**3. Results and Discussion**

### *3.1. Characterization of ITR and Terephthalic Acid (TER) Mixtures Following Neat and Liquid-Assisted Grinding*

Mechanochemical methods of cocrystal production have gained a considerable interest in recent times [25,26]. Therefore, as the first approach, a neat, solvent-free (NG) grinding of ITR and TER in a few stochiometric ratios, followed by the liquid-assisted grinding (LG) was performed. The ITR "as received" material was identified as form I of itraconazole [27]. The PXRD analysis of the samples prepared by neat grinding (NG (Figure 2)) revealed that the mixtures post-processing had similar di ffractograms to those of ITR with two Bragg peaks at 17.5 and 17.95◦ 2θ, corresponding to TER. The samples prepared by LG had additional, but weak, Bragg peaks at 3.5, 7.0 and 21.2◦ 2θ, which were absent in the parent materials, indicating the formation of a new phase. Interestingly, the TER component in all samples post NG and LG appeared to be at least partially amorphous, as evidenced by the weak or absent di ffraction peaks of TER. Therefore, the LG method is more e fficient in producing the ITR–TER cocrystal than NG most likely because the cocrystallisation process by LG was facilitated by the presence of methanol. Indeed, the use of solvents in cocrystallisation screening is common and the impact of the solvent is described as catalytic, since it is used in a very small quantity and is also not part of the final cocrystal [28].

A DSC analysis of ITR and TER on their own showed that the drug melted at around 166 ◦C [12], while TER melted with sublimation at around 350 ◦C [29] (Figure 3). A thermal analysis of the binary ITR–TER mixtures post-processing showed thermograms with a sharp endothermic peak at 198 ◦C (Figure 3), which was absent in the parent compounds and indicated the melting of the new phase. However, all these mixtures also had an endothermic peak with an onset at 164 ◦C (Figure 3), assigned to the melting of ITR. This peak in the ITR–TER 1:1 (LG) and ITR–TER 1:2 (LG) systems was almost imperceptible, with enthalpies of 2.6 and 0.7 J/g, respectively. In contrast, the ITR peak in the samples ITR–TER 1-1 (NG) and ITR–TER 1-2 (NG) had an enthalpy of 13.5 and 18.4 J/g, respectively, indicating that the methanol used in LG contributed to a greater conversion of ITR into the new phase. A small exothermic peak was detected for all NG samples and ITR–TER 1:2 (LG) immediately followed the ITR peak, consistent with the previous studies on the thermal behavior of binary mixtures where the formation of a cocrystal was observed upon heating [30]. Thus, in relation to NG mixtures, the cocrystal melting peak, visible in DSC traces, is of the new form which appears upon heating. Additionally, what can be concluded from the DSC study is that a broad endothermic peak with an onset at around 310 ◦C for the ITR–TER 1:1 samples and an onset at around 320 ◦C for the ITR–TER 1:2 systems was detected, assigned to the excess of TER. It may sugges<sup>t</sup> that the potential stoichiometry of the ITR–TER cocrystal might be 2:1.

**Figure 2.** Powder X-ray diffraction (PXRD) of the ITR and TER starting material as the received powders ("ar") and binary ITR and TER mixtures following neat grinding (NG) and liquid-assisted grinding (LG). **\*** Indicates the distinct peak of the new phase (cocrystal), broken black lines show the position of the ITR diffraction peaks, while the broken red lines show the position of the TER diffraction peaks.

**Figure 3.** Differential scanning calorimetry (DSC) of the ITR and TER starting material, as received powders ("ar") and binary ITR and TER mixtures following neat grinding (NG) and liquid-assisted grinding (LG). The area highlighted in grey indicates the ITR melting region, in purple the cocrystal melting range and in light red the TER melting and degradation. Arrows indicate the presence of an exothermic peak immediately following the ITR melting event.

### *3.2. Properties of ITR and TER Samples Made by Slurring, Evaporation and Ball Milling Methods*

A PXRD analysis of samples obtained by the slurring of ITR and TER in methanol and acetone was very similar to that of TER and indicated an incomplete conversion to the cocrystal, with only weak intensity Bragg peaks of the new phase visible at approximately 7.0, 10.4, 12.4◦ 2θ (Figure 4). A more successful method was the slow evaporation of ITR and TER from methanol, and the diffraction pattern of the 2:1 ITR–TER system is presented in Figure 4. A number of diffraction peaks characteristic of the cocrystal phase can be discerned, however the peak at 15.0◦ 2θ can be described as the ITR starting material powder. The method that gave the purest cocrystal, based on the PXRD results, was ball milling (Figure 4). The ITR–TER 2:1 system had a diffraction pattern distinct from those of the starting material powders, particularly due to the peaks at 3.5, 7.0, 10.5, 12.4, 17.8, 19.3 and 21.2◦ 2θ. Therefore, ball milling, as the most efficient method of cocrystal production, was used to prepare the cocrystal for further characterizations.

**Figure 4.** PXRD patterns of the ITR and TER starting material, as received powders ("ar"), samples produced by the slurry method from methanol (ITR–TER SL MeOH) and acetone (ITR–TER SL acetone), the ball milled sample containing ITR and TER in a 2:1 stoichiometric ratio (ITR–TER 2:1 BM), the sample obtained by the slow evaporation of ITR and TER in a 2:1 stoichiometric ratio from methanol (ITR–TER 2:1 SE) and a powder diffraction pattern of the ITR–TER cocrystal calculated from the single crystal X-ray data (ITR–TER CC SC). **\*** Indicates the peak of ITR, broken lines indicate the position of key diffraction peaks of the cocrystal, while the insert shows the presence of the cocrystal peaks in the samples obtained by the slurry method.

A DSC analysis (Figure 5) confirmed the above XPRD data. The samples prepared by the slurring of ITR and TER in methanol and acetone had low magnitude peaks with an onset at 195 ◦C, which corresponded to the melting of the cocrystal phase. In both samples this peak had the normalized enthalpy of transition of 8.1 J/g. DSC also confirmed that these samples were mainly composed of TER. The sample obtained by solvent evaporation displayed a characteristic event of cocrystal melting, although the peak was broader than that of the ball milled sample. The thermal analysis of the ITR–TER 2:1 ball milled system (Figure 5) verified that ball milling was more efficient in producing the cocrystal, with a sharp melting peak with a normalized enthalpy of 83.7 J/g.

**Figure 5.** DSC thermograms of the ITR and TER starting material, as received powders ("ar"), samples produced by the slurry method from methanol (ITR–TER SL MeOH) and acetone (ITR–TER SL acetone), the ball milled sample containing ITR and TER in the 2:1 stoichiometric ratio (ITR–TER 2:1 BM) and the sample obtained by slow evaporation of ITR and TER in the 2:1 stoichiometric ratio from methanol (ITR–TER 2:1 SE). The area highlighted in grey indicates the ITR melting region, in purple the cocrystal melting range and in light red the TER melting and degradation.
