AIM: To estimate the mechanochemical interconversion of halogen bonded cocrystals with the help of periodic-DFT calculations. In this study, we have examined a more complex case of the formation of three component cocrystals starting from the two component cocrystal.
Methods: We have performed the periodic DFT calculations for the halogen bonded cocrystals under study using CASTEP code. To verify the theoretical outcomes the mechanochemical interconversions were performed and the materials produced were characterized using the X-ray diffraction measurements (SCXRD and PXRD) and thermal studies (DSC and TGA). Further to estimate the accuracy of the calculated interconversion energy, dissolution calorimetry measurements were performed. Additionally, we have conducted molecular energy framework calculations to clarify the role of various intermolecular interactions in favouring the formation of ternary cocrystals. Details of these measurements and calculations are given below.
Mechanochemical Interconversion Reaction
Milling reactions were conducted using a Retsch MM400 shaker mill, equipped with a 10 ml stainless steel jar containing two 7 mm steel balls, operating at 30 Hz for 30 minutes. All reactants were taken in a stoichiometric equimolar ratio, scaled to a total mass of 200 mg. Additionally, 10 μl of liquid additives (including ethanol, acetonitrile, and/or hexane) were added to the reaction mixture. The milled products were dried in open air and characterized using Powder X-ray diffraction (PXRD).
Powder X-ray Diffraction (PXRD) Analysis
The solids obtained from mechanochemical grinding were analyzed using Powder X-ray diffraction (PXRD) measurements. These measurements were conducted on a Bruker D8 Advance diffractometer, equipped with a Lynxeye detector and utilizing CuKα (λ = 1.54184 Å) radiation source, operating at 40 kV and 40 mA. The scan covered a 2θ range of 3 to 40 degrees, with a step size of 0.04 degrees.
Single-Crystal X-ray Diffraction (SCXRD) Measurement
High-quality single crystals of (tftib)(pyr)½ (tpss)½ were grown through slow evaporation from acetonitrile and used for diffraction studies at 100 K. Diffraction data were collected using the Agilent Technologies Super Nova Single Source diffractometer with MoKα radiation (λ = 0.71073 Å), with the data analyzed using CrysAlis Pro software. The structure determination was performed using the SHELX package, where the structure was solved via direct methods, followed by successive least-squares refinement based on the full-matrix least-squares method on F², utilizing the SHELXL program. The hydrogen atoms were positioned on the carbon atoms as determined from the Fourier map. Figures for this publication were prepared using Olex2 and Mercury software.
Thermogravimetryic Analysis (TGA) and Differential Scanning Calorimetry (DSC) Measurements
~2-10 mg of each sample was taken in a standard 70 μl alumina crucible and heated from 30 to 300 °C at a heating rate of 10ºC per minute using Mettler-Toledo TGA/DSC STARᵉ system. All the experiments were performed in the dry nitrogen atmosphere.
Dissolution Calorimetry Measurements
The calorimetric dissolution enthalpies of the cocrystal and its constituents were determined in acetonitrile at 25°C using TAM IV calorimeter. ~ 2.5 –3.5 mg of the solid sample and 15 ml acetonitrile solvent were placed in the sample cartridges and calorimeter cell, respectively. Then the cell was mounted inside the calorimeter and allowed to reach thermal equilibrium overnight followed by stirring at 60 rpm. The related baseline criteria were set as follows: heat flow (baseline slope within 30 min) < 500 nW/h, heat flow standard deviation < 100 nW. The sample was exposed to the solvent by releasing the sample cartridge into the calorimeter cell after the baseline equilibration period. Integration of the heat flow signal was used to obtain the total heat associated with the dissolution process. Finally, the enthalpy of dissolution was measured by subtracting the heat signal for blank samples. The experiments were performed in triplicate.
Periodic DFT calculations
The structure of the (tftib)(pyr)½ (tpss)½ cocrystal obtained from the SCXRD experiment was used as a starting point for the periodic DFT geometry optimization, while the structures of the other cocrystals and individual coformers were obtained from the Cambridge Structural Database (CSD). Covalently bonded hydrogens were normalized to their neutron diffraction values using Mercury, after which the structures were converted into the input format of CASTEP using cif2cell utility for geometry optimization. The unit cell parameters and atomic coordinates were optimized with a constraint on crystal symmetry. The calculations were performed at 800eV plane wave basis cutoff with ultrasoft on-the-fly generated pseudopotentials, combining the PBE functional with either Grimme dispersion correction (D3) or many-body dispersion correction (MBD*) using CASTEP (version 22.11) code. The convergence criteria for the calculation were set as follows: energy = 2 x 10˗5 eV/atom, force = 5 x 10˗2 eV/Å, displacement = 10˗3 Å, stress = 10˗2 GPA. The energies of cocrystal formation were calculated by subtracting the sum of the coformer energy per molecule from the energy of the cocrystal per molecule.
Molecular energy framework calculations
Interaction energy and energy framework calculations were performed with the CrystalExplorer (version 17.5) software package starting from the periodic DFT optimized geometry. The calculations were performed with the B3LYP/DGDZVP method for all the molecular pairs within the sphere of radius 20 Å for each of the symmetry-distinct molecules in the asymmetric unit. The cylindrical tube size was set to 100 for all energy framework calculations.
Conclusion
Periodic DFT calculations were able to assess the phase interconversion of the halogen bonded cocrystals and the formation of the three-component cocrystal. Theoretically predicted interconversion energies were found in agreement with the experimentally observed values. Further, the molecular energy framework calculations gave us a clear understanding of the preference for the three component cocrystal formation over their two-component counterparts.
Supplementary files available here: https://doi.org/10.1039/D3CP04358D