semi-quantitative analysis; the authors used it for mineral chemistry analysis, therefore, cannot yield quantitative results, which is critical for the goal of the project.
Additionally, to make a strong contribution, the authors need to further polish and streamline the writing of the manuscript along with correction of the minor errors in the manuscript.
Regards,
Major comments:
The major concern of this manuscript is the research approach and analytical methods used. The authors studied chromitite from several geological localities, without giving detailed petrographic documentation of all the samples (although there are four BSE images of samples from Greece). However, the topic on gallium incorporation in chromitite needs to take into consideration of all other potential minerals present in the samples, before reaching to any conclusion of Ga-Al correlation. Therefore, the manuscript will need to be significantly improved with more detailed petrographic description of the studied samples.
Scanning electron microscopy along with EDS is typically used for semi-quantitative analysis. Electron microprobe (EMP), which utilizes wavelength dispersive spectrometer (WDS) instead of EDS, is used for quantitative mineral chemistry. Therefore, mineral chemistry, which is essential to discuss the incorporation of elements into crystalline structure, should use EMP to analyze. Due to the use of SEM-EDS and inaccurate datasets from the instrument, it would introduce errors into the interpreteation on mineral chemistry and element incorporation in chromitite.
Since Ga incorporation in chromite might be critical, why not using laser ablation-ICP-MS to analyze specific chromite? This could help better constrain the Ga in chromite, which is useful for discussing the correlation of Al-Ga in later part of the manuscript.
Detailed comments
Line 25–27: Since the manuscript is more on Ga, it is better for the manuscript to start directed with this element. It might be better to start with a sentence as the following: Gallium (Ga) is a vital element for the economy due to their application in high-technology fields, such as electronic industry, electric cars and solar panels.
Introduction part: the introduction does not seem to be well-written. A more specific niche needs to be created on why the studied was conducted (i.e. what is research all about?). Some other issues could also be discussed: for example, what is the current understanding of Ga incorporation in chromitite? (Does Pauling’s rules work well in the ionic substitution?)
Line 54–59: This long sentence tries to address the research subjects and methods in this manuscript. It is difficult to read because of mixing research subjects and methods. Therefore, I would suggest to separate into multiple sentences. Additionally, optical microscopy and SEM-EDS cannot fully quantify mineral chemistry (See major comments).
Line 69–70: Which kind of detector does the SEM-EDS have? Silicon drift detectors (SDD) could be used to approach more quantitative analysis, claimed by some researchers. More commonly, petrologists/mineralogists uses electron microprobe for mineral chemistry analysis, so why not using electron microprobe here?
Line 70–74: What about the EDS detector, which is critical for determining concentration of elements at low level (<1 wt%)?
Line 75: The beam current seems to be low for SEM/EDS analysis. Did you double check this?
Line 77: This is not a sentence.
Line 77–78: Do you mean that ‘Ferric iron was calculated using charge balance based on spinel stoichiometry’? It seems to me that you have analyzed total FeO using FeKα.
Line 80: Is the ICP-MS used to analyze major elements? Which major elements did you analyze? Reported anywhere?
Line 127–128: Figure 1. Abbreviations of the minerals should be noted in the figure caption or elsewhere in the manuscript. What are the silicate inclusions in Fig. 1b?
Line 142–143: Revise to “Major element composition of chromite in chromitites from Greece, Bulgaria and Kempirsai (Urals)”. What are the textural relationship of these analyzed chromite (i.e. core, rim, inclusions or matrix)?
Line 166: Table 2. What does the wt% apply to? Only Fe? The Table needs to be revised to make this clear.
Author Response
Reply to Reviewer 2
Thank you for the comments and recommendations.
General comment: The major concerns of the manuscript are related to the approach and analytical techniques being used here. Samples from several geological localities were analyzed, but contained little detailed petrographic observation and detailed sample chemistry. SEM-EDS is known for its capability of semi-quantitative analysis; the authors used it for mineral chemistry analysis, therefore, cannot yield quantitative results, which is critical for the goal of the project.
Reply: Detailed petrographic observations and detailed sample chemistry have been published in previous studies, since all chromitites studied herein come from very well-known ophiolites complexes. The provided SEM-EDS analyses for the purpose of the present study are semi-quantitative indeed and that is the reason we provide the results with one only point. We used the available analytical facilities, but we think that we can use the obtained data for the purpose of this study.
Comment: Additionally, to make a strong contribution, the authors need to further polish and streamline the writing of the manuscript along with correction of the minor errors in the manuscript.
Reply: Following this comment additional data are provided and large parts of the manuscript were re-written (red color).
