[Babcock_Hall]

Mix the aloquot with some tbdms chloride and imidazole and see if its is replaced by something much more nonpolar?

That is a creative idea, but I do not have those reagents on hand.
EDT
Would an aminal hydrolyze during the workup with KHSO4?

[Babcock_Hall]

One reagent which came up on page 1 of this thread was LDBBA (An JA Bull Korean Chem Soc 2015 36 2928. DOI: 10.1002/bkcs.10571).  The reaction is performed at 0 °C.  Millipore Sigma offers this reducing agent as a 0.25 M solution in THF/hexanes.  Although pricey, one might argue that it is cheap if it works.

[Babcock_Hall]

This is an update on my second attempt at the reduction of a Weinreb amide.  This reaction was let run for about 24 hours at 0-5 °C, then for about the same amount of time at room temperature.  My reason for doing so was that I wanted to see the starting material disappear entirely, but that never happened.  Either there was some unreacted starting material, or there was a side-product that had a similar R_f value.  The last several TLC plates, which were run over these two days, looked similar.  There was a less intense spot with a higher R_f than the starting material, a more-intense spot with about the same R_f as starting material, and a faint spot with a lower R_f than starting material.  I worked the reaction up with KHSO₄ and purified over silica (50:1 w/w) using 60:40 hexanes/ethyl acetate.  I was able to isolate two products, but I did not see a product corresponding to the slowest-moving spot by TLC.  One fraction was a mixed product by TLC, and I did not combine this with either pool.

I have analyzed the ¹H NMR of the two chromatographic pools in a preliminary way  Pool 1 has a main product that has a characteristic aldehyde proton.  However, there appears to be a second product in which there are two singlets that might correspond to the fragment -N(CH₃OCH₃.  The chemical shift of one of these is close to the predicted shift for the Weinreb amine that would be produced upon over-reduction.  Based on the integration, here is 70-85% presumed aldehyde, and 15-30% presumed Weinreb amine.  Pool 2 has two singlets at the same chemical shift of the two methyl groups as the starting material.

[wildfyr]

Use my bisulfite trick to seperate the aldehyde from the other stuff. You can use sodium bicarb to convert it back I think. I would actually say that it might be possible to do the workup without a column using this technique.

[Babcock_Hall]

By mass I recovered about 72% of what I applied to the column.  Pool 1 was about 16% of the recovered mass, and pool 2 was about 84% of the recovered mass.  In other words, I am getting very little aldehyde, because pool 1 is in the minority and is not pure.

BTW, here is a reference to non-aqueous regeneration after bisulfite isolation:  Kjell DP et al., J. Org. Chem. 1999, 64, 5722-5724, in case anyone needs it.  It provides a little bit of background on the general subject.

ETA
One person with specific experience in this area has told me not to attempt a purification of the aldehyde on silica, because the silica will cause decomposition.

[pgk]

Indeed, silica or alumina may catalyze the autoxidation of aldehydes to the corresponding carboxylic acids in the column and TLC plates. But this does not happen always. It mainly depends on the aldehyde structure and the working conditions. Consequently, the chromatographic purification of aldehydes figures in the literature from time to time.
Alternatively and in order to be sure, you can try the bisulfite adduct (as wildfyr suggests) and regenerate your aldehyde by alkaline treatment or/and in non-aqueous conditions:
Purification process for 3-phenoxybenzaldehyde, US Patent 4162269, (1979)
2-Methyl-3-phenylpropanal, Organic Syntheses, 61, 82, (1983) or in: Organic Syntheses, Collective volume 7, 361, (1990)
http://www.orgsyn.org/demo.aspx?prep=CV7P0361
A Novel, Nonaqueous Method for Regeneration of Aldehydes from Bisulfite Adducts, J. Org. Chem., 64(15), 5722-5724, (1999)
https://pubs.acs.org/doi/abs/10.1021/jo990543v

[Babcock_Hall]

Thank you very much.  The first thing that I have to do is to produce a high percentage of the aldehyde, which I do not seem to be doing.  So far I have tried DIBAL/THF twice, both following a literature protocol for a derivative of an amino acid that differed only in the protecting group (they used FMOC and I used BOC).  The first time I used 1.0 equivalents of DIBAL (I had intended to use 1.5) and kept the temperature below zero (this was close to the procedure that I was trying to follow).  The second time I used 1.5 equivalents and gradually increased the temperature over a couple of days (see above).  I followed the reaction by spotting the reaction mixture onto a TLC plate, but what appeared to be starting material was always the major spot.  It could be something else with a very similar R_f and NMR spectrum.

