English

• Drug polymorphism, the phenomenon where a drug exists in multiple solid forms, has attracted significant attention from the pharmaceutical industry due to its impact on altering physicochemical properties [1-3]. The molecular rearrangements within a crystal lattice lead to different solid forms, while hydrates [4-6] and solvates (pseudo-polymorphs) form [7-9] when small guest molecules, like water or solvents, are incorporated during preparation. These solid forms can undergo phase transitions under specific conditions, such as polymorph interconversion [10-12], hydrate formation [13-15], and solvate desolvation [16,17], which may occur at various stages from pre-formulation to commercial oral preparations. Phase transitions in active pharmaceutical ingredients (APIs) can significantly alter their physicochemical properties, influencing key characteristics such as stability [18-20], flowability [21,22], hygroscopicity [23,24], permeability [25], dissolution rate [26,27], and solubility [28-30]. These alterations can reduce drug efficacy or, in some instances, present clinical risks [31]. For example, carbamazepine and ritonavir were recalled due to the phase transformation, which led to changes in dissolution characteristics and subsequent clinical failure [32-34]. This highlights the crucial need for monitoring and managing phase transitions in pharmaceutical formulations to minimize potential clinical hazards. Failing to address phase transitions can lead to medication degradation, resulting in considerable financial losses and posing threats to patient health. Compared to the extensively studied phase transitions of single-component drug polymorphs [35-38], multi-component drug polymorphs present a greater challenge due to their more complex structures and the multiple interactions between molecules within the crystal lattice [39,40]. Investigating phase transitions in these systems is a crucial step in developing more stable formulations, as it provides valuable insights into the behavior of solid forms and helps prevent undesirable transitions [41,42]. Despite their significance, phase transitions in multi-component drug polymorphs remain relatively underexplored.

  In this study, ciprofloxacin (CIF, Fig. 1a), a broad-spectrum antibiotic from the fluoroquinolone class, known for its efficacy against various Gram-positive and Gram-negative bacteria [43,44], was selected as a model drug. Diclofenac (DFA, Fig. 1b), a non-steroidal anti-inflammatory drug (NSAID) with established antipyretic and analgesic properties, was chosen to form a multi-component system with CIF. The objective was to enhance CIF’s physicochemical properties through potential synergistic effects [45,46]. The fundamental properties of three novel CIF-DFA forms, anhydrate, dihydrate, and methanol solvate, were investigated, with a focus on the influence of water activity and solvent vapor on their phase transitions. Understanding these mechanisms is essential for optimizing storage conditions, guiding formulation development, and improving the stability of the final product.

  Figure 1

  

  Figure 1.  Chemical structures of (a) CIF and (b) DFA. CIF-DFA dihydrate: (c) Hydrogen bonding network, (d) layered structures, and (e) packing diagrams. CIF-DFA methanol solvate: (f) hydrogen bonding network, (g) layered structures, and (h) packing diagrams.

  DownLoad: Full-Size Img PowerPoint

  The crystal structures of the CIF-DFA dihydrate and methanol solvate were solved and analyzed for the first time. CIF-DFA dihydrate crystallizes in the monoclinic P 21/c space group, with an asymmetric unit containing one CIF cation, one DFA anion, and two water molecules. In the CIF-DFA system, the piperazine ring N3 is protonated and forms a charge-assisted hydrogen bond (N3⁺-H···O5⁻, 2.669 Å) with DFA [47]. Two water molecules form hydrogen bonds (O4···H—O6, 2.741 Å; O5···H—O7, 2.714 Å) with DFA, and these water molecules are involved in hydrogen bonding (O7-H···O6, 2.820 Å), contributing to the formation of the dihydrate’s basic structural unit (Fig. 1c). These units further form chain structures (Fig. 1d) by connecting with water molecules (N3⁺-H···O6, 2.766 Å; O1···H—O7, 2.922 Å), extending along the b-axis. The chains are linked into layers through weak interactions, which are stacked to form a 3D structure, as shown in Fig. 1e along the b-axis.