Detailed comments
Line 25–27, comment: Since the manuscript is more on Ga, it is better for the manuscript to start directed with this element. It might be better to start with a sentence as the following: Gallium (Ga) is a vital element for the economy due to their application in high-technology fields, such as electronic industry, electric cars and solar panels.
Reply: It was corrected.
Introduction, comment: the introduction does not seem to be well-written. A more specific niche needs to be created on why the studied was conducted (i.e. what is research all about?). Some other issues could also be discussed: for example, what is the current understanding of Ga incorporation in chromitite? (Does Pauling’s rules work well in the ionic substitution?)
Reply: In the section of Introduction an attempt was made to include in the interpretation of the presented data the T,P,X dependence of Ga partitioning in chromite, in accordance to the title of the paper.
Line 54–59, Comment: This long sentence tries to address the research subjects and methods in this manuscript. It is difficult to read because of mixing research subjects and methods. Therefore, I would suggest to separate into multiple sentences. Additionally, optical microscopy and SEM-EDS cannot fully quantify mineral chemistry (See major comments).
Reply:This paragraph was re-written and the investigation methods were omitted as unnecessary here.
Line 69–70, comment: Which kind of detector does the SEM-EDS have? Silicon drift detectors (SDD) could be used to approach more quantitative analysis, claimed by some researchers. More commonly, petrologists/mineralogists uses electron microprobe for mineral chemistry analysis, so why not using electron microprobe here?
Reply: This is the only choice we have, only the SEM-EDS system is available to our labs. Nevertheless, the obtained results seem to be consistent with the bulk rock analyses, as they are massive chromitites (0.95 vol%).
Line 70–74, Comment: What about the EDS detector, which is critical for determining concentration of elements at low level (<1 wt%)?
Reply: please, see the above answer.
Line 75, comment: The beam current seems to be low for SEM/EDS analysis. Did you double check this?
Reply: Yes, it was double checked.
Line 77, comment: This is not a sentence.
Reply: Sorry, this comment is not clear to us.
Line 77–78, comment: Do you mean that ‘Ferric iron was calculated using charge balance based on spinel stoichiometry’? It seems to me that you have analyzed total FeO using FeKα.
Reply: In the section of 2. Materials and Methods, lines 77-78, was clarified that the Fe2O3 and FeO contents of were calculated on the basis of the spinel stoichiometry.
Line 80, comment: Is the ICP-MS used to analyze major elements? Which major elements did you analyze? Reported anywhere?
Reply: Among major elements were analyzed Cr, Mg, Al, but they are not given as unnecessary, because the result is effected by the associated silicates with chromite, such as silicate inclusions and/or serpentine and chlorite in chromitites. The reported atomic ratios of Cr/(Cr+Al) and Mg/(Mg+Fe2+) from the mineral chemistry provide the required information for the purpose of this study.
Ca and a a serious of trace elements analyzed are reported in the chapted 2 now, but all of them are lower than detection limit. Thus, we think that the only interesting major element is the reported Fe content.
Line 127–128, comment: Figure 1. Abbreviations of the minerals should be noted in the figure caption or elsewhere in the manuscript. What are the silicate inclusions in Fig. 1b?
Reply: It was corrected.
Line 142–143, comment: Revise to “Major element composition of chromite in chromitites from Greece, Bulgaria and Kempirsai (Urals)”. What are the textural relationship of these analyzed chromite (i.e. core, rim, inclusions or matrix)?
Reply: They have been described in previous publications and references are provided: [17. 18, 37].
Line 166: Table 2. What does the wt% apply to? Only Fe? The Table needs to be revised to make this clear.
Reply: It was corrected to make clear that is only Fe.
Kind regards
I-P. Eliopoulos
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors have done a reasonable job in improving their manuscript. Although I had recommended that the focus of the manuscript be changed and the authors have not followed my advice, I have no objections in seeing the manuscript published after minor modifications as indicated in the commented pdf file.
Comments for author File: Comments.pdf
Author Response
Thank you very much for your comments and recommendations on our revised manuscript. All corrections in the commented pdf file were done and they are marked in red color in the new revised version.
We tried our best to match your corrections, the first parts of the discussion and conclusions were re-written with emphasis to factors controlling the spinel chemistry.
Many thanks once again for your valuable help to our first publication, as a contribution to the memory of our father.
Best wishes
Ioannis-Porfyrios Eliopoulos
All corrections in the commented pdf file were done and they are marked in red color in the new revised version.
Comments
Line 13: Be aware that -0.98 is smaller than -0.95 and 0 is certainly larger.
Reply: It was re-written as: R ranges from -0.95 to -0.98.
Line 22: Make two sentences out of this. The 2nd sentence should begin with: In contrast, the competing Cr3+...
Reply: It was re-written.
Line 42: Do you mean the size of the site or the ionic radii?
