[rolnor]

You are very dedicated, good. Still it would be nice to se a chromatogram, preferably a LC-MS. You dont learn so much from this NMR-study, its mostly confusing.
Yes, why do you see the iodide? Thats key, it should be consumed. If you used the sulfide as nuchleophile and then oxidised to sulphone you would get fast and more clear result I think. You could even do this one-pot.

[Babcock_Hall]

The TLC results and the P-31 NMR results were consistent with each other regarding both leaving groups.  With P-31 one might be able to learn what the side-product is when I used toluenesulfonate as the leaving group.  I am consulting with someone with much greater expertise in P-31 NMR than I have, although MS could also identify it.

The sulfonyl fluoride reaction followed by a Horner reaction can be done in one pot, but I am not sure that I want to go that route.  Do you mean sulfide, disulfide, or thiol?

[rolnor]

If you alkylate a sulfide-Na salt  with the tosylate or iodide you get the sulfide, then oxidize this one-pot with Oxone to the desired sulphone. What you need is mercapto ethanol sodium salt I think?
Its probably so, that you dont get the product you want now, just as your NMR shows, but you dont know why, if you run LC-MS you can maybe see what goes wrong.

[Babcock_Hall]

We performed the reaction between 3-mercapto-1-propanol and diethyl phosphonomethyltoluenesulfonate at least twice, once with potassium carbonate as the base and IIRC once with cesium carbonate as the base.  Although the yield has been variable in our hands, it is a satisfactory reaction overall.  The oxidation to the sulfone went smoothly.  It would make sense to use 2-mercaptoethanol in a similar reaction; it is inexpensive and not that volatile.  However, the corresponding sulfonyl phosphonate is not presently one of our targets.  The high cost of many of the sodium sulfinates makes this route (EDT, meaning sulfinate as the nucleophile) less attractive to me than it was before, but I already have sodium cyclopropanesulfinate in hand.

[rolnor]

Excellent! I move my fingers like Mr Burns in "Simpsons".

[Babcock_Hall]

A year ago we alkylated benzyl mercaptan and then oxidized the product sulfide.  This spring we alkylated 2-thiopyridine and then oxidized the product.  I suspect that these reactions are slower owing to steric problems associated with the quaternary phosphorus atom, but they are otherwise OK.

[Babcock_Hall]

One of my previous model reactions used TsOCH₂P(O)(Out)₂.  It produced a mixture with the desired product PhSO₂CH₂P(O)(OEt)₂ as the most abundant form.  I hope to see a greater percentage of the desired product by switching from a harder to a softer electrophile.  I found a reaction in the patent literature (WO 2021/263278) between sodium cyclopropanylsulfinate and BrCH₂P(O)(OiPr)₂ in DMF with tetrabutylammonium iodide as a catalyst.  The mole ratio of the sodium sulfinate to phosphonate to TBAI was 1.5:1.0:0.1.  They worked at 100 °C and ran overnight.  I did not see a yield.

I decided to try a model reaction with sodium phenylsulfinate and ICH₂P(O)(Out)₂ in DMF but without TBAI.  One difference between this reaction and previous reactions is that I used fresher DMF.  After about 16 hours at 77 °C the reaction had proceeded to about 25% completion, as judged by P-31 NMR.  I raised the bath temperature to 109 °C and I will take another time point.  I am not sure of of how much rate enhancement to expect by adding a catalyst.  The catalyst may be helping to make the sodium phenylsulfinate more soluble, but in warm DMF, I did not think that it would make a big difference.  I thought that the reaction in the patent literature primarily used TBAI to obtain a catalytic halogen exchange.  I plan to set up a second model reaction which includes TBAI, but I am unsure about the best choice of temperature. 

The cyclopropane derivative is one of my targets, but the 2-propane derivative is also on my list, and I have both of the appropriate sodium sulfinates in hand.  I just need to optimize my conditions.

[rolnor]

TBAI is not logic to use when you have a alkyliodide, I dont see the point. But a iodide should be more soft than a tosylate?.How about my suggestion, use a sulfide and get 100% S-alkylation, then oxidise this to sulphone with one of many possible oxidizers. You can make the sulfide if you pick a proton from the corresponding methylphosphonate with LDA and the add sulphur, then add the alkylhalide to the so formed lithiosulphide in one pot. Foolproof.

[rolnor]

Sorry I did not see the post with thiopyridine. You should be aware that pyridine can be oxidized to the N-oxide with some oxidizers. Thiophenolate is a good nucleophile so it should work fine if you want to go that way. But thiophenol stinks, you can use bleach to remove the smell.

[Babcock_Hall]

I tried this route again for two reasons.  One is that I used fresher DMF and the other is that I had additional support from the literature in the form of a similar reaction.  I imagined that the TBAI also functioned as a phase transfer catalyst, but at 80-100 °C, maybe not.

I can imagine using HSCH₂P(O)(OEt)₂ and the tosylate derivative of cyclopropyl alcohol as a viable route (we made a small portion of this thiol recently).  The opposite disconnection is not a viable route.

[Babcock_Hall]

Increasing the temperature to about 107 °C did not bring about completion of the reaction with diethyl iodomethylphosphonate.  I am thinking of picking up some diisopropyl bromomethylphosphonate, which is the exact compound used in the patent.

[Babcock_Hall]

Sorry I did not see the post with thiopyridine. You should be aware that pyridine can be oxidized to the N-oxide with some oxidizers. Thiophenolate is a good nucleophile so it should work fine if you want to go that way. But thiophenol stinks, you can use bleach to remove the smell.

We made the sulfonyl phosphonate starting with 2-thiopyridine twice, in a two-step synthesis.  Both times we had to run a column after the oxidation, something that we do not have to do when we oxidized the corresponding S-methyl sulfide to its sulfone.  We are not sure about the identities of the side-products yet, but we may know more next week.

I decided to buy diisopropyl bromomethylphosphonate, and when it comes in, I will try a small test reaction along the lines of the protocol in the patent before committing more expensive starting materials.

[rolnor]

The bromo would be much less reactive then the iodo, but thiolates will probably displace bromo quick anyway.

[Babcock_Hall]

In this case the nucleophile is a sodium alkylsulfinate, and there will be tetrabutylammonium iodide as the catalyst.  No yield was given in this patent.  "A mixture of diisopropyl (bromomethyl)phosphonate (0.5 g, 1.93 mmol), sodium cyclopropanesulfinate (0.37 g, 2.895 mmol) and TBAI (0.071 g, 0.19 mmol) in DMF (4 mL) was heated overnight in a IOO0C oil bath. The mixture was cooled to it and diluted with EtOAc and water. The layers were separated, and the organic layer was washed with saturated NaHC03, brine, dried (Na2S04) and filtered. The solvent was removed and the crude was purified by silica gel to obtain the title compound."

[rolnor]

Yes, the sulfinate is much weaker nucleophile than sulfide so it could be useful with TBAI. Cesium iodide would be easier to remove after the reaction.