[Babcock_Hall]

I am trying to duplicate something a student did some years ago, which is to remove two benzyl protecting groups from a diester of a phosphonate.  The phosphonate has an oxygen that has the trichloromethanesulfonyl leaving group, RO₂P(O)CH₂OSO₂CCl₃.  He used Pd/C and ethanol as the solvent, and he used 40 psi for an hour, but he did not indicate the volume of solvent or the amount of catalyst.  I did a little bit of literature searching, and I found a reference that used 50 psi in a deprotection of a derivative of pyrophosphate.

This is my first time running such a reaction.  Any thoughts on how best to proceed?  Thank you.

[discodermolide]

Just try a normal hydrogenation, with 5% 0r 10% Pd/C and see how it goes.

[Babcock_Hall]

Thanks.  I made this compound using a sulfinylating agent (trichloromethanesulfinyl chloride) that is no longer commercially available, so it is quite precious (I have roughly a gram).  I also have a few milligrams of material I obtained from a back extraction (it has more color, but it produced good P-31 and H-1 NMR spectra).  I am contemplating using this material first or splitting the sample in half, but the jars for the Parr apparatus are very large (they look as if they are about 500 mL bottles), and I don't know if having a relatively large solvent volume will be a problem.  I am considering using 50-100 mL of ethanol.

[discodermolide]

They should be OK. Soak up lots of H₂ with that volume and surface.

[Dan]

Personally I would try running it on 20-50 mg in 1-2 mL EtOH under a balloon of hydrogen first if it is that precious. I've never worked with phosphonate esters though - this is just based on my experience with reductive amination, azide reduction and benzyl ether hydrogenolysis with H₂, Pd/C.

[Babcock_Hall]

Dan, if the reaction were successful at the 50 mg scale, how would you scale up?  Would you continue to use a balloon, or would you move to the Parr apparatus?

[Babcock_Hall]

Dan, I saw your suggestion after I set the reaction up.  My student ran for an hour, and I ran my reaction for about 80 minutes.  By P-31 NMR and by H-1 NMR, I would say that I have a mixture that looks perhaps as if it is partially debenzylated.  Any suggestions?

[Dan]

Dan, if the reaction were successful at the 50 mg scale, how would you scale up?  Would you continue to use a balloon, or would you move to the Parr apparatus?

You could do either. I would probably stick with the balloon and scale up to 500 mg. For me that's really just familiarity - I have only used pressure hydrogenation a couple of times (worked perfectly) because I have never really had convenient access to the kit and have always been able to spare the extra reaction time - waiting  day or two for the reaction to finish was no big deal. I also prefer to avoid significantly altering a procedure when scaling up - it's unlikely that an increase in hydrogen concentration would result in more unwanted side reactions, but I'm not sure I'd want to risk it with half my stock of a precious compound. The reactions are also easier to monitor by TLC when they're under a balloon because you don't have to dismantle everything.

Seeing as you've already set up the pressurised reaction, I would not bother looking at the balloon reaction unless something goes horribly wrong (which I doubt).

Dan, I saw your suggestion after I set the reaction up.  My student ran for an hour, and I ran my reaction for about 80 minutes.  By P-31 NMR and by H-1 NMR, I would say that I have a mixture that looks perhaps as if it is partially debenzylated.  Any suggestions?

If it's a clean reaction, just resubmit and be patient

[Doc Oc]

B_H, you have a couple of options.  The first is most obvious, which is to increase the amount of Pd/C in the reaction to speed it up.  I will note that in double debenzylations I've often had trouble with the 2nd debenzylation, which brings me to my 2nd option.  In the past, I had success using Pd(OH)₂ (Pearlman's catalyst) with just a H₂ balloon.  5-10% catalyst with a balloon for an hour usually was plenty to get both benzyls off.

I've used both the balloon and Parr hydrogenator.  I will say that both work, but the balloon is way easier to setup/use.  People also tend not to flip out about the safety of a balloon vs a pressurized hydrogen apparatus.

[Babcock_Hall]

I took the reaction that looked incomplete and dissolved in about 17 mL of ethanol, which is about half of what I used before.  I ran for a total of almost five hours at 40-50 psi.  I added 50 mg of Pd/C catalyst, and after 1.2 hours I added 70 mg more catalyst.  It was suggested to me that this would lessen the problems of sulfur poisoning the catalyst.  My preliminary analysis of the H-1 and P-31 results is that I have (almost) complete debenzylation, but I also have a significant amount of solvolysis, perhaps 20-25%.  My starting compound was (PhCH₂O)₂P(O)CH₂OSO₂CCl₃.  The position of substitution may be hindered, but there is a good leaving group present; therefore, solvolysis is not entirely unexpected.  I still have some of the starting material left, and I can try other conditions.  Are there other solvents that are considered good ones for hydrogenations, perhaps 2-propanol or THF?

DocOc, I just reread your note now, and I will look into this as well.

[discodermolide]

Try ethyl acetate or some other acetate, that should avoid solvolysis.

[Babcock_Hall]

Try ethyl acetate or some other acetate, that should avoid solvolysis.

discodermolide, Initially I was a little worried about the solubility of the product, a free phosphonic acid, in EtOAc.  Therefore, I used 25 mg of methylphosphonic acid as a model compound and 3 mL of EtOAc, and it dissolved completely after a few minutes.  My plan now is to scale up.  Does that sound reasonable?

[discodermolide]

Yes, that looks reasonable.
Good luck.

[Babcock_Hall]

I tried a hydrogenation at 50 psi for about six hours in EtOAc, and from the H-1 NMR I would say that the reaction looks incomplete.  The mass is also too large.  I can think of at least two explanations for the failure of the debenzylation to go to completion.  One, I used a different lot of Pd/C, although it was the same manufacturer.  Two, Perich and colleagues (Australian J. Chemistry 44(2) 1991, pp. 233-252) used formic acid as their solvent for debenzylation of phosphate esters.  In one part of their paper they talk about removal of a BOC group at the same time (which the formic acid would also do).  But in another part of their paper (on the removal of phenyl esters of phosphate by hydrogenolysis, p. 241), they mention that the acid strength of the protic solvent is a factor in the rate of benzylester reduction (though whether they are referring to carboxlyic or phosphoric esters is not clear to me).  They used 2% acetic acid in ethyl acetate to remove the 4-nitrobenzyl group from a carboxylic ester via Pd/C catalyzed hydrogenolysis.  I am not sure whether or not I need a protic solvent or co-solvent at this point.
http://www.publish.csiro.au/paper/CH9910233

EDT
I have been looking over some debenzylation conditions, and it seems as if some but not all use protic solvents.
http://www.scribd.com/doc/22791404/51/Hydrogenolysis-of-Benzyl-Esters

[Babcock_Hall]

Should I simply repeat the reaction conditions (ethyl acetate as the solvent) but change back to the bottle of Pd/C that I first used, or should I try some change in solvent?