Reply: It was clarified that it is the ionic radii.
Line 48: at very low oxygen fugacity
Reply: It was corrected
Line 51: either Fe3+ or Fe(III)
Reply: It was corrected
Line 66: It is not exactly clear to me what you mean – experimental literature data relevant for the structure of spinels?
Reply: It was re-written.
Line 85: I am not sure that I understand this.
You write that the samples were decomposed in an acid mix. The dry residue was dissolved in HCl, evaporated again and dissolved again in HCl. Also, is "heated to fuming and taken to dryness" something important and different from "evaporated to dryness on a hotplate"?
Reply: It was re-written and the repetition was omitted: The samples were dissolved using an acid mix (HNO3–HClO4–HF) digestion and then the residues were dissolved in concentrated HCl.
Line 88: The order to elements from U to K is a bit strange. Why not order them by atomic number?
Reply: The order by atomic number was corrected: The rare earth elements La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb and Lu as well as Li, K, Ge, Sr, Y, Zr, Mo, Sb, Cs, W, Pb, Th, and U.
Line 117: As I noted previously, the term incompatible is misleading without further qualification. All PGE are certainly highly compatible for the mantle as long as (Ni rich) sulfides or (Fe) metal alloys are present in peridotite. However, if sulfide is lost through partial melting, the PPGE will go into the melt also, thus behaving incompatibly.
If sulfides and alloys precipitate from a magma (and maybe get trapped in spinel) my guess is that all PGE will be strongly removed from the melt (IPGE more than PPGE) which means all PGE are compatible.
Reply: It was corrected and re-written.
Lines 184-187, Figure 2: It was re-written.
Line 197: In my opinion it does not matter whether one element is incompatible and another one compatible. It is just important that you use two elements with highly different D values for the process you have in view, such as Pd and Ir.
Reply: It was corrected and re-written.
Line 317: You should either spell out the full names of the journals for all references or use the short names only.
Reply: Full names of the journals for all references are given now.
Thank you
Author Response File: Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
In the submitted manuscript, the factors controlling the presence of the critical metal Ga in high-Al chromitites are approached. Emphasis is given on the relationship between Ga and Al in natural spinels from selected ophiolites in Greece. Original data is given and the overall study is of interest to the scientific community.
However, parts of the manuscript need further clarification and revision.
For example. it is not very clear to the reader which data are original and which come from the literature. If the geochemical characteristics of the greek chromitites given in Table 2, are all from the literature referred to in the brackets, then which whole-rock analyses were performed in ACME Laboratories (Section 2.2). Are some of the analyses given in Table 2, performed by the authors? Please clarify tis point. Besides, I can't see the elements Cu, As, Ca mentioned in the Materials and Methods, given and commented in the manuscript. On the other hand, Sc is commented in line 106, but no analysis of it is given in the manuscript.
Conclusions are very brief and of poor wording. Please rewrite this section.
Specific comments/suggestions:
line 10: please change explanation in with explanation for
line 11: please change in this present study with in the present study (or with in this study)
line 12: please write a strong negative correlation (R2≥-0.90)
line 51: please change understanding og with understanding of
line 55: please replace comma with and (and scanning electron microscope)
line 61: why is the As X-Ray line is given here? There is no As content given in the tables. please delete
line 63. Please add which standards were used for the other elements (eg Al, Ti, Ca, Si, ...)
line 76: please put the sentence (up to 63.2 ...ratio 0.81 in square brackets
line 77: please put the sentence (up to 34.5 ....ratio 0.4) in square brackets
Table 1: please write how trivalent iron was determined (?measured?)
line 85: please put a dot after [7]. Delete which.
line 90: please change while with whereas and any with no
line 91: please delete not
Figure 1: please add (b) after Tibet in the legend
line 100: please change trend on with trend in. Write a decreasing trend.
line 103: what do you mean by modified cores?
line 106: please write Co, Mn and Zn with Co, Mn, Zn (delete and)
line 111: difficult meaning. please clarify.
line 114: please add in magmas. Please write a trend in instead of a trend on.
line 115: please write depletion in instead of depletion of
line 116: please replace extent with degree
line 117: please write the described one
line 124: please replace forms with with
line 125: please replace every with each
Table 3: in the table write parameter (instead of paramete)
line 159: please add in before the mobilisation
line 182-184: difficult meaning
line 190: replace contents with content. Write ...Cr content in spinels affects the occupancy...
line 198. replace contents with content.
Figure 2: add as a function (in the legend)
line 209: what do you mean by magmatic structure ? better magmatic composition?
line 218: please add in after resulting
I suggest manuscript publication after the authors take into consideration the above mentioned comments.
Reviewer 2 Report
see attached pdf
Comments for author File: Comments.pdf
