Sciencemadness Discussion Board

Failed attempts at oxidation of alcohols with DMSO/P2O5

Lionel Spanner - 8-3-2024 at 13:31

Hello all.

On my journeys on the internet, I came across a relatively obscure 1987 paper by Taber, Amedia and Jung claiming that alcohols could be oxidised to aldehydes, and aldehydes to ketones, by means of DMSO and phosphorus pentoxide in DCM, then releasing the free aldehyde into solution with a trialkylamine base; the whole procedure being carried out at fridge to room temperatures. Said paper is attached.

This paper has been cited by others and the method has apparently been reproduced by them, however I have made no less than 8 attempts to reproduce it myself, testing its effectiveness by converting cinnamyl alcohol (smells like marzipan and spring flowers) to cinnamaldehyde (smells like cinnamon) and have completely failed to do so, as the reaction has resulted in little to no product. I have been using DCM dried over phosphorus pentoxide, and have done experiments both with and without an argon atmosphere, though the latter doesn't seem to have much effect.

There is a definite cinnamon smell in the residues, but no measurable yield of product.

Later attempts incorporated some modifications from a 2002 paper by Carpenter et al. (also attached - the experimental procedure is the very last one in the text), namely premixing the alcohol in DMSO before adding it to suspension of P2O5 in dry DCM, and pouring the final mixture into water rather than adding dilute acid to it. The result is the same: little to no actual product.

Has anyone else on here managed to get this procedure to work, and if so, how?

Many thanks!

Attachment: alcohol oxidation by P2O5 and DMSO.pdf (278kB)
This file has been downloaded 213 times

Attachment: example of Onodera reagent oxidation.pdf (271kB)
This file has been downloaded 166 times

clearly_not_atara - 8-3-2024 at 14:04

Have you tried any other substrates? It's plausible that cinnamyl alcohol just acts weird here. Not least of all because it can in principle cyclize to indene.

For allyl alcohols I always thought activated MnO2 was preferred.

Lionel Spanner - 8-3-2024 at 14:27

Quote: Originally posted by clearly_not_atara  
Have you tried any other substrates? It's plausible that cinnamyl alcohol just acts weird here. Not least of all because it can in principle cyclize to indene.

For allyl alcohols I always thought activated MnO2 was preferred.

I did previously try salicyl alcohol, with the same result.

The idea in both cases was to produce a product that smelt substantially different to the starting material, as proof that reaction actually worked - in both cases, it did produce the intended product, but with next to no actual yield.

As my salicyl alcohol was homebrewed and of questionable purity, and I was unfamiliar with the reaction, I decided to test the reaction out on a commercially available alcohol of known composition and high purity instead.

bnull - 8-3-2024 at 15:30

Maybe the P2O5 is the problem. Do you have other dehydrating agents? Acetic anhydride, for example, could work.

I leafed through March's Advanced Organic Chemistry and found some other DMSO-based oxidations that may interest you. Maybe they can give you any clue about the issue. The pages in question are attached. Good luck.

Attachment: DMSO.pdf (90kB)
This file has been downloaded 168 times

Lionel Spanner - 8-3-2024 at 20:19

Quote: Originally posted by bnull  
Maybe the P2O5 is the problem. Do you have other dehydrating agents? Acetic anhydride, for example, could work.

I leafed through March's Advanced Organic Chemistry and found some other DMSO-based oxidations that may interest you. Maybe they can give you any clue about the issue. The pages in question are attached. Good luck.


The phosphorus pentoxide is from an official Carl Roth distributor (who also sells to individuals, get in!), and is a free-flowing powder - inside the container, at least, before it touches atmospheric moisture. It behaves as expected with water, and conc. sulphuric acid does an excellent job of removing it from the walls of the flask after drying a solvent. I have no doubts about its authenticity.

It's funny that you mentioned acetic anyhdride - my next line of investigation will be the Albright-Bergman procedure, which uses DMSO and acetic anhydride, and is a better established method.

I will, however, try this at a later date. I've done nothing in the last two chemistry-working weeks but failed attempts at alcohol oxidation, and exploring other reactions in the interim will likely prevent burnout when I do return to the subject.

kmno4 - 9-3-2024 at 03:00

I do not know why you use so sophisticated procedure for such oxidations, maybe you feel like it, but it is beyond my interest.
Cinnamyl alcohol can be converted to aldehyde with any persulfate.
Just water, persulfate (NH4, K, Na), alcohol and heating. The yield is high or very high.

bnull - 9-3-2024 at 04:39

Quote: Originally posted by Lionel Spanner  
an official Carl Roth distributor (who also sells to individuals, get in!)
Just saw the price of 500g of P2O5. That plus shipping (and some stupid tax that the "gummint" here would be glad to include) would break my heart. The price here in my country (including shipping, no restrictions :D) is about 15 euros, give or take a few cents. And were not even self-sufficient in phosphates. But thanks anyway. I'll save it for the exotic, hard to find, once-in-a-lifetime kind of reagent they probably have.
Quote: Originally posted by Lionel Spanner  
The phosphorus pentoxide [...] I have no doubts about its authenticity.
What about the trialkylamine? Did you use Et3N as in a Swern oxidation, or was it DIPEA as in the second paper? Maybe cinnamyl alcohol is picky about amines.

