Sciencemadness Discussion Board

Recrystallization

cnidocyte - 22-8-2010 at 12:50

I'm unsure of how recrystallization really works. Lets say I have a mixture consisting of 40% Salt A and 60% Salt B and need to separate the 2. First scenario lets say they are equally soluble in the solvent I choose and I dissolve the mixture in the hot solvent then cool to allow the salts to crystallize out. Will 1 salt crystallize out before the other? Would it be Salt B because theres more of it? Would a mount of only Salt B crystals keep forming until the solution is no longer supersaturated or would Salt A simultaneously start to crystallize out. If its the latter then would Salt A crystals form in a different area to Salt B?

2nd scenario lets say the 2 salts have different solubilities in the solvent I choose. Would it be only the less soluble salt that crystallizes out until the solution is no longer supersaturated?

12AX7 - 22-8-2010 at 16:52

What's the solubility curve of each salt?

Shall we assume they do not form a compound salt (like KCl + MgCl2 <==> KMgCl3), and will not result in a double displacement reaction (like 2KCl + Na2SO4 <==> 2NaCl + K2SO4)?

Using KCl + NaCl as an example, NaCl's solubility curve is level, so you will precipitate KCl during cooling (with constant solvent), and NaCl during evaporation (at constant temperature). This breaks down at the edges, where NaCl's temperature coefficient is noticable (<5%), or where KCl is too weak to precipitate, etc.

Subsequent recrystallizations of the products will increase purity, by precipitating the majority product, partitioning the impurity into the solvent (at the expense of some of the major product, which can be fed back into the process to seperate it again).

This process overall can be very tedious, where both salts have low tempcos or similar solubilities, or where compounds are formed, etc. I've seen at least a few occasions where a range of salts are precipitated simultaneously, e.g., NaCl, CuCl2 and Na2CuCl4. Simultaneous precipitation is analogous to freezing a eutectic mixture; an unpleasant fine-grained mixture results.

Tim

[Edited on 8-23-2010 by 12AX7]

crazyboy - 22-8-2010 at 18:23

Generally recrystallization is not used for the type of scenario you described unless they have very different solubility. Generally I would use recrystallization when I have a solid crystalline product contaminated with an oil or other liquid or if the contaminant is highly soluble in the solvent at a range of temperatures but the product has a steep solubility curve.

psychokinetic - 23-8-2010 at 01:35

Yer, in most cases I'd think you'd get both crystals.

cnidocyte - 24-8-2010 at 08:51

Quote: Originally posted by crazyboy  
Generally recrystallization is not used for the type of scenario you described unless they have very different solubility. Generally I would use recrystallization when I have a solid crystalline product contaminated with an oil or other liquid or if the contaminant is highly soluble in the solvent at a range of temperatures but the product has a steep solubility curve.


So how do you separate salts without the addition of another salt? Lets say you had a 1:1 mixture of NaOH and NaNO3 and needed to separate the 2 how would you do it? Adding enough nitric acid you could get pure NaNO3 but you'd lose the NaOH. You could try melting one salt but in this case theres only a 10 degree difference in melting point. You can probably come up with a long, complicated array of metathesis reactions that will result in the separation of the 2 but this kinda thing is a pain in the ass, is there not a simpler way to do this kinda thing?

ScienceSquirrel - 24-8-2010 at 09:03

It all depends on the different solubilities, for example Lo Salt which is a mixture of potassium and sodium chloride is fairly easily fractionally crystallised to obtain quite pure potassium chloride.
A hot saturated solution will preferentially deposit potassium chloride leaving the sodium chloride in solution.
Potassium chlorate has a very steep solubility curve so potassium chlorate will crystallise almost completely from any solution containing potassium and chlorate ions leaving sodium chloride, potassium chloride, etc in solution.

See the solubility table here

http://en.wikipedia.org/wiki/Solubility_table

[Edited on 24-8-2010 by ScienceSquirrel]

cnidocyte - 24-8-2010 at 10:40

Ah right I get it. In my example the solubility curve of NaOH has a slope of 1.5 whereas NaNO3 curve has a slope of 0.8 so I'm guessing NaOH could be crystallized out in this case.