[Jd1828]

So let’s say that you have this reaction that you just ran and you check it on tlc.  It looks like a mess (several spots) but you need to separate everything out to see if your product in even in there at all.  What is your typical method for running a column to clean it up?  Would you run a flash column or a long gravity column?

Just wondering what others do.  For me I will almost always go for the flash column but that’s only because I can't stand to watch a column drip for 10 hours.  I've always heard flash is better anyway. 

[macman104]

Flash should give a better separation, as the more time it takes, the more the compound can diffuse.  I always run a flash column (if we're defining flash not as those fancy automated systems, but just air pressure applied).  In order of preference, I load neat, then wet, and if I have to, I'll load with the compound dried into silica gel.

EDIT:  Also, if it's really messy, I'd see if I could get some partial solubility of different spots in different solvents.  I recently ran a PCC oxidation, and it was very messy.  However, upon a minimal dilution in DCM, and then adding a larger amount of ether, a solid crashed out of solution, and I was left with a much cleaner column to run.

[Jd1828]

I think diffusion may be my problem I have.. I ran a gravity column a few days ago and the most non polar spot to come off the column in 6 fractions.  It seems like the rest of the sample maybe have diffused to the point where it can't be seen on TLC.

Do you have any specific reference that you like to follow for running a flash column?  I normally try to follow Still's paper from 1978.

[reflux]

I run alot of "messy" reactions with 6-10 spots.  If I need to characterize everything then I usually just run a prep plate (giant TLC) on about 100 mg or less and scrape off the desired components.  It helps if they are UV-active but if not you can just cut a small vertical strip from your prep plate and stain with the appropiate stain.

Otherwise, you can run more than one column.  Say you get three nonpolar spots in one fraction and three polar spots in another fraction on the first attempt.  You can collect them keeping the spots with the similar Rf together then run another column for each type.  It sounds like alot of work but I think three 20 minute columns are better than one 3 hour column.  Also if you take crude nmrs after the first purification it might tell which group of spots the stuff you want is in.

[macman104]

Do you have any specific reference that you like to follow for running a flash column?  I normally try to follow Still's paper from 1978.

Nope, just got a run through from a postdoc when I started working in the lab, and that's how I've been doing it since.

[Mitch]

I would filter through celite and mass spec your product to see if the intended target is there.

[Jd1828]

I would filter through celite and mass spec your product to see if the intended target is there.

That would be my first choice too.  The bad part is we don't have an LCMS in lab and the mass spec characterization lab takes at least a week run a sample.  The PI doesn't really like people sending out reaction mixtures for mass spec either.  Not sure why