Lionel Spanner - 9-3-2024 at 06:19

Quote: Originally posted by kmno4  
I do not know why you use so sophisticated procedure for such oxidations, maybe you feel like it, but it is beyond my interest.
Cinnamyl alcohol can be converted to aldehyde with any persulfate.
Just water, persulfate (NH4, K, Na), alcohol and heating. The yield is high or very high.

The idea was to carry out a Swern-type activated DMSO oxidation, which is very versatile as a general procedure - it can handle a wide variety of substrates, and can oxidise multiple groups on the same molecule without polymerisation or other undesirable side-effects that occur in other procedures.

I don't have the capability to do Swern itself (no easy way to get dry ice, no way to get or prepare oxalyl chloride, and no desire whatsoever to generate carbon monoxide), and many of the similar reactions require exotic reagents or reaction conditions. This seemed like one of the more amateur-friendly procedures.

I have no specific desire for cinnamaldehyde, I just decided to use it for the purpose of testing whether the reaction itself has worked.
Quote: Originally posted by bnull  
Quote: Originally posted by Lionel Spanner  
The phosphorus pentoxide [...] I have no doubts about its authenticity.
What about the trialkylamine? Did you use Et3N as in a Swern oxidation, or was it DIPEA as in the second paper? Maybe cinnamyl alcohol is picky about amines.

I used triethylamine. What I got back was largely the starting material, suggesting either the alcohol/DMSO/P2O5 intermediate had failed to form at the first stage, or the product, once formed, had been reduced to the starting material.

[Edited on 9-3-2024 by Lionel Spanner]

clearly_not_atara - 9-3-2024 at 07:15

The Swern conditions are known to attack phenols as well:
https://cir.nii.ac.jp/crid/1390282679113907712

In general "activated DMSO" i.e. Me2SX reagents act as strong electrophiles and are not compatible with molecules having multiple nucleophilic groups.

How did you even make salicyl alcohol? Isn't it usually made by reduction of the aldehyde? To me a normal test substrate would be something like menthol.

Lionel Spanner - 9-3-2024 at 16:42

Quote: Originally posted by clearly_not_atara  

How did you even make salicyl alcohol? Isn't it usually made by reduction of the aldehyde? To me a normal test substrate would be something like menthol.

By reduction of salicylic acid using the method described by Kanth and Periasamy in J. Org. Chem., 56 (1991), 5964-5965:
https://pubs.acs.org/doi/pdf/10.1021/jo00020a052

Lionel Spanner - 29-5-2024 at 14:28

Hello all. I tried another DMSO-based method for alcohol oxidation (Albright-Goldman), had multiple total failures with that one too, and worked out that most the likely cause was that my DMSO was not dry enough. I had a similar problem with t-butyl alcohol from the same supplier, which turned out to be azeotropic and contained a lot more water than expected, despite being advertised as 99% pure.

There are easier ways to oxidise alcohols, so it's most likely I'll not be pursuing either method any further unless I feel really stubborn (masochistic?) and determined to prove that they actually work.

bnull - 29-5-2024 at 16:02

Well, I'm sadistic enough to offer you some more references. They come from the Encyclopedia of Reagents for Organic Synthesis (EROS), under the entry "Dimethyl sulfoxide-Phosphorus Pentoxide". https://onlinelibrary.wiley.com/doi/abs/10.1002/047084289X.r...

Keras - 30-5-2024 at 01:29

I’ll try the original procedure (the microscale version) with benzyl alcohol later this month and tell you if I get any benzaldehyde. Did you try sodium hypochlorite? IIRC, I got a decent yield of benzaldehyde from benzyl alcohol and sodium hypochlorite at room temp.

sykronizer - 4-8-2024 at 23:02

I seem to recall DMSO mediated benzyl alcohol oxidation to benzaldehyde at room temperature, just by using a large excess of DMSO and bubbling oxygen or air into the stirred mixture. It took a fair amount of time, but if I recall the purity was quite high, yield around 80 percent