update atomic masses

[rowlesmr]

Also changed key to element symbol. Need to update core enumeration to match.

They are the abridged values from https://doi.org/10.1515/pac-2019-0603

The masses for the radioactive elements are the masses of the isotope with the longest halflife as detailed https://iupac.qmul.ac.uk/AtWt/

The mass for deuterium was taken from https://www.ciaaw.org/atomic-masses.htm and given to the same decimal places as hydrogen.

core to be updated if accepted.

[rowlesmr]

Probably the most potentially controversial change is Tc. I don't know how Tc-chemists do their thing, so there may be a preferred value. But should be documented in core as to what is going on.

I've contacted Gordon T and Brendan K for their opinions.

The maxiimum differences are

Z    element      new        templenum    diff   
61    Pm          145          147        -2     
94    Pu          244          242         2    
98    Cf          251          249         2    
43    Tc           97          98.906     -1.906 
93    Np          237          237.048    -0.048 
32    Ge           72.63        72.59      0.04  
76    Os          190.23       190.2       0.03  
88    Ra          226          226.025    -0.025 
22    Ti           47.867       47.88     -0.013
34    Se           78.971       78.96      0.011 

with the remainder haved an absolut difference of 0.01 or less.

[vaitkus]

Thank you for the work, but this does not seem like a backward compatibility breaking approach since other dictionaries outside the IUCr may also be importing these values. Small changes would generally be ok, but this PR completely re-keys the enumeration set.

A more conservative approach would be to just create a new save frame (e.g. atomic_symbol_to_atomic_mass) and add the new values there. This way, the importing dictionaries can choose to migrate to the new version or not.

[rowlesmr]

no dramas. Can fix

[vaitkus]

Thanks for humoring me, the PR looks good now. The next step would probably be to agree in coreCIF if the default values should be re-keyed on the element symbol. For now, I would probably designate this PR as a draft to not accidentally preemptively merge it, but have it at hand.

[rowlesmr]

Brendan says most chemistry is done on 99-Tc extracted from fuel rods. Using 99-Tc in experiments is backed up by https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/ejic.202400780

[rowlesmr]

Thanks for humoring me, the PR looks good now. The next step would probably be to agree in coreCIF if the default values should be re-keyed on the element symbol. For now, I would probably designate this PR as a draft to not accidentally preemptively merge it, but have it at hand.

Ta a lot :)

I've put in PRs in core, and marked those as draft.

[vaitkus]

I've put in PRs in core, and marked those as draft.

Thanks, but I was actually trying to suggest it the other way around. Mark PRs on the Enumeration_Templates as draft and leave the coreCIF ad regular ones to make it clear they are open for discussion. But I guess this works too. The important thing is not to merge in things we are not ready to merge in.