Active Pharmaceutical Ingredients (API) are produced in the solid-state in a variety of crystalline forms or as amorphous solids. Different crystalline forms are possible including pure compounds polymorphs’, co-crystals, solvates (hydrates) and salts.
At any given pressure and temperature, the most stable polymorph is the thermodynamically preferred and all others (called metastable) should spontaneously convert into it. However, some metastable polymorphs can be kinetically stable and present different physicochemical properties (solubility, dissolution rates, hygroscopicity, physical and chemical stability …) affecting their mechanical properties and bioavailability.
Therefore, the discovery of new polymorphs becomes critical. An early identification of the most appropriate crystalline form of an API is essential in the development of new drug candidates and for the patent protection.
The polymorph prediction methods can be useful to find the most appropriate polymorphic forms and, compared to the slow, expensive and non-environmental friendly conventional experimental screenings, several virtual methods have been developed.
The “Blind Tests” carried out by the Cambridge Crystallographic Data Centre have confirmed that polymorph prediction methods are a reliable tool to predict the most stable polymorphs and the most stable of the metastable polymorphs, with a reasonable accuracy (< 50% of success).
The virtual system (REG) offered by CIRCE matches or improves that ratio of success by reproducing the energy of molecular solids, their optimum geometry, and the energy landscape of various molecular crystals. Most polymorph prediction technologies are based on intermolecular atom-atom potentials. Recently, ΔH energy minimization techniques have been employed with GRACE potentials.
Our theoretical approach, a ΔH energy minimization technique, uses intermolecular energies computed using the semi-classical Pixel intermolecular potential. Such approach describes the interaction energy between two molecules as the sum of the exchange-repulsion, electrostatic induction, charge transfer and dispersion components. We have used this approach in a parallel computer code (PIXCRYPAR) that predicts the structure and energy of any molecular crystal, using as starting point the geometry of the molecule.
The quality of the PIXCRYPAR predictions was demonstrated in the prediction of the aspirin polymorph. Zaworotko et al. (J.Am. Chem. Soc., 2005 (127), 16802-16803) reported the existence of a new crystal form, form-II, and Desiraju and Boese (Angewandte Chemie-Int. Ed., 2007, 46, 615-617) demonstrated that the form-II had mixed domains of the old form-I and the new form-II and was almost isoenergetic with form-I. Using the Pixel potentials and PIXCRYPAR, we predicted the new polymorph and other properties detected experimentally. The ability to evaluate polymorphs that exhibit such small difference demonstrates the high throughput of the PIRCRYPAR polymorph prediction program.
The results obtained by PIXCRYPAR studies can also be used as a starting point for ΔH geometry optimizations. The unit cell can also be optimized when needed and, on top of this, we can also perform Ab-initio Molecular Dynamic (AIMD) studies of crystals of special interest.
These theoretical tools have been developed in the Molecular Materials Structure Group (School of Chemistry, University of Barcelona), under the direction of Professor Juan J. Novoa. All the previous approaches allow a reduction in the costs, residues and time necessary to obtain new stable crystalline forms with the desired physicochemical properties. For companies working on drug discovery (or in developing functional molecular materials), such reduction gives them a great improvement that allows them to be more competitive, in the highly competitive global market environment.
In CIRCE, as a result of our high principles and since we are a company that offers pharmaceutical services, we are interested in collaborating with our clients in a very active manner, becoming their partners and creating long-term relations. For this reason, we propose global solutions through unique and state of the art technologies, to solve problems related to your APIs as well as developing protection strategies and/or extending the intellectual properties of your New Chemical Entities (NCEs).
You can download our Virtual Polymorph Screening technical file here.
For further information, please Contact us.