  The CIF-DFA methanol solvate crystallizes in the triclinic P 1 space group. Its asymmetric unit includes one CIF cation, one DFA anion, and one methanol molecule. Similar to the dihydrate, in the methanol solvate structure, the N3 protonated piperazine ring (N3⁺-H···O4⁻, 2.680 Å) is stabilized with DFA through charge-assisted hydrogen bonding, and the methanol molecule is connected to DFA via O6-H···O5 (2.700 Å), forming the basic structural unit (Fig. 1f). The chain structure (Fig. 1g) is generated by hydrogen bonding (N3⁺-H···O1, 2.785 Å), with layers formed through weak interactions. These layers stack through further interactions to create a 3D structure (Fig. 1h) along the c-axis. Although a perfect CIF-DFA anhydrate single crystal was not obtained, powder X-ray diffraction pattern (PXRD) was used to calculate its cell parameters, as summarized in Table S1 (Supporting information). It crystallizes in the monoclinic P 21/c space group, with cell parameters similar to those of CIF-DFA dihydrate. PXRD, thermal analysis, Fourier transform infrared spectroscopy (FT-IR), and elemental analysis (EA) confirmed these new phases (Table S2 and Figs. S1-S3 in Supporting information).

  The phase transition relationship among the solid forms of CIF-DFA salts such as anhydrate, dihydrate, and methanol solvate, was examined to better understand the transition mechanisms influenced by water and methanol solvent. Dynamic vapor sorption (DVS) and humidity experiments were conducted to observe these phase changes. Both CIF-DFA anhydrate and methanol solvate displayed minimal moisture absorption (1.2% and 0.36%, respectively) at 90% relative humidity (RH), indicating slight hygroscopicity (Fig. 2a) [48]. CIF-DFA dihydrate, however, showed significant mass change with increasing humidity, reaching 5.1% at 10% RH, which corresponds to absorption of approximately one water molecule per unit (calculated as 4.9%). As humidity reached 90%, CIF-DFA dihydrate absorbed up to 6.5% water, equal to approximately 1.3 water molecules. Based on the PXRD patterns (Fig. 2b), this mass increase is attributed to physical adsorption, which indicates the absorbed water is unbound [49]. Further humidity testing (Table S3 in Supporting information) showed that CIF-DFA anhydrate remained stable between 8% and 45% RH. However, at 84% RH, it transitioned to a mixture of anhydrate and dihydrate, with PXRD analysis highlighting the distinct peaks of each form (marked in pink and black dashed circles, Fig. 2b blue). This indicates that CIF-DFA anhydrate should be stored in dry conditions to avoid phase transition. In contrast, CIF-DFA dihydrate demonstrated strong physical stability, retaining its phase after 10 days at 84% RH without any observable transition. To further explore the phase transition relationships among CIF-DFA anhydrate, dihydrate, and methanol solvate, methanol vapor at 45% RH and 25 ℃ was employed over periods of 0, 5, and 10 days. PXRD patterns indicated that both CIF-DFA anhydrate and dihydrate transformed into methanol solvate by day 5, remaining phase stable until day 10 (Fig. 2c). The presence of methanol, a polar solvent, acted as a "molecular loosener", disrupting the hydrogen bonding network of the anhydrate and dihydrate forms. This disruption facilitated the formation of stronger hydrogen bonds between CIF, DFA, and methanol, promoting the preferential formation of methanol solvates [50].

  Figure 2

  

  Figure 2.  (a) Water sorption (filled points) and desorption (empty points) curves of CIF-DFA anhydrate, dihydrate, and methanol solvate at 25 ℃. (b) PXRD patterns of CIF-DFA anhydrate and dihydrate at different RHs. Pink dashed circles represent anhydrate characteristic peaks, black dashed circles represent dihydrate characteristic peaks. (c) PXRD patterns of phase transitions of CIF-DFA anhydrate and dihydrate in methanol vapor. Black triangles indicate anhydrate characteristic peaks, black squares represent dihydrate characteristic peaks, and black asterisks represent methanol solvate characteristic peaks. (d) PXRD patterns of phase transitions of CIF-DFA anhydrate at different water activities, with a corresponding phase diagram showing the stability of anhydrate as a function of water activity. Blue dotted boxes indicate anhydrate characteristic peaks, red dotted boxes indicate dihydrate characteristic peaks.

