Are the following notations official? Can I find them in a colored IUPAC book? Do they have a name, I could google for? (I need this to judge whether to support this with mhchem or not.)

  1. Stability of carbocation intermediates: secondary alkane versus primary allyl?

    • < as in $\ce{CH3+ < CH3CH2+}$
    • ≈ as in $\ce{C6H5CH2+ ≈ (CH3)3C+}$
  2. Is Buckminsterfullerene aromatic or not?

    • @ as in $\ce{M@C28}$
  3. Confused about parity of atom and chirality

    • @@ as in $\ce{O[C@@](F)(Cl)Br}$
  4. Why does hydrogen so easily take the place of oxygen in organic molecules during reduction reactions? and What are the mechanisms for capture of As(III) and As(V) by magnetite from water?

    • > as in $\ce{>C=O}$
    • > as in $\ce{>Fe-O-AsO^{3-}-O\bond{...}HO-Fe<}$

    (This seems to differ from 1. Is this a bond? IUPAC?)


2 Answers 2


It really is tricky to judge what is official and what is not. Here is my take on the several issues.

  1. This is not something that should be implemented in mhchem, as I think it is just sloppy typesetting. The used symbols are binary mathematical operators like “=” just as in $3<4$ and $\pi\approx3.14$. Therefore the correct typestting should be
    • $\ce{CH3+} < \ce{CH3CH2+}$ as in $\ce{CH3+} < \ce{CH3CH2+}$ (meaning the methyl cation is less stable than the ethyl cation in this case)
    • $\ce{C6H5CH2+} \approx \ce{(CH3)3C+}$ as in $\ce{C6H5CH2+} \approx \ce{(CH3)3C+}$ (meaning the tert-butyl cation is approximatly as stable as the benzyl cation in this case)
  2. The notational usage of @ as in $\ce{M@C28}$ is indeed very common for endohedral fullerenes. (Maybe not on this site, but if you search for endohedral fullerene on scholar you will get many hits.) I don't know about an official IUPAC recommendation; the blue book covers general aspects of fullerene nomenclature, but not endofullerenes. See comment by Loong for exact details.
    As I don't know about the internals of mhchem, I don't know how this character would or should be treated, I assume, from the looks of it, it works as it should.
  3. The third example is SMILES notation. More specific, it is part of the notation of stereochemistry in SMILES. The purpose of SMILES is to convert a 2D/3D structural representation to a string of letters, numbers, and symbols. The IUPAC developed a similar representation, which is called InChI. If mhchem does not support SMILES (or InChI), then it should not be used to write it (I personally see no reason why mhchem should support these formats). Hence plain text should suffice for this purpose. The problem was probably, that brackets have to be escaped, which is not necessary in MathJax.

    • O\[C@@\](F)(Cl)Br results in O[C@@](F)(Cl)Br, but
    • O[C@@](F)(Cl)Br results in OC@@(Cl)Br (which is not correct, because the renderer interprets [...](...) as a link)
    • another workaround (to keep the symbols equal) would be using the text wrapper; $\text{O[C@@](F)(Cl)Br}$ results in $\text{O[C@@](F)(Cl)Br}$
  4. Yes, that is different from 1. It is an attempt to reproduce the structure of a molecule.

    • $\ce{>C=O}$ is therefore a rather ill-looking short form of $\ce{R'R''C=O}$ which stands for an unspecified ketone
    • the second example is one of my own misuses because of the lack of more sophisticated typesetting on this site. I state “with $\ce{>Fe}$ being the nano particle”. While I have to admit is looks quite ugly, for the time being, it suffices.

While we are at it, some things I noticed. While <=>, <-, ->, etc. are supported very well, there is one missing

  • $\ce{A <--> B}$ where $\rightleftarrows$ should be used (maybe stretched), but it renders as $\ce{A <--> B}$

Unfortunately the left-right-arrows (Unicode decimal 8644) cannot be streched with the extpfeil extension:

    A \Xleftrightarrows{\Delta T = 298.15~\mathrm{K}} B$$

$$\require{extpfeil} \Newextarrow{\Xleftrightarrows}{1,1}{8644} A \Xleftrightarrows{\Delta T = 298.15~\mathrm{K}} B$$

