A new and systematic method for co-crystal screening has been developed, based on improvements in the understanding of the thermodynamic factors that influence co-crystal formation. This method works from the premise that pure component solubilities determine the concentration regions to screen for new co-crystals, rather than the stoichiometry of the co-crystal. An extended phase diagram screen gives the composition ranges in which the co-crystal is the stable crystalline form. The method is based on the measurement of saturation temperatures which are experimentally easily accessible using standard laboratory equipment. The method has been validated using both carbamazepine and cinnamic acid with a number of co-formers in ethanol solvent. New co-crystals of carbamazepine with isonicotinamide, benzamide and 3-nitrobenzamide, and of cinnamic acid with 3-nitrobenzamide have been discovered.

Dedicated to Professor Manfred T. Reetz on the occasion of his 60th birthdayThe exact positioning of molecules on surfaces has become essential for the development of molecular electronics [1] and biochip applications, [2] as well as for conducting single molecule experiments. Commonly, one of two possible routes is followed, either by chemically reacting molecules to surfaces or to anchoring molecules present at a surface, or by physisorption, that is, the use of nonspecific physical interactions between molecule and surface, and between molecules themselves. Covalent chemical modification usually requires synthetic effort, is hard to control at surfaces, and does not allow either self-correction or (intentional) desorption. Physisorption does allow self-correction, which is common for all self-assembly processes, but the thermodynamic and kinetic parameters governing adsorption and desorption are difficult to control. Moreover, neighboring molecule±molecule interactions are usually needed for obtaining stable and ordered layers, thus the formation of densely packed layers is a prerequisite. [3] Supramolecular interactions serve as an intermediate case with the possibility of employing advantages from both routes. Supramolecular interactions, such as those observed in host± guest complexes, are specific and directional, and a wealth of information is usually available on their binding strengths and kinetics. The application of molecules that allow the formation of multiple supramolecular interactions provides a tool to tune adsorption and desorption process parameters because thermodynamics and kinetics are, in principle, straightforwardly related to the number of interactions, and the strength and kinetics of an individual interaction.It is therefore our goal to apply host surfaces as ™molecular printboards∫ at which multivalent guest molecules can be positioned. Prerequisites will need to be identified to reach thermodynamically and/or kinetically stable assemblies that can be employed in nanotechnology. It is envisaged that one can control adsorption and desorption by environmental stimuli, for example, by competition with another host in solution. Herein we address the thermodynamic and kinetic COMMUNICATIONS (TUE) are gratefully acknowledged for supplying the adamantyl-terminated dendrimers. Frank Geurts (Akzo Nobel, Arnhem) is acknowledged for performing the XPS measurements, and Paul Koster for assistance with the grazing-angle FTIR measurements. Szczepan Zapotoczny is acknowledged for performing the AFM measurements. The Dutch Technology Foundation STW is gratefully acknowledged for funding the Simon Stevin Award project on Nanolithography.

No abstract

Isonicotinamide (INA) co-crystallizes with carbamazepine (CBZ), as do nicotinamide (NA) and benzamide. The structure of CBZ-INA form II is solved from powder and is shown to be isostructural with CBZ-NA. However picolinamide (PA), despite its similarity to the other pyridine carboxamides in the homologous series, does not appear to form a co-crystal with CBZ. We compare and contrast the use of computed crystal energy landscapes and binary and ternary phase diagrams to explain this behavior. Two 1:1 co-crystal structures of CBZ and INA were predicted to have lower or comparable lattice energies than the sum of the pure component lattice energies. These structures corresponded to the known co-crystal structures. On the other hand, lattice energies of predicted CBZ-PA co-crystal structures were less stable than the pure component lattice energies, implying that CBZ and PA would not form a co-crystal. This is consistent with the experimental evidence. Examination of the hypothetical crystal structures for CBZ-INA and for CBZ-PA explains this in terms of intermolecular hydrogen-bonding capability. Thus computed crystal energy landscapes are more reliable than simple crystal engineering concepts in understanding co-crystal formation, and can provide a useful complement to experimental co-crystal screening.