Can someone please advise / point me to information regarding the difference between these two compounds:


enter image description here


enter image description here

The reason for this question is that rdkit assigns them the same 'inchikey', as if they were identical, or spontaneously interconverting substances (like tautomers).
But they aren't tautomers, as there is no atom being transferred, only electrons; and they are distinct commercial substances, sold in fairly large quantities, so I would guess they are individually sufficiently stable.

Many variously decorated commercial compounds containing either of these moieties are also found, reinforcing the above conclusion.

So it would seem that the machinery that calculates the inchikey is wrong in treating these two molecules as identical.

Any ideas?

EDIT summary of insights from the comments

According to both the people who commented so far, these two species are resonance structures of the same molecule.
The fact that different cas numbers, names, catalogue numbers from a single vendor are found for the two, would therefore seem to be incorrect, misleading and unjustified.
On the contrary, the fact that rdkit returns the same inchikey for the two seems to be correct.

  • 3
    $\begingroup$ Just learn to draw mesomeric structures, instead of using some "combi-blocks" - the second one has one bond too many. $\endgroup$
    – Mithoron
    Commented Jul 24, 2023 at 17:56
  • 1
    $\begingroup$ @Mithoron - surprising though it may sound to you, I do know how to draw chemical structures. I used the pictures from the vendor to show how they represented these molecules. And FYI, Scifinder (CAS) also draws these two molecules in the exact same way as Combi-Blocks does. If you have more information to provide, please do so. It is not really clear to me how making such snide comments or insinuations about people's intelligence or skills helps answering the question I am asking. $\endgroup$ Commented Jul 25, 2023 at 6:23
  • 2
    $\begingroup$ Sorry for sounding harsh, but it boils down just to the knowledge about resonance (not intelligence, as you mentioned), Some structure drawing tools are just tools, users need to discern if results are good or not (they frequently aren't). In this case a slightly alleviating factor could be people buying into deprecated hypervalency - this second mesomeric structure is the "hypervalent" one. Also how some vendors draw it and sell, isn't exactly relevant, apparently can lead rather to confusion than to learning. $\endgroup$
    – Mithoron
    Commented Jul 25, 2023 at 12:11
  • 1
    $\begingroup$ Could you link to some of those distinct syntheses from papers? And also, just because there are two resonance structures, doesn't mean that both are equally prevalent. I like the structure without the hypervalent sulfur better. $\endgroup$
    – Curt F.
    Commented Jul 25, 2023 at 12:40
  • 2
    $\begingroup$ For the record: "Homework questions must demonstrate some effort to understand the underlying concepts." was the reason people provided for closing the question. Wrong. The very first thing that I did was research these two substances, draw myself the resonance structures to see how they were conceptually the same (one electron pair from the S atom of the first one can be moved to the ring, all other pairs shift around until the last one reconnects to S, forming the N=S=N moiety). I should have stopped at that and ignored the other misleading evidence pointing to them being different. $\endgroup$ Commented Jul 25, 2023 at 19:16

1 Answer 1


In practice, there is no difference between the two structures - they represent the same compound if you were to purchase them. That said, in terms of "correctness", the non-hypervalent structure is preferred, even though its Lewis structure still does not accurately represent the full electronic structure of the molecule. You could add some non-hypervalent charge-separated resonance forms which improve the description, as shown below.

The overarching truth is that occasionally you can find compounds with unusual Lewis structures available for purchase because they weren't sanity-checked. This is especially likely with combinatorial chemistry suppliers as they will often automatically populate their catalogues with thousands of structures. For example, sometimes you can find a cyclohexa-2,4-diene-1-one on offering, though on further inspection a chemist would realize these are actually a phenol tautomer, which typically only exist in part-per-trillion amounts.

But with benzothiadiazole specifically, the problem goes a bit deeper. While a trivial and satisfying Lewis structure can be assigned to it (or more precisely, benzo[c][1,2,5]thiadiazole), what about the related structure benzobis(thiadiazole) (more precisely, benzo[1,2-c:4,5-c']bis([1,2,5]thiadiazole)), which contains another fused benzothiadiazole ring opposite to the first? This is a well-known compound, of which hundreds of variants have been explored in organic electronics.

It turns out in this case that there is no "trivial" Lewis structure which can be assigned to the compound, at least not without a hypervalent sulfur atom. And curiously, that's what the literature has settled on - a quick Google image search of benzobisthiadiazole will turn up dozens of articles using the hypervalent structure shown below. It has become the standard in the field, and nobody particularly cares. It is possible to draw a non-hypervalent charge-separated resonance form, but it is ignored. The true electronic structure is even more complex, apparently including non-negligible biradicaloid character, which can be shown on a non-hypervalent Lewis structure.

As a last point, x-ray crystallographic data of benzobisthiadiazole compounds show that the two thiadiazole rings are typically completely equivalent (there is no single hypervalent ring, as the widely-drawn structures suggest), indicating that it really all boils down to Lewis structures just being limited tools at representing reality. You always have a trade-off between simplicity of representation and accuracy, and sometimes this trade-off is particularly unsatisfying.

  • $\begingroup$ Excellent, thank you very much for this detailed explanation! "occasionally you can find compounds with unusual Lewis structures available for purchase because they weren't sanity-checked" - that's quite bad indeed, but while one may forgive vendors for not being so strict with their checks, it remains shocking (to me) to find distinct cas numbers for species that do not even exist individually. I created a post about that. The phenol vs cyclohexadienone example is not strictly applicable: tautomers are indeed distinct species with, as you point out, measurable proportions in mixtures. $\endgroup$ Commented Jul 26, 2023 at 6:36

Not the answer you're looking for? Browse other questions tagged or ask your own question.