[Babcock_Hall]

Today one of my students tried to run a silica gel column, but it stopped after about 100 mL of solvent had flown through.  I was able to get the column restarted briefly, but it stopped again.  The solvent was 1:8:91 acetic acid/methanol/acetone.  The product was the amide formed from 4-nitrophenyl isonicotinate and BOC-Lysine (having an unprotected carboxylic acid).  Therefore the side product is 4-nitrophenol.  We ran a similar column with a slightly higher percentage of methanol last week without any apparent problems.  The only odd thing I noticed this time was a slight haziness at the bottom of the glass column below the stopcock, and a whitish precipitate in a beaker used to catch solvent.  My student said that she saw some cloudiness in the previous column, but I do not recall seeing something like this before.  I wonder whether or not the solvent was somehow contaminated or not made up properly.  I know that a high percentage of methanol can dissolve silica, but I would not have thought that 8% would be a problem.

Any thoughts about what went wrong?  Any thoughts about how to recover the product?  Thanks for any ideas.

[discodermolide]

I reckon you have dissolved some of the silica and blocked the frit.
I think you will have to empty the column and stir the silica in methanol or ethanol to recover your compound.
Why did you not use reverse phase for the purification?

[Babcock_Hall]

I reckon you have dissolved some of the silica and blocked the frit.
I think you will have to empty the column and stir the silica in methanol or ethanol to recover your compound.
Why did you not use reverse phase for the purification?

We have used silica perhaps half a dozen times before without this happening (most recently last week, where the percentage of methanol was higher).  We based our conditions on something similar in the literature, a paper from Karl Folkers's lab.  We don't have a great deal of experience in using reverse phase-based purifications (although we were contemplating it for the next step in the synthesis).  We do have some RP TLC plates a a few Sep Pak cartridges, however.  Can you suggest a good book chapter or other resource on reverse phase, by any chance?

[discodermolide]

I can only suggest you look at the suppliers web site. I know the Merck page has info about TLC, maybe worth a look there.
Acetic acid /methanol/acetone is quite an unusual solvent system for normal silica.
Maybe your compound crystallised and blocked the column.
You will just have to unpack the column and get the compound back.

[Babcock_Hall]

The original authors (Janecka et al., J. Med. Chem., 1994, 37, 2238-41) used a slightly less aggressive mobile phase than we did, but we did not get much movement of the compound under their conditions IIRC.  It was 94:5:1 CHCl₃/MeOH/HOAc.  I just reread this section, and they also used:  n-BuOH:py:AcOH:H₂O in the ratio 30:10:3:12, and they saw a higher R_f in this system than in the chloroform-containing system.

Between you and me, one can never rule out that an undergraduate student made a mistake in how the solvent was prepared, but it is difficult to be certain that this was the problem (they might have used the wrong acid, for example).  I was thinking of using roughly 25% methanol in a couple of portions to pull it away from the silica once we unpack the column.  However there will be dissolved silica at this point.  If we remove the solvent, I would predict that some silica will come out with it.  Maybe we can ignore this silica. although it will add to our mass.

[curiouscat]

Silly question, perhaps, but no way to analyse the solvent to see if there was indeed a prep. error by the undergrads?

[Babcock_Hall]

At one point I contemplated whether the student had used another acid, not acetic acid.  But unfortunately, at 1% it is hard for me to detect acetic acid by smell on top of acetone and methanol.  I suppose that we could try GC, but we might just be chasing our tails.

[orgopete]

I cannot say categorically it wasn't the silica, but I've had columns become blocked and it wasn't the silica. In my case, it was simply crystallization. Your circumstance doesn't seem as though this should have occurred, but if the blockage were similar to mine, then I'd search for a change in solubility. If one of the components had dissolved silica and binding of the component caused a supersaturated silica solution, okay, it's the silica. If the components increased their mutual solubility, then separation can result in a supersaturation condition. Simply dissolving silica in the eluent should result in silica in your fractions. I am surprised if you saturated the solvent and do not find silica in the fractions.

On second thought, was this vacuum chromatography? Evaporation of the solvent could cool the eluent and cause precipitation of silica or one of your components.

[Babcock_Hall]

We typically use gravity, and we use 40-63 micron.