[Rory]

Mesylation of the following primary alcohol yielded the desired mesylate as well as the recovered starting alcohol (ca. 50%). The starting material couldn't be consumed even with raised reaction temperature and prolonged reaction time (at r.t. for 2 days, 50% conversion). Another portion of MsCl (8 eq) and TEA (16 eq) was then added, only to consume just part of the remaining alcohol. Then a third portion of MsCl (8 eq) and TEA (16 eq) was added and most of the alcohol disappeared, but about 10% of the alcohol didn't react.

I don't know what's wrong with such a seemingly simple reaction. Anyone has any idea?

[Dan]

The problem is probably water. Either the MsCl got wet and contains lots of mesic acid (its shelf life is not fantastic), or your triethylamine is not anyhdrous, or you DCM is not anhydrous. NB if you are using DCM from a bottle, make sure it is not stabilised with 05-1% ethanol/methanol, use the amylene stabilised variety instead. Ideally you wan't alcohol free DCM from a still. It really depends on the age and storage care of the reagents/solvent. I find that old triethylamine is usually the culprit, and usually distil it over CaH₂ before use.

You should run this reaction under a dry inert atmosphere as well, otherwise water vapour will condense inside the flask.

[Rory]

Thank you for your suggestion, dan. TEA is newly distilled. DCM was distilled about one year ago and stored in a 500ml schlenck flask. MsCl is from a new bottle with a label that shows the production date 2011-09-22. Is MsCl itself so readily to hydrolyze? I mixed the MsCl with water and two phases seperated without any white mist (HCl). So I think MsCl is quite stable with water in the absence of a base. And addition of NaHCO3 to MsCl didn't afford CO2. Maybe MsCl is not the culprit. I did run this reaction under a dry inert atmosphere. I still don't know what's wrong.

[Dan]

Hmmm.... Not sure then. Wet starting material?

I've never actually tried to examine the stability of MsCl to water, but really just assumed it was quite sensitive. I have done a lot of tosylation, mesylation and triflation - and it is only MsCl that ever causes problems for me. I do find that sometimes reactions stall or are very slow, and always assumed this was to do with water in the system. I usually add the MsCl at 0^oC and then warm to RT.

I have one last idea (long shot) - is the product you isolate from this reaction, after the long reaction time and large excesses, defnitely the mesylate or it it the alkyl chloride? It occurred to me that perhaps the starting material and the mesylate co-spot on TLC, and the product that takes the time to form is the alkyl chloride by displacement of the mesylate?

If a tosylate would do, I'd just hit it with TsCl - much less fussy reagent.

[Rory]

Hmmm.... Not sure then. Wet starting material?

I've never actually tried to examine the stability of MsCl to water, but really just assumed it was quite sensitive. I have done a lot of tosylation, mesylation and triflation - and it is only MsCl that ever causes problems for me. I do find that sometimes reactions stall or are very slow, and always assumed this was to do with water in the system. I usually add the MsCl at 0^oC and then warm to RT.

I have one last idea (long shot) - is the product you isolate from this reaction, after the long reaction time and large excesses, defnitely the mesylate or it it the alkyl chloride? It occurred to me that perhaps the starting material and the mesylate co-spot on TLC, and the product that takes the time to form is the alkyl chloride by displacement of the mesylate?

If a tosylate would do, I'd just hit it with TsCl - much less fussy reagent.

The starting alcohol and the mesylate co-spot on TLC with PE/EA as the developing solvent, but are easily seperated on the TLC plate using DCM/Et2O as the developing solvent. So no alkyl chloride was observed. Actually I wanna transform the alcohol to a six-membered cyclic nitrone. According to related literature it could be synthesized by reacting the targeting mesylate with NH2OH. Would tosylate or triflate also work?

[Dan]

Actually I wanna transform the alcohol to a six-membered cyclic nitrone. According to related literature it could be synthesized by reacting the targeting mesylate with NH2OH. Would tosylate or triflate also work?

Triflate would definitely work, but it's overkill and Tf₂O is far more expensive than TsCl or MsCl. I would give TsCl a go - though less reactive than a mesylate, I expect a primary tosylate would work.

Secondary mesylates and iodides cyclise to 5 membered rings very well (e.g. Wei-Chieh Cheng Tetrahedron 65 (2009) 93–100), so a primary tosylate should be fine.

 

[Rory]

Actually I wanna transform the alcohol to a six-membered cyclic nitrone. According to related literature it could be synthesized by reacting the targeting mesylate with NH2OH. Would tosylate or triflate also work?