  DownLoad: Full-Size Img PowerPoint

  What factors contribute to the phase transition of CIF-DFA salts? Water activity (a_w) is a crucial determinant between the anhydrate and dihydrate forms, while solvent vapor plays a significant role in solvate formation. The hydration state of the crystalline form is influenced by water activity at a given temperature, and the activation barrier for phase transition is notably lower when the anhydrate is exposed to isopropanol/water solutions. Specifically, when the a_w exceeds 0.24, the anhydrate spontaneously and irreversibly transitions to the dihydrate (Fig. 2d), suggesting that the dihydrate is thermodynamically favored at higher a_w levels. Below this threshold, the anhydrate remains stable in solution. This critical water activity threshold serves as an important parameter for driving phase transitions between anhydrate and dihydrate forms. Moreover, these findings imply that manipulating water activity can promote the transition from anhydrate to hydrate, offering a strategy to control phase transitions and improve the stability of hydrated crystalline drug forms. In the DVS experiments, the transition from CIF-DFA anhydrate to CIF-DFA dihydrate was not observed, likely due to the high activation energy barrier associated with this process. Calculations of hydrogen energy suggest that overcoming an energy barrier of at least 4 kcal/mol is necessary to remove one molecule of lattice water, highlighting the challenges in this phase transition.

  To better understand the phase transition, more general solution-mediated transition experiments [51] were conducted to investigate the morphological differences among the three CIF-DFA solid forms. The observed morphologies included micro-crystals for CIF-DFA anhydrate, needle-like crystals for dihydrate, and block crystals for methanol solvate. The transition from anhydrate to dihydrate and methanol solvate occurs in three distinct steps, as illustrated in Fig. 3a. The first step involves the dissolution of the anhydrate in pure water or methanol, resulting in a supersaturated solution relative to either the dihydrate or methanol solvate. This is followed by the nucleation of the dihydrate (Fig. 3a(ⅰ)) or methanol solvate (Fig. 3a(ⅱ)). Finally, stable crystal growth of the dihydrate or methanol solvate occurs (Fig. 3b) [52]. The transition from anhydrate to dihydrate was achieved by placing the anhydrate in pure water. As shown in Fig. S4 (Supporting information), upon selecting an appropriate time to remove the CIF-DFA anhydrate (microcrystals) from the glass vial, it was observed that it transformed into CIF-DFA dihydrate (needle-like crystals) after just 0.1 h in pure water. Over time, the needle-like single crystals continued to grow in size. The number and size of the dihydrate crystals increased progressively, and by 5.5 h, the anhydrate had completely transitioned into dihydrate.

  Figure 3

  

  Figure 3.  (a) Anhydrate to dihydrate/methanol solvate: (ⅰ) Anhydrate dissolution releases CIF and DFA, which recrystallize with water molecules to form dihydrate; (ⅱ) Anhydrate dissolution releases CIF and DFA, which recrystallize with methanol molecules to form methanol solvate. (b) Different forms of the three salts and phase transition processes in dihydrate, methanol solvate: (ⅲ) Dihydrate solubilization and transformation to methanol solvates; (ⅳ) Transformation of methanol solvate to dihydrate.

  DownLoad: Full-Size Img PowerPoint

  To investigate the phase transition of CIF-DFA salt from anhydrate to solvate, methanol/water solutions with varying ratios (v/v) were utilized. In solutions with methanol volume fractions ranging from 10% to 60%, the CIF-DFA anhydrate transformed into dihydrate (Figs. S5a-f in Supporting information). As the methanol concentration increased, particularly at 70%, both needle-like and block crystals were observed, indicating the presence of both dihydrate and methanol solvate (Fig. S5g in Supporting information). In solutions with methanol volume fractions between 80% and 100%, the anhydrate fully converted to methanol solvate within the same timeframe (Figs. S5h-j in Supporting information). Raman spectra (Fig. S6 in Supporting information) confirmed that the transition from anhydrate to methanol solvate occurred through a two-step process. This indicates that the methanol content in the solution serves as the driving force for the nucleation of the methanol solvate [53]. In addition, the conversion experiments were carried out up to day 7 at the same methanol fraction. In particular, in the 70% methanol solution (Fig. S7 in Supporting information), on day 7, the transformation from the coexisting state of dihydrate and methanol solvate was fully converted to methanol solvate.

