Case for characterisation of PZQAmine (or any other resolved intermediate)

Published by ndt228 on 7 November 2010 - 7:38pm



Now that both racemic and single enantiomer PZQAmine are available, it is possible to determine phase diagrams - either binary, based on MPt, or ternary based upon solubility.  But why do this ?
The essence of the matter is that options for resolving a racemic mixture or or purifying a partially resolved mixture are defined in the phase diagram, which is a represention of the significance of the solid crystalline form(s) which may exist.  Pasteur's original resolution of tartaric acid was achieved by hand-picking apart individual crystals of the two enantiomers, but this was only possible because it naturally crystallises as what is called a conglomerate (ie it spontaneously resolves into separate crystals)(and that these were recognisably different, and physically large enough to handle).  This resolution self-evidently happens  because this is a "lower energy option" than if the molecules of (+) and (-) were to try to arrange themselves into the same crystal.  If any of the crystals are characterised they will be found to be identical (xrpd, IR, MPt, DSC etc), with the exception of the sense of their chirality.
More often, a mixture of enantiomers will be happier to crystallise as a "racemic compound" in which (typically) pairs of opposite enantiomers first associate (think shaking hands), and then the pairs stack to form the larger crystal lattice.  In such cases, if the single enantiomer can be isolated or made, it will typically crystallise in quite a different way, since it is unable to pair with itself in the same way it can with the opposite form.  Thus crystals of the racemate and the single enantiomer will have different physical properties (xrpd, DSC, IR, MPt etc).   
If you then consider a mixture of single enantiomer and racemate crystals, the two types of crystals behave effectively as impurities to each other, and an intimate mixture will display a depressed melting point relative to either "pure" crystal type.  Somewhere on the scale from 100% racemate to 100% single enantiomer the melting point of the mixture will exhibit a minimum, called the eutectic composition.  This is highly significant because if by some means you happen to obtain a euctic mixture it will be impossible to change its composition by crystallisation, either from the melt or from a solution.  [caveat: - unless you can change the crystal form, but that's another story] 
If you obtain a composition between that of the eutectic and the single enantiomer, then it will in theory be possible to crystallise your mixture up to enantiopurity.  It is even possible to predict the ultimate possible yield/efficency given some basic data.  However, if you have a composition between that of the eutectic and the racemate, you will never be able to obtain a pure enantiomer by crystallisation - instead it will be the racemate that crystallises first, and then once the liquor composition has reached the eutectic composition, it will be the eutectic that crystallises.
NB There is no way I know to predict where the eutectic will lie - it has to be determined experimentally. 
So, when performing a diastereomeric salt resolution, it remains important to understand the crystallisation properties of the "free amine", as well as to know the behaviour of your salt.  Without this, you cannot tell whether you need to recrystallise your salt up to enantiopurity before you crack it back to the amine, or whether you can rely upon crystallisation of the amine to play its part.  Obviously it would be desirable to get 100% de salt from your resolution - but this too is subject to the energetics of that system, as expressed in its phase diagram.  The latter are (even) more complicated than what I have described above, and beyond my scope to adequately explain, though there is plenty to be found in textbooks.
I have attached a file with a picture of a typical binary phase diagram, and some explanations...


MatTodd's picture

Nick - thanks for pointing this out, which we had completely failed to address. I had always been working on the assumption that the effects you are talking about were small, but that's a bad assumption here. It'll be fascinating to see how this applies to PZQamine, and whether our crystallizations are going to go in the direction we want. [see here for Nick's offer to do the analysis]

ndt228's picture

Just a quick note to acknowledge receipt of the rac-PZQ and (R)-PZQ samples sent on 31st Jan.  They both look very pretty under a microscope with crossed polaroids, but from the very similar melting points declared on the jars it may be more challenging than I had hoped to get a phase diagram from DSCs at various ratios.  I will do my best, and let you know in due course.
Nick (Almac Sciences)