[Babcock_Hall]

I would like to remove a para-methoxybenzyl (PMB) group from a carboxylate ester.  The compound of interest also has a pyridinium group, an amide, a second (free) carboxylic acid, and a BOC-protected amine.  I hope to remove the PMB and the BOC groups together to make the final product in a four-step synthesis.  The final product should be somewhat water-soluble, and we intend to purify by ion-exchange on Dowex-1.

I have been reading through the literature on PMB esters, and I noticed that several groups use acids such as TFA or formic acid, followed by removal of volatiles.  Some methods use phenol as a solvent: Torii et al., J. Org Chem. 56, 3633 (1991), or have anisole present: Stewart, Australian J. Chem. 21, 2543 (1968).  Some methods do not have a scavenger: Wakamiya et al., Bull. Soc. Chem. Japan 68(9), 2699-2703 (1995) and Stelakatos and Argyropoulos, J. Chem. Soc. C, 964 (1970).  If one uses TFA or formic by itself, the reaction would seem to be a transesterification that is driven by mass action.

1.  With respect to the methods that lack either anisole or phenol, I must be missing something.  If one simply removes the volatiles, I cannot see why the equilibrium would not get displaced back toward reactants.  In other words, the excess acid (TFA or formic) would be removed but not the PMB formate or PMB trifluoroacetate ester.  I could accept that if one separated the PMB trifluoroacetate or PMB formate ester first, then one could potentially remove volatiles and isolate the desired product.
2.  Should we be looking at different removal strategies altogether?

[discodermolide]

I would remove the Boc and the PMB separately.
Hydrogenate off the PMB then use TFA for the Boc.

[Doc Oc]

Hydrogenate off the PMB then use TFA for the Boc.

I know hydrogenation is used to remove benzyl groups, but for some reason I always thought PMB was resistant to standard hydrogenolysis.  I almost always see DDQ used.  Of course, I've not used PMB so I don't have any expertise, just curious.

[discodermolide]

Hydrogenate off the PMB then use TFA for the Boc.

I know hydrogenation is used to remove benzyl groups, but for some reason I always thought PMB was resistant to standard hydrogenolysis.  I almost always see DDQ used.  Of course, I've not used PMB so I don't have any expertise, just curious.

I have a procedure for the hydrogenolysis of PMB groups (ethers).
DDQ works as well but it is messy and the quality of the DDQ is important, as are the conditions, which are sometimes quite difficult to work out!

[Babcock_Hall]

We have made both the benzylester and the 4-methoxybenzylester (PMB ester) versions of the compound of interest (see attached tif file).  We have stayed away from hydrogenation because of the pyridinium ring.  Given the pyridinium ring in NAD as a model, I would estimate that the reduction potential at pH 7 (E°') is -0.32 volts.  That having been said, if we reduced our ring by two electrons, I would not throw the compound away, but it would be different from the intended target.

[discodermolide]

We have made both the benzylester and the 4-methoxybenzylester (PMB ester) versions of the compound of interest (see attached tif file).  We have stayed away from hydrogenation because of the pyridinium ring.  Given the pyridinium ring in NAD as a model, I would estimate that the reduction potential at pH 7 (E°') is -0.32 volts.  That having been said, if we reduced our ring by two electrons, I would not throw the compound away, but it would be different from the intended target.

Why don't you make the pyridinium salt as the last step?

[Babcock_Hall]

Right now the alkylation of the nitrogen is the second-to-last step.  If we used a free carboxylic acid with a terminal iodo group as the alkylating agent, the last step would be simply the removal of the BOC group at the other end of the molecule.  That is close to what you are suggesting, if I understand it correctly.  I suppose I was afraid that the BOC group would fly away at the elevated temperature (ca. 80 °C) of the alkylation, in the presence of the weakly acidic -COOH group.  However, it just occurred to me that we already have one -COOH group present in the molecule, so my reasoning is beginning to look suspect.

[discodermolide]

Right now the alkylation of the nitrogen is the second-to-last step.  If we used a free carboxylic acid with a terminal iodo group as the alkylating agent, the last step would be simply the removal of the BOC group at the other end of the molecule.  That is close to what you are suggesting, if I understand it correctly.  I suppose I was afraid that the BOC group would fly away at the elevated temperature (ca. 80 °C) of the alkylation, in the presence of the weakly acidic -COOH group.  However, it just occurred to me that we already have one -COOH group present in the molecule, so my reasoning is beginning to look suspect.

Maybe TMSI would do it all in one step, deprotect the PMB, remove the Boc and form the Pyridinium iodide. I don't know if TMSI will cleave PMB esters, but it does for others.
Or you could use a different ester, one which relies on elimination, such as the TMS ethyl ester?

[orgopete]

…the last step would be simply the removal of the BOC group at the other end of the molecule. … I suppose I was afraid that the BOC group would fly away at the elevated temperature (ca. 80 °C) of the alkylation, in the presence of the weakly acidic -COOH group. 

Wouldn't loss of the BOC be a bonus?

[Babcock_Hall]

Removal of the BOC group might be a plus on the condition that the alpha-nitrogen (the one at the far end of the molecule) did not get alkylated.  Obviously, we only want to alkylate the pyridine nitrogen.

[Babcock_Hall]

With respect to TMSI, there is a paper that uses TMSCl and NaI in acetonitrile to transesterify various carboxylate esters.  IIRC, they were able to remove a benzylester at 45 °C.  If anyone needs the reference, please let me know.  I also found a reference in Philip Kocienski's book on protection groups to succinic acid protected with PMB at one end and a tetrahydronaphthyl group at the other.  At room temperature with one equivalent of TMSCl and NaI, only the latter group was removed, but not the PMB group (p. 413; Tet. Lett. 1999 40 5691-04).

This strategy is one that we are considering, but I was unable to find a literature reference that deals with what would happen to the BOC group under these conditions.

[discodermolide]

With respect to TMSI, there is a paper that uses TMSCl and NaI in acetonitrile to transesterify various carboxylate esters.  IIRC, they were able to remove a benzylester at 45 °C.  If anyone needs the reference, please let me know.  I also found a reference in Philip Kocienski's book on protection groups to succinic acid protected with PMB at one end and a tetrahydronaphthyl group at the other.  At room temperature with one equivalent of TMSCl and NaI, only the latter group was removed, but not the PMB group (p. 413; Tet. Lett. 1999 40 5691-04).

This strategy is one that we are considering, but I was unable to find a literature reference that deals with what would happen to the BOC group under these conditions.

I would guess the Boc would fall off.

[Babcock_Hall]

I would anticipate that the BOC group would fall off, also.  However, I am not sure whether the tert-butyl cation would make isobutene gas or be trapped by iodide.  Therefore, I am not sure of how many equivalents of iodide and TMS-Cl to use.

[discodermolide]

I would anticipate that the BOC group would fall off, also.  However, I am not sure whether the tert-butyl cation would make isobutene gas or be trapped by iodide.  Therefore, I am not sure of how many equivalents of iodide and TMS-Cl to use.

You mean TMSI!
Make sure it's relatively new.
Use 6-10 equivalents. I don't think the isobutene would be a problem. Remember you will get everything silylated, so you will need some sort or protolytic work up, butanol or such with a trace of acetic acid to re-generate the free acids etc.

[Babcock_Hall]

I had been thinking in terms of TMS-C/NaI.  Do you mean true TMSI and not this alternate recipe?