How did you get interested in this research project?

Up until the year 2000, most academics interested in crystallisation of organic solids had restricted themselves to neutral molecules, despite the overwhelming importance of molecular salts as active pharmaceutical materials. At about this time, AstraZeneca supported one of my Ph.D. students, Ed Collier, in a thesis involving the isolation and solubility measurements of 17 salts of ephedrine. We had a lot to learn, particularly on the definition of a precisely controlled crystallisation process in a salt system. Fortunately, Ed was joined by a second Ph.D. student, Helen Jones, funded by Sanofi-Aventis to explore further aspects of salt crystallization. One of the materials identified from a database search was this polymorphic compound ethylene diammonium dinitrobenzoate. Helen proceeded to measure important thermodynamic parameters and to carry out crystallisations at defined pH and supersaturation values, something that to our knowledge had not been done before.

What is the most important result in the paper?

We found that of the two forms, the monoclinic one was always the most stable. The rule of thumb that is known as 'Ostwald's Rule' would lead us to believe that in such cases, the least stable form would always appear first. In this case, however, we identified conditions in which the most stable phase appeared directly from solution. I know of no other case in which this happens. I suppose the most important point that this reflects is how very little we know about the process of nucleation - this remains a central challenge in physical chemistry today.

This paper also points out that stereochemically related additives or impurities can have a significant impact on the outcome of such a process. Given the work of Leiserowitz, Weissbuch and Lahav over the last twenty years this is perhaps not surprising but nobody had ever shown it before.

What are the implications of the results you present in this paper?

We have much to learn about the structural pathways that are available at the point of nucleation. In the pharmaceutical industry, the central commercial activities of salt selection and process development cannot be robust and hence quality cannot be designed into salt development activities without taking good care of the kinds of crystallisation issues raised here.

Are there any particular challenges facing future research in this area?

A couple of really key things: what happens in nonaqueous solvents in terms of speciation and its impact on nucleation, and how can we learn more about structural aspects of nucleation.