*3.1. Solid-State Characterization of the Solids Obtained from Slurry or Mechanochemical Methods*

The cocrystallization of 9-ethyladenine and oxalic acid was carried out in 1:1 or 2:1 (9ETADE:OXA) molar ratios. On the one hand, by mechanochemical process by neat grinding (NG) or liquid-assisted grinding (LAG), which have previously demonstrated to be fast methods for salt/cocrystal screenings. On the other hand, slurry experiments in water or acetonitrile were also performed with the purpose to check whether other plausible solid forms or solvates were possible. In solution, we found that for these two molar ratios and the two solvents, up to four different powder patterns were obtained (see Figure 1). The solids were identified as shown in Scheme 1b (compounds **1**–**4**).

**Figure 1.** Comparison of the experimental and calculated powder patterns of 9ETADE, oxalic acid dihydrate and the as-synthesized compounds **1**–**6**.

During the mechanochemical screening, for the 2:1 molar ratio by NG or LAG in methanol or water, the powder patterns of the resulting solids were identical to the diffractogram of compound **4**. However, for the 1:1 ratio, the situation was much more complex. Use of water or methanol afforded powder patterns not only different between them but also to the ones obtained in solution with water or acetonitrile. The solid obtained by LAG in water was identified as compound **5** and the solids obtained by NG or by LAG in methanol, as compound **6**. The diffractograms of solids **5** and **6** are shown in Figure 1).

Between the powder patterns of solids **2** and **5** some differences can be observed, and therefore they will be treated as two different forms. Even if these two solid forms show some characteristics in common not only in their FT-IR spectra but also in their TGA-DSC traces, as it will be commented below. Among all the other forms or respect to the starting products, characteristic and distinguished peaks are clearly observed. Furthermore, the agreement among the experimental and calculated patterns from single crystal data for compounds **1** and **3** was excellent. Surprisingly, the shortest dicarboxylic acid we used resulted the most fruitful in rendering different solid forms.

All the new compounds were analyzed by Fourier Transform Infrared spectroscopy in Attenuated Total Reflectance mode (ATR-FT-IR) and thermal methods (TGA-DSC) to elucidate the nature of these materials. Changes in the IR spectra of the new solids respect to the starting compounds have provided evidence that new H-bond interactions have taken place. Moreover, the presence or absence of several characteristic vibration modes, as for instance NH2, COOH, C=N, gave further information of these new contacts. The IR spectra are included in Figure 2.

**Figure 2.** FT-IR spectra of (**a**) 9ETADE, oxalic acid dihydrate and the as-synthesized compounds in the full range, and in the ranges (**b**) 3600–3000 cm−<sup>1</sup> or (**c**) 1850–1500 cm<sup>−</sup>1.

The main difference among compounds prepared from aqueous slurries (**1** and **3**) with respect to all the others is the absence of one or two bands in the region 3450–3400 cm−<sup>1</sup> (see Figure 2). While for these two solids no bands are observed, confirming the anhydrous state observed in the thermal analysis, for all the other compounds (**2**, **4**–**6**), two bands at ca. 3446 and 3407 cm−<sup>1</sup> were present, which could be assigned to OH from water, and from the hydroxyl of unprotonated oxalic acid. Some differences in the NH region can be also detected.

In the spectra of the starting compounds, the also characteristic ν(C=O), ν(C=N), and δsciss (NH2) vibrations appear at 1654, 1610 and 1669 cm−1. For some of the compounds (**1**, **4** and **6**), only a broad band coincides at approximately at 1686 cm−1, or for **3** and **5**, at 1678 cm<sup>−</sup>1, in agreement with previous observations [9]. However, for the solid **2** a doublet overlapped band at 1694 and 1678 cm−<sup>1</sup> can be detected. Moreover, for compounds **1**, **2** and **4**–**6**, a small band appears between 1728–1770 cm<sup>−</sup>1, which confirms again the presence of C=O from unprotonated –COOH from oxalic acid [40]. No band is observed in this region for compound **3**.

By simultaneous thermogravimetric analysis (TGA)—differential scanning calorimetry/differential thermal analysis (DSC/DTA)—the stability of the new salts was evaluated. Both compounds **1** and **3** were anhydrous according to their TGA traces (Figures S1 and S2, respectively). In the DSC trace for compound **3** only one big endothermic peak corresponding to the melting process (melting peak value of 227.5 ◦C) and concomitant degradation was observed. This is the highest melting point for the solid forms prepared in this work. Thus, this is the most stable form. For compound **1**, the DSC showed a previous small *endo* with an onset temperature of 176.2 ◦C followed by two overlapped endothermic processes with an onset temperature for the first one at 211.6 ◦C. The second endothermic peak (Tpeak, 227.8 ◦C) agrees with the one observed for compound **3**. This event would suggest the loss of an oxalic acid molecule to achieve the salt **3** in 2:1 molar ratio.

On the contrary, compounds **2** and **4** showed each one a loss on drying (lod), of 2.74 and 4.19%, respectively, until 125 ◦C (Figures S3 and S4). They were assigned to dehydration processes of 1/3 to 1/2 of molecule of water for **2** (being probably a nonstoichiometric hydrated) and one molecule of water for **4** (theor. lod of one H2O, 4.16%). In their DSC after these desolvation processes (at Tpeak 107.2 ◦C for **2** and Tpeak, 111.5 ◦C for **4**), the melting of the solids followed by degradation was observed.

The TGA-DSC trace for compound **5** resembles to the one of **2**. In its TGA a loss of 2.6% was observed closer to the theoretical value for a third hydrate compound (2.32%) and appeared in the same region. Besides, in the DSC an endothermic desolvation process was observed (peak temperature at 111.7 ◦C) and a subsequent melting (melting peak temperature, 187.9 ◦C) and its degradation (Figure S5).

Finally, for compound **6**, in the TGA trace a mass loss of about 5.8% was observed before complete degradation of the solid, suggesting that a new solvate was formed. The theoretical loss for a monohydrate solid form is 6.6%. The DSC thermogram showed several endothermic events, unveiling a complex desolvation process for this solid form (Figure S6).