Triflate would definitely work, but it's overkill and Tf₂O is far more expensive than TsCl or MsCl. I would give TsCl a go - though less reactive than a mesylate, I expect a primary tosylate would work.

Secondary mesylates and iodides cyclise to 5 membered rings very well (e.g. Wei-Chieh Cheng Tetrahedron 65 (2009) 93–100), so a primary tosylate should be fine.

 

I'll try the tosylate. Thanks very much!

[Doc Oc]

I've never actually tried to examine the stability of MsCl to water, but really just assumed it was quite sensitive.

This has been my experience as well.

I have one last idea (long shot) - is the product you isolate from this reaction, after the long reaction time and large excesses, defnitely the mesylate or it it the alkyl chloride? It occurred to me that perhaps the starting material and the mesylate co-spot on TLC, and the product that takes the time to form is the alkyl chloride by displacement of the mesylate?

In agreement with this statement, I almost always observed alkyl chloride, both with mesyl and tosyl chlorides.

One other thing for you to consider, I tried to cyclize a similar substrate to this and had tremendous difficulty with it.  Once the mesylation occurs, you could very well be getting alkene from E2 elimination.  Depending on your TLC conditions, it might run at the same Rf as your SM.  Might be worth trying to isolate it and get an NMR.

[Rory]

One other thing for you to consider, I tried to cyclize a similar substrate to this and had tremendous difficulty with it.  Once the mesylation occurs, you could very well be getting alkene from E2 elimination.  Depending on your TLC conditions, it might run at the same Rf as your SM.  Might be worth trying to isolate it and get an NMR.

Could you please show me the structure of the substrate you failed to cyclize and corresponding conditions, Doc Oc? Have you done any mesylation that caused the same trouble for you?

[Doc Oc]

I'm sorry, but I can't, it's IP that I can't share.

Starting with your compound, if you were to take away the ketone, and instead of an oxygen have an acetylamine there, that'd be the crux of what I was doing.  I wanted to cyclize that into a 6-membered ring, but couldn't.

[stewie griffin]

If you ever have trouble with MsCl or TsCl causing formation of the alkyl chloride you can always just try switching to the anhydride (Ms₂O). The resultant anion leaving group is not as good a nucleophile as chloride.

[Dan]

Starting with your compound, if you were to take away the ketone, and instead of an oxygen have an acetylamine there, that'd be the crux of what I was doing.  I wanted to cyclize that into a 6-membered ring, but couldn't.

I think oximes and amides are quite different beasts, but while I know there is precedent for cyclisation to 5-membered nitrones, I'm not sure about their more challenging 6-membered counterparts, so things may well be tricky (haven't looked hard though).

Rory, what is your ultimate target? If you're just after a piperidine there are more reliable methods...

[Doc Oc]

Starting with your compound, if you were to take away the ketone, and instead of an oxygen have an acetylamine there, that'd be the crux of what I was doing.  I wanted to cyclize that into a 6-membered ring, but couldn't.

I think oximes and amides are quite different beasts, but while I know there is precedent for cyclisation to 5-membered nitrones, I'm not sure about their more challenging 6-membered counterparts, so things may well be tricky (haven't looked hard though).

You're right, I certainly wasn't trying to make that comparison.  My main point was that I have worked with a similar primary alcohol and that as soon as I activated it, the base in the reaction led to E2 elimination and left me with an alkene.

[Rory]

Starting with your compound, if you were to take away the ketone, and instead of an oxygen have an acetylamine there, that'd be the crux of what I was doing.  I wanted to cyclize that into a 6-membered ring, but couldn't.

I think oximes and amides are quite different beasts, but while I know there is precedent for cyclisation to 5-membered nitrones, I'm not sure about their more challenging 6-membered counterparts, so things may well be tricky (haven't looked hard though).

Rory, what is your ultimate target? If you're just after a piperidine there are more reliable methods...

I know there're lots of precedents for cyclization to 6-membered nitrones. Things are very different between Doc Oc's case and mine.
Strangely, tosylation of the same alcohol was uncooperative, too. I can always see 3 spots on the TLC plate: the alcohol, TsCl and the tosylate. Prolonging reaction time or addition of another portion of TEA to consuming the SM and TsCl was unsuccessful. Maybe it has something to do with the alcohol itself, but not with the impurity including water.
The nitrone is my ultimate target.

[sudi]

Hi everyone. I am trying to mesylate a conduritol using Mesyl anhydride, TEA in  dry dichloroethane. I get two products every time instead of one. Has anyone had a similar experience?