Polymorphism Prediction in Crystal Systems of Technological Interest
What we offer:
Predicting the most probable polymorphic forms of a molecular crystal is a field which has been studied extensively in recent years. As a result, a large number of predictive software applications have been developed and are currently available. However, the performance and the applicability of these emerging applications are still limited, as it has been demonstrated by the results of tests performed and published by the Cambridge Crystallographic Data Centre.
The research centre which is member of the Reference Network on Theoretical and Comptutaional Chemistry (XRQTC) has identified the source of error which is mainly caused by the use of atom-atom intermolecular potentials, which are limited when describing weak hydrogen bonds or Van der Waals interactions. Thus, the integration of intermolecular potential within the calculations delivers similar results to those obtained with advanced ab initio calculations. For the predictive studies Pixel intermolecular potentials are quantified. The code developed PixCryPar is an MPI-parallelized version that allows to predict in a cost-effective way the most likely crystal structures in space groups adopted by the substances.
To perform a prediction satisfactorily, one has to describe the interactions within the crystal, i.e. the intermolecular interactions. By means of PixCryPar these potential interactions are described using new generation algorithms, known as potential-pixel, which are based on the quantification of the molecular electron density concentrated ab initio for a given volume or pixel, and compute power as a sum of intermolecular interactions between pixels. This allows a very precise description of intermolecular energy for any molecular crystal or co-crystal.
The predictive method has been tested for a set of 36 molecular crystals of varied structure, complexity, and application (drugs, dyes, etc.). The results show excellent reproducibility and accordance with those obtained experimentally.
Currently, the team is optimizing the algorithm to gain a powerful and reliable software that would allow companies from the pharmaceutical sector, and from other sectors for which the solid state of a substance has a strong influence on the production processes, to predict the crystalline structure of solids.
What are we looking for:
We are looking for companies interested co-developing together the software and we are interested also in provide these services to companies.
- ] Pharmaceutics
- ] Computational chemistry