  The CIF-DFA anhydrate served as the starting material for all the aforementioned solution-mediated phase transitions. If the starting solid form is CIF-DFA dihydrate, the process is simplified to a one-step transition. Notably, the dihydrate did not exhibit any phase transition in 10% to 70% methanol solutions (Figs. S8a-g in Supporting information), maintaining its needle-like crystal morphology. However, when the volume fraction of methanol increased to 80%, the dihydrate completely transformed into methanol solvates, resulting in a full conversion to block crystals (Figs. S8h-j in Supporting information). This transition was confirmed by Raman spectra (Fig. S9 in Supporting information). The methanol/water solutions with methanol volume fractions ranging from 70% to 79% were prepared to explore the conditions where dihydrate and methanol solvate could coexist. As shown in Fig. S10 (Supporting information), both needle-like crystals (dihydrate) and block crystals (methanol solvate) were observed in solutions with methanol volume fractions between 71% and 73%. This observation is further supported by the corresponding Raman spectra in Fig. S11 (Supporting information), confirming that the dihydrate and methanol solvate coexist within this methanol concentration range. We attempted to observe the transition of dihydrate crystals to methanol solvate by placing them in a 100% methanol solution under a polarized light microscope. As shown in Fig. 3b(ⅲ), the needle-like crystals of the dihydrate began to dissolve from the exterior in methanol solution. After 8 min, the needle-like morphology disintegrated, and by 12 min, block crystals had formed around the original morphology. This process indicates that methanol disrupts the hydrogen bonding connections of the water molecules in the dihydrate. As the crystals dissolve, the concentration of APIs in the surrounding solution increases, thereby accelerating the nucleation rate of the methanol solvate [54]. The phase transition of CIF-DFA methanol solvate to dihydrate can be observed in pure water (Fig. 3b(ⅳ)). After 5 h, the surface of the methanol solvate crystal began to exhibit unevenness. By 13 h, dihydrate crystals were visibly nucleating and growing on the surface of the methanol solvate. Over time, the dihydrate crystals continued to expand in size, and ultimately, at 20 h, the methanol solvate crystal completely transformed into dihydrate crystals.

  Further analyses of voids and hydrogen bond energy calculations provided support for the observed phase transitions of CIF-DFA salts. The solvent-accessible void volumes were calculated to be 3.9% for CIF-DFA dihydrate and 7.1% for CIF-DFA methanol solvate (Table S4 in Supporting information for details). As illustrated in Figs. 4a and b, both solid forms are categorized as isolated-site solvates [55], with their voids being closed and the solvent molecules encapsulated within stable cavities that are not readily desolvable. In terms of void volumes, dihydrate has a lower volume than methanol solvate, theoretically rendering it a more stable solid form [56]. The strength of the hydrogen bonds between the solvents and APIs was assessed using established methods [57]. The calculated hydrogen bonding energy required to dissociate the dihydrate was found to be 4.0 kcal/mol (Fig. 4c), while the energy for the methanol solvate was 8.0 kcal/mol (Fig. 4d). This suggests that the water molecule can be more easily detached from the CIF and DFA compared to the methanol, which binds more strongly to these compounds [58-60]. Moreover, considering the phase transition results, it appears that large voids can be tolerated when the hydrogen bonding network is sufficiently robust [61]. This indicates that the methanol solvate represents a more stable state when compared to both the anhydrate and dihydrate forms.

