[Babcock_Hall]

Our starting material is a formylpyridine with a hydroxyl group at the 5-position* (edit: the aldehyde is at the two position*), and we wished to protect the oxygen with TBDMS.  We found an identical reaction in the literature that used TBDMSCl and triethylamine.  We have not yet fully characterized the product, but it was less soluble in CDCl₃ than expected.  There was what appeared to be a tertiary-butyl group by H-1 NMR, but nothing that looked like CH₃Si.  Our options are to set the reaction up again, to try a new base (probably imidazole, which is often used), or to try entirely different conditions, such as TMSOTf and lutidine (we have used this combination previously against aliphatic alcohol groups).  Does anyone have any thoughts?
*my original post was in error, and I am sorry about any confusion that this created.

[rolnor]

Is the formyl in the 3-position? In that case you have a strong hydrogen bond between the formyl and the OH so its much more difficult to protect, you need a stronger base, maybe NaH/DMF  TBDMSiTriflate can also work even with TEA. Or maybe heat, reflux/Toluene

[Babcock_Hall]

My apologies!  I corrected my post, and then I saw your response.  The present one has the aldehyde at 2 (it is a picolinaldehyde) and the hydroxyl group at 5 (the two functional groups have a para-relationship).  However, the second aldehyde in this series may well have the problem you describe, in that the two functional groups have an ortho-relationship.  We have a literature reference (Ozumerzifon Inorganic Chem. 2019.  https://doi.org/10.1021/acs.inorgchem.9b00340), and they used catalytic DMAP, which we also used.  They had an 87% yield over two steps, the first step being an oxidation of the primary alcohol to the aldehyde.
EDT
If we had the silica-based purification to do over again, we might have added 0.5% TEA to the mobile phase.

[Babcock_Hall]

We had followed the reaction using TLC.  The NMR data we had taken right after the purification were ambiguous in that the peak that I tentatively assigned for the tertiary-butyl group had an integral of 9, but there was not an obvious candidate for the two CH₃Si groups.  I took NMR spectra of the product and starting material in DMSO (the previous spectrum was taken in CDCl₃.  The two look identical at first examination.  Taking the TLC data into account, I would hypothesize that we lost the TBDMS group during workup or during chromatography, and I favor the latter idea.

I see two alternatives.  One is to repeat the work, but to use 0.5% TEA in the mobile phase used for the silica gel column.  The other is to use tertiary-butyldiphenylsilyl as the protecting group and probably also use TEA in the mobile phase.  We will run the reaction tomorrow.  If that does not work, we will have to use a different protecting group strategy.

[Babcock_Hall]

Update: We tried the reaction with TBDPSCl yesterday, and we used TEA-treated TLC plates.  I will let everyone know how the purification goes.

[rolnor]

OK, yes, t-BDSi is not very stable, even less so on phenols like your pyridine. TEA is a good idea to keep pH on the alkaline side. DMPA is a good I have used it when acylating 2-OH acetophenons, these are also hard to acylate because of the hydrogenbond. t-BDPSi can be bulky and induce steric hindering, but if its para its no problem.