Here’s a bit of a tricky situation. Every now and then I have a product which is clean, crystalline, and highly hygroscopic. I definitely don’t want to make more work for myself, redissolving the beautiful crystals just to pass them through a flash column, but every time I try to do a simple filtration I get a gooey mess.

The best approach in my experience is to run an inverted filtration, drawing the solvent away from the crystals under inert atmosphere. It’s not a terribly common technique, but I’ve been playing around with it over the last little while and I think I’ve got a pretty robust system put together. If anyone has run into this problem in the past though, I’d love to hear about your solution.

The basic design is cribbed from the advanced practical organic chemistry book, with a couple of modifications to fit it in with my glassware. In essence, the longest needle I can find is inserted backwards through a rubber septum, then lowered into the mother liquor. I stuff a little ball of kimwipe into the leur lock to serve as a filter, and generally try to use a 16G needle (anything up to a 22G will work, but the flow rate is really slow). The sharp end of the needle fits through another septum into a three way valve, which is either connected to the vacuum source or vents to the open air. A standard RBF at the bottom of the three way valve collects the filtered solvent.

For air sensitive compounds you can forgo the vacuum line in favour of a nitrogen source. I’m using a two-necked RBF in the example below, so I just attached the line with rubber tubing, but you could also think of double-piercing the septa of your solution flask. The key is to have a direct line from the gas source; putting a bubbler in the middle of the chain prevents solvent flow.