"Dry-column" Flash Chromatography, L.M. Harwood, Aldrichimica Acta, 1985, 18, p. 25
Now this paper isn't particularly detailed so we had to mess around with it for quite a long time before we got it right but it was worth it. We were columning really large quantities of stuff this way. I spoke to my former industry colleague about it the other day and he told me that they were doing 500 gram columns now collecting 1 and 2 liter fractions in conical flasks. Anyway, we thought that it made sense to publish a paper with more detailed instructions so that someone might actually try to use it successfully. Fortunately, the referees also liked the idea so we got this paper out on the topic: Dry Column Vacuum Chromatography, D.S. Pedersen and C. Rosenbohm, Synthesis, 2001, pp. 2431-2434
We even took the liberty of renaming the technique because we thought flash was a bit misleading. Basically Dry Column Vacuum Chromatography (DCVC) is an alternative to flash where the solvent is sucked through the column. The solvent is added one fractions at the time (making very precise gradient elution dead easy) and hence the column is sucked almost completely dry between fractions. For the full details read the paper. So what does it look like? Above there's a picture of a 20 gram column I ran the other day. Now this is a big column but it also works with small columns. My favourite size columns have a diameter of 4-6 cm. They are easy to work with and I would typically collect 20 ml fractions. If you want to give DCVC a go but don't want to ask the glass blower for a fancy piece of glassware at first you don't have to. Some of the bits and pieces you need for a simple set up are shown above. A nice and sturdy spatula is essential for good column packing. Quite a few chemist I know use water aspirators for vacuum rather than a fancy diaphragm pump and it seems to work for them so that's also an option. The trick to make this technique work is to get the right silica.
If you use the stuff to the right it will all work fine. However, if you use regular flash silica it will not work. Don't even try it's a waste of time. Regular flash silica is simply to coarse and you can't pack a proper column.
Cool features about DCVC are:
(1) Large scale columns are possible
(2) It's faster than flash. Mainly because you don't collect many fractions (usually 20-30)
(3) You don't use much silica (column is only ~5 cm high) or solvents
(4) Because the column gets sucked dry between fractions you can easily do other stuff whilst running it. Even leaving it for a few hours will not ruin the resolution.
My personal record is seven DCVC's in one day (didn't go home until 10 pm though). However, I wouldn't recommend doing more than four in a day. You start going insane after the fourth column. The technique works so well that I haven't run a flash column in five years and I have no intention of ever doing one again. The columns can normally be recycled. Just give them a good rinse. I usually do something like: MeOH, EtOAc and finally hexane. Seal it up nicely and stick in the fridge. They easily last for a week in the fridge if you seal them up properly. This is how I do it: An old suba seal on the stem a bit of para film on the top with a bit of tape and stick in the fridge in a beaker (see picture).
One important detail that we have discovered since publishing the paper is that one should NOT concentrate the sample to be columned on silica. Instead, you should use Celite (or Kenite). Opposed to silica Celite is very non-polar and hence you can use quite a lot for loading you sample without compromising the resolution. This way you can ensure that you get as much of your compound as possible on to the actual column. If you stick a piece of filter paper on top of the silica column prior to loading the celite it makes it really easy for you to scrape the celite off after running the column so that you can use it once more. Also you should stick a piece of filter paper on top of the kenite so that you don't mess the surface up when you pour the solvent on. Okay I'd better stop here. You obviously need to read the paper before you try this method. Just send an email to firstname.lastname@example.org to request a copy. D!