Recrystallization and the melting point experiment

[psychokitty]

Does it really make a difference?  I need to determine a compounds authenticity and it looks like checking its mp is the way to go.  However, I can't help but wonder how pure a sample has to be in order for the mp to be right on the mark.  Does anyone have any practical experience to offer in this area? 

Also, can anyone tell me whether or not applying the mp experiment has personally helped them in establishing the identity of an unknown substance?  And if so, how pure did that substance have to be--in terms of recrystallizations that had to be performed--in order to get an accurate measurement?  Also, if the mp is depressed, how depressed is still okay to know what you've got?

[hest]

To get a sharp mp you sampel need to bee clean. mp within 0.1-.03°C is a good sign (usual 2 times rextalisation).
Iff you have a good clue abouth the product and the mp is right then it's the right product.

[psychokitty]

So what you're saying is that if a literture mp of a substance is at say 195 deg C and the observed melting point is 189-193 deg C then this is a good indication that the substance is what it should be?

[Rhodium]

I believe you rather should recrystallize the substance 2-3 times, so that it will become pure, and then it will have a mp of say 194-195°C if the literature value is 195°C.

If your substance is not pure, then its melting point is of little value for identification purposes. If you have an authentic sample, a mixed melting point could be taken (and the melting point should not be depressed compared to either of the samples), and that is a good identification method. If the samples aren't the same, the mix mp will be lower.