I don't do research on Fe isotopes, so I can't say much from first-hand knowledge, but a somewhat recent paper Ni et al. (2020; Nat. Geosci., doi: 10.1038/s41561-020-0617-y) indicates that Fe-57 can be fractionated from Fe-54 so that solid Fe would have higher 57Fe/54Fe than liquid Fe.

Thank you for the very kind comment.

Thank you for the comment. I'll keep that in mind for future recordings (and possibly re-record some of these).

Be sure to check my calculations but I think to convert Fe2O3 to elemental Fe, you would multiply by 0.4499055, and then the inverse of that number should be the conversion of Fe to Fe2O3

This is actually an excellent question and the problem is that the cell G8 is a relict from an earlier version of the sheet. I experimented with allowing T or P to vary with Fe2+/Fe3+ (as it likely does) but the parameter was not statistically different compared to leaving it out. So the short answer is "no" it doesn't affect the models (and I need to delete it from the spreadsheet). Thank you for finding that necessary correction.

It might be easiest for me to find the problem if you send me a copy of your spreadsheet via email: [email protected]

Are you coding it yourself, or are you using the spreadsheet that is available here? : sites.google.com/mail.fresnostate.edu/keithputirka/home

Sorry I didn't see your answer... I am using the spreadsheet available in your website. There is something wrong with the iterative calculations, maybe it's about my version of excel. What do you suggest for this? Thank you for all your work and thank you for sharing with us by the way, this is the most helpful.@@keithputirka

That's a renormalization on a volatile-free basis of estimates of the Solar photosphere composition ("Sol" for short"); I believe that particular composition is the "recommended" value from Lodders et al. (2009).

Yes, you are correct. 4 m, not 3m. I thought I re-recorded and re-posted this video, but I must have re-posted the old one? Anyway, I'll post a note in the description. Thank you for pointing this out.

@@keithputirka You are welcome. Overall I like your videos. They are informative and cover good topics.

In the conversion table that appears in in columns P to AA, you would use row 6 instead of row 5, so you would take oxides (which could be entered in columns B-L instead of elements) and multiply those oxides by the correction factors in row 6. Then be sure to edit the headers, so if you are performing the multiplication in columns Q to AA, then the header in cell Q9 becomes "Element, weight proportion" (instead of "Oxide, weight proportion"), and cell A9 wold be changed to "Oxide %".

Hi Amina - I later realized that this video says nothing about how to deal with Oxygen. I've recorded (and will soon post) a new video to illustrate the conversion of oxides to elements, and also how to deal with O, including some caveats for systems that are possibly very O-rich. Thank your for your question. And my apologies for not thinking it through earlier.

Thank you for that correction. I'll add a note to the description

If I understand you correctly, then yes, you want to know the charge of the element, since it affects the radius of the atom. However, be aware that there is circularity in that for a given charge, the radius does depend upon the coordination. So in a real-life problem there is a bit of guesswork to analyze the structure and see what works.

@@mineralogy so if I’m meant to compare the radii of Al3+ & O2-, would I multiply the radii due to needing more to balance, or is that taken care of as Al3+ already has the different Radius? Sorry if my wording is difficult to understand.

@@sethlopez8287 if you mean, "does it matter that we need two Al for every 3 O, to create Al2O3", the answer is 'no'; the numbers of atoms per formula unit (what we sometimes call the "stoichiometry") are not a part of the calculation.

That's a great question and probably requires a new video, which I'll try to record soon. In brief, An-Ab is a complete solid solution because there is no limit on the amount of Ab that can dissolve into An or vice versa. But An-Or is (highly limited) since only small amounts of Or can dissolve into An and vice versa (and Ab-Or is nearly complete at high T, but increasingly limited as T decreases).

I didn't present the derivation but it's purely mathematical - a matter of using trigonometry to convert the triangular coordinates to x-y coordinates.

This video does not show the derivation, but gives the basic idea: kzbin.info/www/bejne/eIaap2qfjrRrbbs

Could you explain me or give me some guidance for the derivation please? ):

Hi Luis - I like to assign this as an extra credit to my students, so I don't want to say too much, but the trick is to pick an arbitrary point within the triangle and then try drawing triangles with sides that run parallel to the x-y axes and triangles whose sides are the lines that bisect the apices of the triangle. The ternary units give you the known lengths along those bisectors. So you want to use trig to take the known lengths of the bisectors to determine the lengths of the sides that parallel x and y. I hope that helps.