[zuriel]

Hey everyone,

I'm hoping someone can give me advice on the best way to recrystallise a crude product I have.
I know the basis of the technique and I understand that you need to dissolve it in the minimal amount of hot solvent which the compound ius not so soluble in at RT. However I have such a small sample of product (150 mg appox.) and I'm wondering if there is any technique I can use that won't require 'solubility tests' which can be time consuming?

One of the postdocs in my group suggested that I use the technique he often uses - dissolve the compound in a minimum amount of DCM...add hexane until the mixture goes cloudy and then keep in the fridge overnight to allow the crystals to nucleate.

I'm not sure if this is the best method though.

So - if any of you have any GENERAL recryst. advice - please feel free to share it. 

[Dan]

One of the postdocs in my group suggested that I use the technique he often uses - dissolve the compound in a minimum amount of DCM...add hexane until the mixture goes cloudy and then keep in the fridge overnight to allow the crystals to nucleate.

I actually do it the other way around. I suspend the crude in refluxing hexane, then add the polar solvent dropwise until it dissolves, then cool (in fridge if necessary).

This thread may be useful:

http://www.chemicalforums.com/index.php?topic=58244

[opsomath]

That's a small sample! You might be best off just running a column right off.

When you don't know what to use, use absolute or 95% ethanol. When that doesn't work, dissolve it in a small portion of DCM, add an equal volume or so of methanol (stop adding MeOH before it goes cloudy) put it in a suitably sized flask, stick a cotton ball or wad of paper towel in the neck, and put it someplace quiet for 24-48 hours. It's amazing how often this gets crystals.

This is assuming that your stuff is nearly insoluble in MeOH and fully soluble in DCM, which is true of an amazingly wide range of organic compounds. I like MeOH better than hexane for the poor solvent, it seems to cause a steeper solubility gradient vs. solvent composition and reduces the chance of oils.

If you want to do a traditional boil'n'cool recrystallization and your stuff isn't soluble in any common alcohols or it oils, try a chloroform/methanol or dichloroethane/ethanol blend (or even chloroform/ethanol if you are feeling lazy) as this mixture will dissolve damn near anything in the right proportions. There will usually be losses due to the enhanced solubility. Be sure to do this in the hood so you're not inhaling chlorinated solvent vapors. A good way to do it is to dissolve your stuff in chloroform, add methanol until it hazes, heat it up til it boils and goes clear, then let it cool slowly. (Substitute your halogen solvent/alcohol as needed)

[zuriel]

Thanks very much for the advice guys, much appreciated. 
I'm going to try some stuff in the lab next week.

[rucik5]

Maybe someone will come back to that topic, so I thought I'll add a little something:
I had some success with using Diethyl Ether as solvent, managed to do a triple recrystallisation of 20 mg of a sample (was left with only 8 mg at the end though). What I do is I simply dissolve the sample in several drops of hot Diethyl Ether (even 2-3 drops + warming up is enough sometimes) and then about 3-5 times more (by volume) of petroleum ether. Allow it to cool down slowly, put in in the freezer (2h minimum, or overnight) and pipette out the solvent, wash the solid with several drops of petroleum ether... and repeat. That works if you need to grow nice crystals for x-ray (or good NMR).

Another method is vapour diffusion. Dissolve your sample in a volatile solvent it dissolves really well in, in a little vial/tube. Put the vial with the sample/solvent in a large beaker with anti-solvent, cover both and leave. As the solvent from the sample slowly evaporates it is slowly replaced by the anti-solvent vapours and usually crystallizes your compound. Again, I had success using Diethyl Ether or DCM as solvents, petrol ether as anti solvent.

This picture explains it:
http://xray.chm.jhu.edu/images/vapor_diffusion.png

[408]

Need more compound detail before we can attempt to predict a solvent system.

Also, make more material.