  Figure 4

  

  Figure 4.  (a) Void map of dihydrate after removal of water. (b) Void map of methanol solvate after removal of methanol. (c) Hydrogen bond energy of dihydrate. (d) Hydrogen bond energy of methanol solvate.

  DownLoad: Full-Size Img PowerPoint

  In this study, after the anhydrate transformed into the dihydrate, the dihydrate showed greater stability in RH experiments. When the anhydrate formed a methanol solvate, the solvate had lower hygroscopicity and better moisture resistance compared to the anhydrate in DVS. It should be noted that unlike single-component phase transitions, the multi-component phase transition of CIF-DFA depends not only on external factors but also on the ratio of each component, intermolecular proton transfer, hydrogen bonding, and other interactions. In conclusion, phase transitions in multi-component salts are a common and significant phenomenon affecting clinical drug safety. Water activity and solvent vapor play a critical role in these transitions. Understanding these factors helps optimize manufacturing, ensures drug stability during storage and distribution, and reduces clinical risks. Therefore, choosing thermodynamically stable crystalline forms is crucial to prevent dehydration and desolvation during storage and production.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Shifang Song: Writing – original draft, Data curation. Chenyu Wu: Writing – original draft, Data curation. Li Zhang: Data curation. Dezhi Yang: Writing – review & editing, Data curation. Yang Lu: Conceptualization. Zhengzheng Zhou: Writing – review & editing, Supervision, Investigation.

  Acknowledgments

  This work was supported by the Young Scientists Promotion Fund of Natural Science Foundation of Guangdong Province (No. 2023A1515030128), Natural Science Foundation of Guangdong Province (No. 2024A1515011590), National Natural Science Foundation of China (No. 81703438) and CAMS Innovation Fund for Medical Sciences (No. 2022-I2M-1-015).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110911.

  
  
• 1. [1]

     D. Singhal, Adv. Drug Deliv. Rev. 56 (2004) 335–347.

  2. [2]

     Z. Zhou, M. Calatayud, J. Contreras-Garcia, et al., J. Pharm. Sci. 108 (2019) 3340–3347.

  3. [3]

     H. Wang, D, Yang, W. Zhang, et al., Chin. Chem. Lett. 34 (2023) 107258. doi: 10.1016/j.cclet.2022.02.063

  4. [4]

     H. Liu, H.S. Chan, L. Zhang, et al., Chin. Chem. Lett. 34 (2023) 108057. doi: 10.1016/j.cclet.2022.108057

  5. [5]

     L. Lange, G. Sadowski, Cryst. Growth Des. 16 (2016) 4439–4449. doi: 10.1021/acs.cgd.6b00554

  6. [6]

     K.K. Arora, S. Thakral, R. Suryanarayanan, Pharm. Res. 30 (2013) 1779–1789. doi: 10.1007/s11095-013-1022-7

  7. [7]

     I. Sathisaran, S.V. Dalvi, Cryst. Growth Des. 21 (2021) 2076–2092. doi: 10.1021/acs.cgd.0c01513

  8. [8]

     Z. Zhou, H.H. Tong, L. Li, et al., Cryst. Res. Technol. 50 (2015) 538–548. doi: 10.1002/crat.201500034

  9. [9]

     Z.Y. Fang, B.X. Zhang, W.H. Xing, Chin. Chem. Lett. 33 (2022) 2159–2164. doi: 10.1016/j.cclet.2021.10.056

  10. [10]

      A. Joseph, C.E. Bernardes, A.I. Druzhinina, et al., Cryst. Growth Des. 17 (2017) 1918–1932. doi: 10.1021/acs.cgd.6b01876

  11. [11]

      L. Ma, Q. Zheng, D.K. Unruh, Chem. Commun. 10 (2023) 3640–3654.

  12. [12]

      J.H. Horst, P.W. Cains, Cryst. Growth Des. 8 (2008) 2537–2542. doi: 10.1021/cg800183v

  13. [13]

      K.M. Lutker, R. Quiñones, J. Xu, A. Ramamoorthy, A.J. Matzger, J. Pharm. Sci. 100 (2011) 949–963. doi: 10.1002/jps.22336

