Talk:2,1,3-Benzothiadiazole

synthesis

Although well reported, the synthesis for this compound seems weird. As depicted, it precedes with reduction of the S centre, which is fairly unusual for thionyl. Presumably there's an intermediate sulfoxide? (although I've never seen anything have N₂S-OH group). I can't find a paper discussing the mechanism but I'd be interested it see it. --Project Osprey (talk) 23:21, 29 August 2020 (UTC)

Sulfoxides are R₂S⁺–O^– (not neutral S with an OH) and the O is nucleophilic enough to react with a range of acylating agents including SOCl₂ (see Pummerer rearrangement) and in the Swern oxidation and related protocols. The next step in all these cases is an elimination reaction in which the acylated oxygen is the leaving group. An alternate synthesis uses disulfur dichloride instead of thionyl chloride (see doi:10.1021/jo01284a040 for proposed analogous intermediate). A large amount of thionyl chloride is used (413 mmol, vs 92 mmol o-phenylenediamine according to [1]), so there is easily a second equivalent available to dehydrate the R–HN–S(O)–NH–R (I think this is called an aminosulfinamide?) to R–N=S⁺–NH–R, which can then become the 1,2,5-thiadiazole (or a resonance structure of it) by loss of that remaining NH hydrogen. DMacks (talk) 05:13, 30 August 2020 (UTC)

@DMacks and Project Osprey: Yes, Dmacks is correct. I had to think carefully about the mechanism, too, when adding the reference. It is interesting to bear in mind that at the time the ring system was normally drawn as a benzene ring fused to a ring drawn with -N=S=N- bonds. That makes it a bit easier to envisage the mechanism. [2] has a good recent review which I'm going to add to the article Michael D. Turnbull (talk) 12:15, 30 August 2020 (UTC)

Yes, that fits. You'd need a very good leaving group as the dehydration step also sees the benzene unit adopt the quiniod configuration. I think the sulfur diimide is a minor resonance form. --Project Osprey (talk) 22:16, 30 August 2020 (UTC)

One might think that the reaction would proceed via a compound called 1,3-Dihydro-2lambda4,1,3-benzothiadiazole 2-oxide (Pubchem "here".) but I suspect that such a beast doesn't in fact exist as I can find only that one reference and it or its tautomers are not in Chemspider. Time to get into a lab and try adding the thionyl chloride very slowly at low temperature ;-) P.S. @DMacks: might be worth updating the diagram to show the proper stoichiometry and the byproduct SO₂, as I've done in the text. Michael D. Turnbull (talk) 10:28, 31 August 2020 (UTC)

I'll update the diagram tonight. I see a few known chemicals containing "sulfamoyl chloride" (N–SO₂–Cl) but none containing just one oxygen. DMacks (talk) 13:54, 31 August 2020 (UTC)

I swapped in the balanced-equation image. DMacks (talk) 16:32, 6 September 2020 (UTC)

My first guess had been a sulfinylamine continuing via 'C1=CC=CC2=C1N=[S](N2)O' (whatever that's called). I honestly don't have any idea how likely that is.--Project Osprey (talk) 18:57, 31 August 2020 (UTC)

Sulfur diimides would have the sulfur sp-hybridized and the two π bonds at right angles to each other, which is consistent with its behavior more similar to a cumulene than any sort of all-conjugated 3-atom system. And that does not seem compatible with the orbital or electron-count requirements for aromatic nature of this chemical. DMacks (talk) 13:42, 31 August 2020 (UTC)

It is briefly mentioned in one of the refs doi:10.1002/ejoc.201201161 (pg 237), although the mechanism of tautomerization doesn't make immediate sense to me (again) - valence tautomerisation perhaps? I think it's sufficiently minor that we don't have to worry about it too much. --Project Osprey (talk) 18:57, 31 August 2020 (UTC)

@Project Osprey: My library doesn't subscribe to that journal. Could you send me the page? DMacks (talk) 16:33, 5 September 2020 (UTC)

@Project Osprey and DMacks: I have found an Eli Lilly patent [3] that claims both sorts of substructure (e.g. examples 78-81 and 154-157). Normally, I would think of Eli Lilly chemists as being reliable but in this case I think it's fantasy stuff: they have no detailed experimental for any of these! Michael D. Turnbull (talk) 17:38, 5 September 2020 (UTC)

There's an open access version here but as I said, I don't think it's very important--Project Osprey (talk) 21:26, 7 September 2020 (UTC)

That's actually quite helpful. For the sulfur dimine form of the thiodiazole ring, I assume you are referring to the "delocalized form" in Scheme 4 of that paper? Those N–S bonds are delocalized (one bond is dotted line), not a diimide (two full lines of a double bond). Instead of two pi bonds, there is only "some pi bonding"; as a start, consider it like one sulfur lone-pair spread across the whole N–S–N region (3-center–2-electron in the pi world). By resonance, it's like N3 of imidazole, that has pi bonding to both adjacent C but has a "frozen" lone-pair on it rather than a full C=N=C bonding pattern. DMacks (talk) 02:31, 8 September 2020 (UTC)

No, Scheme 27 but I can't get into the references that discuss it. --Project Osprey (talk) 12:24, 8 September 2020 (UTC)

I couldn't get your link to work (it seems to have an expiry date) but it was available from academia.edu which I found by Googling the doi. The article is by Neto et al, who seems to be writing a lot on this heterocycle recently: I added that review to our references. All he is saying in Scheme 27 is that when co-ordinated to metals (Ru in this case), the electronics is better explained by the resonance form with -N=S=N- rather than =N-S-N=. Michael D. Turnbull (talk) 14:19, 8 September 2020 (UTC)

That's all I was saying though; that the sulfur dimine appeared to be a valid resonance form. It's probably very minor - but also one of those awkward ones that's hard to explain with simple curly arrows. I'm not saying that it's important, I just find it interesting. --Project Osprey (talk) 23:46, 8 September 2020 (UTC)