[Babcock_Hall]

Last fall a worker made two compounds, 4-methoxybenzylesters of 4-iodobutanoic acid and 5-iodopentanoic acid.  The NMRs looked as expected.  We tried using one of them in an alkylation a week or two ago, and got unexpected results the first time:  we did not see most of the NMR peaks associated with the alcohol group of the ester.  We repurified the 5-carbon ester over silica, and most of the brown color was removed (I don't recall having seen a brown color in previous syntheses of these two esters).  However, we again got unexpected results from a second alkylation, namely the lack of peaks that we could attribute to hydrogen atoms of the ester.  Then we checked the presumed repurified ester by H-1 NMR.  I don't see anything which could be the benzyl group, Ar-CH₂-O-.  The CH₃O- group is no longer a sharp singlet, but is broad and a multiplet.  The aromatic protons which are expected to be at roughly 7.3 are not present, and the aromatic protons between 6.9 and 7.0 are no longer a sharp almost-doublet, but instead are broad.

We had saved behind a small sample of the ester prior to its repurification on silica, and the NMR looks similar to what we see after repurification (which did remove much of the color).  We took the NMR of the 4-carbon ester made at about the same time, and it was similar to the 5-carbon ester:  no evidence that the PMB group stayed intact during storage.  We have made these compounds before on several occasions.  If we had decomposition prior to this, we were entirely unaware of it.  Any thoughts on what to do and/or what happened?  Thank you for any ideas.

[Archer]

It's not unheard of for alkyl iodides to disproportionate into I₂ and an alkyl radical.

That may explain the brown colour. Also you may have many products from the radical interactions which are broadly similar in polarity.

Have you tried GC-MS?

[discodermolide]

As Archer said iodine can be eliminated. Iodine is an oxidant for benzylic carbons, I have used it for that purpose. Can't find the reference atm. PMB is easily oxidised.

[Babcock_Hall]

The iodo group appears to be intact on the basis of the presence of a signal at the expected chemical shift, by H-1 NMR.  Also, the compound that we were trying to alkylate, BOC-ornithine-isonicotinate, seems to have alkylated on the pyridine nitrogen, on the basis of a triplet at about 4.65 ppm, which we assign to a R-CH₂-N(pyridinium) on the basis of model compounds.  In storage the ester would only have whatever impurities it picked up in silica gel chromatography.  In other words, I don't think it was stored in the presence of an oxidant.

From a practical point of view, we would like to complete the synthesis by deprotecting the molecule, which means removing the BOC group and the alcohol which makes up the ester, assuming that there actually still is some alkyl group protecting the carboxylic acid.  Without knowing the identity of this group, I really have no idea whether or not this is a fool's errand.

EDT

Here are the C-13 data in CDCl₃:  178.9, 32.8, 32.6, 25.5, 5.9 ppm.  There are some very small peaks in the aromatic region.  The H-1 data are:  3.21 triplet, 2.41 triplet, 1.92 multiplet, 1.89 triplet.  Based on the NMR data, it looks as if the whole 4-methoxybenzyl group is gone, and we now have the free acid.  I don't understand how we (a) were able to move the free acid through silica (25% to 75% EtOAc in hexanes) or to see the free acid on TLC plates using only UV-active plates.  These results don't make sense.

[discodermolide]

Have you made the lactones? For example:

and

[Babcock_Hall]

Thanks; I would not have thought about lactone formation.  We don't have a peak around 69-70 ppm in the C-13 spectrum, and we don't have a peak around 4.2 in the H-1 spectrum.  However, we also do not seem to have a proton for a free carboxylic acid.  We don't have an easy way to acquire mass spectral data, but we are working on it.  Would the iodo group or a putative carboxylic acid cause problems for mass spectrometry?

EDT
Another possibility is that we formed a carboxylic acid anhydride somehow, although the C-13 shift is too downfield for the carbonyl carbon.

[discodermolide]

Mass spec. depends upon the conditions, you may not see a M⁺ because the iodo group can leave readily and the acid may well loose CO₂.
The anhydride is a possibility but I find the lactones more appealing. In the case of 5-iodopentanoic acid imagine if the iodo left forming a terminal C=C and then it catalyses iodolactonisation you would get

but this has a triplet at about 6.2, so it may be a silly idea?

[Babcock_Hall]

At this point there are no silly ideas.  However, we don't see a peak at 6.2 ppm.  Right now I am planning to finish the synthesis and purification.

[Babcock_Hall]

We found a flask with a small amount of this ester that was made about two years ago.  The flask had been sitting out on the bench for months.  When we ran the H-1 NMR of this sample, it looked just fine.  I suppose it must mean that the other two syntheses somehow had an impurity that led to decomposition, perhaps by some catalytic process.

[Archer]

Did you ever run GCMS in the end?

[Babcock_Hall]

Not yet.  We don't have walk-up or service GC/MS.  But thanks for the reminder; I will have to figure out how to get the student trained.  We might also do high resolution mass spec on the final product, which we have made previously.