Similar to the above it is currently not possible to typeset what is referred to as an stoichiometric equation in the goldbook

  • $\ce{A = B}$ should render as $\ce{A} = \ce{B}$, but it actually does render as a double bond $\ce{A = B}$, which is equal to the version without spaces $\ce{A=B}$

The spacing breaks when mhchem is used in an alignment, most notably for <=>

   \begin{align}\ce{A &<=> B}\\ \ce{A &-> B}\\ A &= B\end{align}

which results in the odd looking \begin{align}\ce{A &<=> B}\\ \ce{A &-> B}\\ A &= B\end{align}

  • 4
    $\begingroup$ The @ notation for “fullerenes with metals inside” was proposed in J. Phys. Chem. 1991, 95 (20), 7564–7568 and became generally accepted. The corresponding IUPAC proposal, which was presented in Pure Appl. Chem. 1997, 69 (7), 1411–1434, does not use the @ symbol, but it was not as successful as the first one. $\endgroup$
    – user7951
    Mar 14, 2016 at 11:13
  • $\begingroup$ 1. So it does mean "some property of $\ce{CH3+}$ is less than some property of $\ce{CH3CH2+}$. No bit deal to support that. A few people used it here. See my update for 4) 2. It is not at all very common here. Just 2 instances within a set of 43.5k usages of \ce. $\endgroup$
    – mhchem
    Mar 14, 2016 at 23:09
  • $\begingroup$ @mhchem Please see my update. There have indeed been few questions about endohedral fullerenes, but in other places it is quite common. $\endgroup$ Mar 15, 2016 at 5:17
  • $\begingroup$ @Martin-マーチン Thanks a lot! (Yes, I will say it from time to time, even if it isn't encouraged by SE). Regarding 4, what was the intended output? Regarding your improvement suggestions, this is already implemented for the upcoming update. (But other MathJax developers might have to add their part, for instance stretchable arrows.) Any suggestion are welcome. Here or by mail. (BTW what happens to this answer once the rendering will be correct?) $\endgroup$
    – mhchem
    Mar 15, 2016 at 5:44
  • 2
    $\begingroup$ @mhchem You are welcome. This is meta, it is all about drama, so thanks are very much appreciated. This being said, thanks for developing MathJax and mhchem! It is an interesting question about what happens to this thread once the update is complete. Accepting the answer was the first step. I'd suggest, you'll add a note in which versions the changes have been implemented, so that any one who comes across this knows that this is for an old version. [...] $\endgroup$ Mar 15, 2016 at 5:56
  • $\begingroup$ @mhchem [cont.] Regarding 4, I'll make some changes to the two posts. I don't know the original intention of how it should be rendered in the first post, so I don't know. I'll convert it to an accepted representation of ketones and methylenes. The second post was one of my earlier ones, where I was still learning how to grind the wheels, so there are more issues with it, which I now want to correct. For me the > was just a way to show that it is not a single iron atom, but a nano-particle. MathJax is coming along fine, so there are a few nice things one can do now. $\endgroup$ Mar 15, 2016 at 6:00
  • $\begingroup$ @Martin-マーチン So with all instances of 4. gone, I'll not support the >X and X< "bond". $\endgroup$
    – mhchem
    Mar 16, 2016 at 1:26
  • $\begingroup$ @mhchem Yes, I have corrected that and no support is necessary. I came across another issue I'd like to run by you: meta.chemistry.stackexchange.com/q/2803/4945. Maybe you have already seen this, but I thought just in case you might want to know about that. $\endgroup$ Mar 18, 2016 at 7:45
  • $\begingroup$ @Martin-マーチン Yes, all of that is already incorporated into the upcoming version. See also meta.chemistry.stackexchange.com/questions/2803/… $\endgroup$
    – mhchem
    Mar 18, 2016 at 17:11

I have seen the 4.1 notation rather frequently in rather old (c. 1950 and earlier) articles of Chemische Berichte and probably other journals. It seemed to be a rather simple way to draw molecules with the methods available back in the day. While it should no longer be considered necessary it still seems like a ‘nice hack’. If the hack breaks, it breaks, so be it.

In any case, it was intended to be a ‘bond’ without a single typographic partner on the other side. But actually as two bonds, one going upwards, one downwards.

Martin covered the other cases very well.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .