I had a compound that I was preparing on a 10-50 g scale and kept getting one spot with a consistently lower Rf which when present would stop my compound from recrystallizing. I heard about this method from a colleague who gave me a copy of Daniel Sejer Pedersen's 2001 Paper in Synthesis (Synthesis, 2001, (16), 2431-2434). Daniel has further detailed this technique on his blog.
I made the closest adaptation I could with our existing glassware and it worked great. In this photo 13 g of crude compound is loaded onto 40 g of silica. It took 300 mL to see any trace of the compound and after 500 mL I increased the solvent polarity from 70% hexanes : 30% EtOAc to 50% hexanes : 50% EtOAc to recover 12 g of compound that recrystallized easily. I have since managed to upscale to 25 g with exactly the same setup.
I used DCM to load my compound to the silica and care had to be taken when removing the DCM which bumps violently on the rotavap even when the vacuum is lowered with care, so you might want to use a splashguard with cotton in it. The dry loaded silica is then distributed evenly over the wide column of clean silica using a glass funnel. The Synthesis paper suggests placing a filter paper on top of the loaded silica when pouring solvent. I tried once with this and once without and found that in my case I got a more evenly spread pour with no paper but by using a wide mouthed beaker moved in a circular fashion as poured. It is true what it says in the paper that when a compound comes through a foaming is seen in the solvent. It would definitely be better to have a separatory funnel on the bottom with a vacuum inlet near the neck as this would more conveniently control the release of each fraction.
I made the closest adaptation I could with our existing glassware and it worked great. In this photo 13 g of crude compound is loaded onto 40 g of silica. It took 300 mL to see any trace of the compound and after 500 mL I increased the solvent polarity from 70% hexanes : 30% EtOAc to 50% hexanes : 50% EtOAc to recover 12 g of compound that recrystallized easily. I have since managed to upscale to 25 g with exactly the same setup.
I used DCM to load my compound to the silica and care had to be taken when removing the DCM which bumps violently on the rotavap even when the vacuum is lowered with care, so you might want to use a splashguard with cotton in it. The dry loaded silica is then distributed evenly over the wide column of clean silica using a glass funnel. The Synthesis paper suggests placing a filter paper on top of the loaded silica when pouring solvent. I tried once with this and once without and found that in my case I got a more evenly spread pour with no paper but by using a wide mouthed beaker moved in a circular fashion as poured. It is true what it says in the paper that when a compound comes through a foaming is seen in the solvent. It would definitely be better to have a separatory funnel on the bottom with a vacuum inlet near the neck as this would more conveniently control the release of each fraction.