March 28, 2024, 05:13:10 AM
Forum Rules: Read This Before Posting


Topic: Alternative method of recrystallization at room temperature (using 2 solvents)  (Read 3843 times)

0 Members and 1 Guest are viewing this topic.

Offline blaisem

  • Regular Member
  • ***
  • Posts: 87
  • Mole Snacks: +5/-0
The standard method for recrystallization taught is to heat a solvent containing an insoluble sample until it dissolves, then allow it to cool so that crystals from the pure compound can precipitate while small amounts of impurity remain solvated.

Another method I've become aware of is to use a multisolvent system at room temperature, but I don't know the exact details of the technique, and searches seem to land me exclusively by the standard method. Wikipedia mentions it as a multi-solvent recrystallization but doesn't go much into the details.

I'm assuming this method is more desirable when your compound is unstable under high heat. Does anyone know the specifics of the technique, eg:

  • How much of Solvent A to use vs. solvent B
  • Must both solvents be miscible with each other?
  • Techniques to add the secondary solvent. I surmise dripping it dropwise down the side of the glass is preferable, but is this best done in a round bottom flask, a tube, or something else?
  • What is the mechanism for recrystallization exactly? I wanted to do it in a tube, but I thought if the crystals nucleate at the interface between the solvents, they will simply fall down into the more abundant solvent and redissolve. I'm a little confused on that.
  • Temperature control: is light heating initially or cold at the end ever advisable?
  • How long does the method take? Do you let it sit overnight or wait for observable nucleation then crash out the crystals with cold?

I'm currently looking at a solvent mixture of ether:petroleum ether, where the sample is soluble in ether but not in petroleum ether. However, I am trying to extend the method to another compound I have, but since I am unfamiliar with the technique I'm not entirely sure what properties are ideal for a solvent mixture in terms of volatility and miscibility for optimization. This second sample has impurities soluble in hexane, so I thought maybe something like ethanol and hexane. My plan was to dissolve the sample in ethanol, then pipette hexane dropwise down the sides, let it sit until all the crystals had developed, then pipette out any remaining solvent. I wanted ethanol because I thought the miscibility would be important, and it's not as strong a solvent as DCM so my product yield will be higher. Does this sound like the right thinking?

Sorry for the rambling post, and thanks for any advice!

Offline ATMyller

  • Chemist
  • Full Member
  • *
  • Posts: 223
  • Mole Snacks: +31/-6
Depends on what you are crystallizing. Choice of appropriate crystallization solvent is basically a trial and error process.

Yes the solvents need to be miscible. Initially you dissolve the material into solvent A and add solvent B until slight cloudiness appears. Then you either add solvent A dropwise or heat the solution until you get a clear solution and then cool down the solution to start the crystallization.

It is usually desireable to choose the solvents so that both have very similar boiling and freezing points, but so that one dissolves the material well and the other does not, usually the difference is in polarity.

Typical solvent pairs include: acteone/diethyl ether,  water/ethanol, ethanol/acetone, ethanol/petrol ether, chloroform/petrol ether, and chloroform/CCl4.
Chemists do it periodically on table.

Offline jeffmoonchop

  • Full Member
  • ****
  • Posts: 334
  • Mole Snacks: +37/-5
  • Gender: Male
the crystallisation of a compound depends on its metastable zone width. ie how supersaturated the solution can become before spontaneous nucleation occurs. If the system is only slightly supersaturated it is less likely it will start to crystallise. That's why when you add a miscible antisolvent it needs to be something that will make the system more supersaturated. If the metastable zone width is too wide sometimes you need a very harsh antisolvent. Usually the solubility of something in water and ethanol are too similar so try using longer chain alcohols ie lowering the polarity compared to the water. If the solvents become too immiscible dissolve it in ethanol and try antisolvents then. Water isn't miscible with many organic solvents but ethanol is much better. But if you want high quality crystals I wouldn't use antisolvent addition in turbulent flow because the nucleation rate is too high. I think its better to crystallise by evaporation especially if you have to use water. It takes a long time but the quality of crystals are usually improved. Don't heat the system just find out the solubility of the thing you're crystallising and weigh enough out to be fully dissolved in approx. 20ml or so. Then leave it out for a few days. This way increases supersaturation much more slowly and allows random primary homogeneous nucleation to occur when it hasn't reached the metastable point. This will mean your crystals grow slower giving the molecules more time to arrange themselves properly and not improperly. You'll get the thermodynamic product instead of the kinetic unstable product.

Offline blaisem

  • Regular Member
  • ***
  • Posts: 87
  • Mole Snacks: +5/-0
Thanks for the advice guys. I'm giving it a try now. Apparently another term for the technique is liquid-liquid diffusion, and with that as my search term I found more hits. Some earlier threads from these forums popped up as well which were helpful.

One question: Do you cap the multisolvent mixture to prevent evaporation? It seems like a necessity but ideally I'd like to evaporate some of the solvent as crystals fall out so I can maintain yield. Right now I have a perforated aluminum foil cover as a middleground.

Offline jeffmoonchop

  • Full Member
  • ****
  • Posts: 334
  • Mole Snacks: +37/-5
  • Gender: Male
if you need to know exactly how much solvent you're adding id cap it. Its best to control temperature and evaporation if you want an accurate measurement of how much antisolvent you add because then you can calculate the metastable zone width accurately. But if you don't care about all that and just want crystals faster just leave it open as you add the antisolvent.

Sponsored Links