Right now I do not know what to try next.  I could try DIBAL/THF again, and I am open to suggestions on how to change the conditions.  DIBAL in toluene has precedent, but I do not have this reagent on hand.  Lithium aluminum hydride has been recommended to me and has precedent in the literature.  There is also a paper cited upthread that uses LDBBA.  I would like to spare the BOC-protected material because it is the most valuable.  I also have the FMOC-protected Weinreb amide of the same compound, and that is somewhat less valuable.  I can also make a simple model Weinreb amide.

It is also possible that spotting onto a TLC plate is producing results that are misleading in some way.

[pgk]

The problem is that if your aldehyde is destroyed in the column and the TLC plate, you will obtain low yields, whatever the reduction method is. Thus, it would be better to initially focus to a non-chromatographic monitoring/purification method of the final product.

[Babcock_Hall]

The problem is that if your aldehyde is destroyed in the column and the TLC plate, you will obtain low yields, whatever the reduction method is. Thus, it would be better to initially focus to a non-chromatographic monitoring/purification method of the final product.

Are aldehydes inert to alumina?  I am not sure how I would monitor by TLC if alumina also decomposes aldehydes.  The only other thing that comes to mind is for each time point to do a mini-conversion into the sulfite-adduct and then run the adduct on a TLC plate.
EDT
The protocol I am following used a flash column and hexanes/EtOAc.  However, I realize that I may have to depart from the script...

[pgk]

Alumina may further catalyze the autoxidation of aldehydes to the corresponding esters (Tishchenko Reaction).
So at the beginning, you can run the reduction overnight without monitoring, as described above and purify the final product via the bisulfite adduct, in order to get an idea about the “real" reaction yield.
In a next trial, monitoring can be effectuated by HPLC, where degassed eluents are used and columns are packed under pressure and theoretically, there is no air (oxygen) trapped therein.

[wildfyr]

Have you taken NMR of the crude and just seen the conversion of weinreb amide to aldehyde? This will tell you if the reaction is actually low conversion, or just falling apart on silica.

You can also run a 2D TLC to see if your product falls apart on silica.

I've never understood why people want to column everything, I avoid it as much as possible. Expensive, requiring high expertise, and most of all, time consuming. Clever washes and precipitations are by far the best (and most scalable) methods. You'd be surprised what you can accomplish, such as removing PPh₃O with a zinc chloride wash! https://pubs.acs.org/doi/pdf/10.1021/acs.joc.7b00459

[Babcock_Hall]

For the second reaction I checked the crude product (after washes but before the column) with H-1 NMR, and it was suggestive of about 15% aldehyde, which is not that different from what I guessed, based on TLC.

[Babcock_Hall]

For the second reaction I checked the crude product (after washes but before the column) with H-1 NMR, and it was suggestive of about 15% aldehyde, which is not that different from what I guessed, based on TLC.

This experiment rules out decomposition on silica as being the sole problem.  Although decomposition is a possibility, it was not mentioned in the paper I am following.

It has been suggested to me that the presence of the BOC or FMOC group could cause the consumption of one equivalent of DIBAL through removal of the carbamate proton.  If this happened, the work-up would reprotonate it.  I am in the process of looking over other examples that used DIBAL.  Thoughts?

[Babcock_Hall]

I tried the reaction on a smaller scale using the FMOC-protected Weinreb amide and 2.5 equivalents of DIBAL.  I do not have the NMR in front of me at the moment, but my recollection is that by integrating the aldehyde peak, I saw about 20% conversion.  Some literature searching that I did suggested to me that LAH will work.  I thought that I cancelled my order of lithium diisobutyl-tert-butoxyaluminum hydride in THF/hexanes, but it showed up last week.  Under the circumstances I am inclined to try LDBBA next and LAH after that.

[rolnor]

The Fmoc is base-labile so this is maybe a problem?