Introduction
The 1967 attempt of structural analysis of the solid-state complex of caffeine (caf) and pyrogallol (pyg) was a pioneering structural investigation in supramolecular chemistry of caffeine. While this might be one of the earliest attempts, if not the earliest, to report structural properties for a molecular cocrystal of caffeine, the heavily used molecule in crystal engineering, the reported structural model showed multiple structural issues such as distortion of aromatic rings of caffeine and pyrogallol from planarity and the non-optimised molecular geometry. Our aim therefor was to revisit this structural model demonstrating this long-overlooked complex is most likely a tetrahydrate with a different structure and composition than initially proposed and provide the crystal structure of the anhydrous cocrystal.
Methods
Computational techniques such as density functional theory, structure determination from powder X-ray diffraction data (PXRD) and Rietveld refinement techniques were employed. In addition, mechanochemical and solution crystallisation methods were used for sample preparation. Multiple instrumentations such as single crystal XRD, variable-temperature (VT-PXRD), thermal analysis, dynamic vapor sorption (DVS) and Fourier-transform infrared attenuated total reflectance (FTIR-ATR) spectroscopy were used for structural, physicochemical characterisation and phase behaviour study.
Results
The crystal structure of the solid-state complex was found to be a channel hydrate with a tetrahydrate (caf·pyg·4H2O) composition. The water molecules arrange into columns extending along the crystallographic c-axis, exhibiting edge-fused hydrogen-bonded pentagons and heptagons. The newly discovered anhydrous structure (caf·pyg) is monoclinic and consists of tetrameric (caf)2(pyg)2 units. Reversible hydration/dehydration behaviour is demonstrated through DVS and VT-PXRD studies.
Conclusion
In conclusion, we re-investigated caffeine-pyrogallol solid state complex, showing that the complex is a tetrahydrate rather than pentahydrate as initially reported. We also reported an anhydrous form of this historically important system and studied their phase behaviour.
