[chepda06]

I am currently working on synthesising a benzenesulfonamide, however the compound crystallises in two crystalline forms.  One polymorph has a melting point range of 197-203°C, while the other melts at 203-210°C - the polymorph I want is the former.  The compound is slightly soluble in water, freely soluble in acetonitrile and methanol and very slightly soluble in DCM.
My current method of recrystallisation is to suspend the compound in water and heat to reflux for 1 h with stirring.  I then cool to 80°C and add enough ethanol to completely dissolve the compound, before cooling further to 75°C and seeding the solution without stirring.  The solution is then left to cool to room temperature, but I usually end up with a mixture of crystalline forms.  Does anybody have any ideas or suggestions on how I could isolate the correct crystalline form?

[sHaY]

I would look for a solvent the the unwanted crystalline form is insoluble and the wanted form is soluble. This way you can filter out the unwanted form while having the wanted one suspended in solution and than isolate it with little of the unwanted crystalline present.

[orgopete]

I would look for a preparation of the lower melting polymorph and use that solvent for the recrystallization. Alternately, you could just try a number of solvents in doing recrystallizations.

This would be my thinking. If benzenesulfonamide preferred to crystallize into the higher melting polymorph, it would likely be unnoticed that a lower melting form existed. Since a lower melting form exists, it must form more readily. I would further guess that higher crystallization temperature form the higher melting form. You might just repeat the recrystallization at a much lower temperature by dissolving it in ethanol and adding water to the cloud point at a much lower temperature. Then cool in an ice or ice-salt bath to crystalize the sulfonamide. (Incidentally, dissolving my compound in the more soluble solvent and then adding a miscible lower soluble solvent is how I do all of my recrystallizations. Heating for 1 hour in a solvent your compound is not very soluble in would be an exercise in futility by comparison.)

[chepda06]

Thanks a lot for your help.  I shall try the recrystallisation at lower temperatures.  I will also try dissolving the compound in ethanol first and then adding water to the cloud point before cooling.  However, I have never been able to obtain the correct polymorph without seeding the mixture, so this may pose a problem.

Also, my final compound must pass several tests, such as melting point, GC, sulphated ash etc.  I am wondering if it is possible for a compound to pass an IR test but fail a melting point test?

[orgopete]

Certainly, you can seed the saturated solution to enable crystallization.

The molecular properties of a polymorph will only apply to properties associated with this crystalline form, specifically the mp. I don't know if there are any IR effects with polymorphs. If someone with experience could answer? The GC or solution properties would be unaffected.

[AC Prabakar]

Yes,if there is a change in polymorph the finger print regions of IR will vary.You may not get the corelation factor 0.99 in IR.

The seeding problem can be addressed in the following manner:-

Take the seed compound in small amount if solvent and to this solution add your dissolved compound.This protocol will help to avoid high seeding % with out affecting the purpose of seeding.

To get one polymorph consistently you have to study lot of parameters like,
1.Cocn. of soln.
2.Thermodynamically Stable form and kinetically stable form(To operate at different temperature.)
3.Addition and cooling ramp(Especially to selectively crystallise one polymorph)

[chepda06]

Thanks very much for the helpful advice from everybody.  One more thing - is it possible for a sample to pass a melting point test but fail an IR test?

[AC Prabakar]

It can be!!.
The reason is,in melting point we will have some range(generally as we will pronounce as MR),where as in IR it is not possible.But there will not be any change in the characteristic functional groups stretching.(you can check this phenomenon by simple example/or case study)