  14. [14]

      S. Karki, T. Friščić, W. Jones, W.S. Motherwell, Mol. Pharmaceut. 4 (2007) 347–354. doi: 10.1021/mp0700054

  15. [15]

      P. Sanphui, S.S. Kumar, A. Nangia, Cryst. Growth Des. 12 (2012) 4588–4599. doi: 10.1021/cg300784v

  16. [16]

      Z. Li, J. Zhou, X. Zhang, S. Wu, J. Gong, Chin. Chem. Lett. 34 (2023) 107983. doi: 10.1016/j.cclet.2022.107983

  17. [17]

      P. Yang, C. Qin, S. Du, et al., Crystals 9 (2019) 367. doi: 10.3390/cryst9070367

  18. [18]

      D. Rossi, T. Gelbrich, V. Kahlenberg, U.J. Griesser, CrystEngComm 14 (2012) 2494–2506. doi: 10.1039/c2ce06659a

  19. [19]

      S. Aitipamula, P.S. Chow, R.B. Tan, CrystEngComm 14 (2012) 691–699.

  20. [20]

      S. Majodina, L. Ndima, O.O. Abosede, et al., CrystEngComm 23 (2021) 335–352. doi: 10.1039/d0ce01387k

  21. [21]

      X.B. Zhou, J.R. Zhu, J.M. Gu, X.R. Hu, Acta Crystallogr., Sect. C: Struct. Chem. 74 (2018) 690–695. doi: 10.1107/s2053229618006836

  22. [22]

      J. Yang, C.T. Hu, A.G. Shtukenberg, et al., CrystEngComm 20 (2018) 1383–1389. doi: 10.1039/c7ce02107k

  23. [23]

      H. Liu, J. Nie, H.S. Chan, et al., Int. J. Pharmacuet. 601 (2021) 120537.

  24. [24]

      Z. Li, H. Qu, L. Fang, S. Wu, J. Gong, Chin. Chem. Lett. 34 (2023) 107504. doi: 10.1016/j.cclet.2022.05.018

  25. [25]

      S.K. Rai, S. Allu, A.K. Nangia, Cryst. Growth Des. 20 (2020) 4512–4522. doi: 10.1021/acs.cgd.0c00313

  26. [26]

      N. Qiao, K. Wang, W. Schlindwein, A. Davies, M. Li, Eur. J. Pharm. Biopharm. 83 (2013) 415–426.

  27. [27]

      M. Omori, T. Uekusa, J. Oki, D. Inoue, K. Sugano, J. Drug Deliv. Sci. Technol. 56 (2020) 101566.

  28. [28]

      D. Huang, D. Yang, Y. Lv, et al., J. Mol. Struct. 1250 (2022) 131891.

  29. [29]

      Y. Li, P.S. Chow, R.B. Tan, Int. J. Pharmaceut. 415 (2011) 110–118.

  30. [30]

      J. Li, Y. Huang, Q. An, et al., Int. J. Pharmaceut. 628 (2022) 122310.

  31. [31]

      M.M.H. Smets, S.J.T. Brugman, E.R.H. Van Eck, et al., Cryst. Growth Des. 15 (2015) 5157–5167. doi: 10.1021/acs.cgd.5b01188

  32. [32]

      M.C. Meyer, A.B. Straughn, E.J. Jarvi, et al., Pharm. Res. 9 (1992) 1612–1616.

  33. [33]

      D. Murphy, F. Rodrıguez-Cintrón, B. Langevin, et al., Int. J. Pharmaceut. 246 (2002) 121–134.

  34. [34]

      J. Bauer, S. Spanton, R. Henry, et al., Pharm. Res. 18 (2001) 859–866.

  35. [35]

      N. Zencirci, T. Gelbrich, D.C. Apperley, et al., Cryst. Growth Des. 10 (2010) 302–313. doi: 10.1021/cg901062n

  36. [36]

      Y. Liu, A. Li, S. Xu, et al., Angew. Chem. Int. Ed. 59 (2020) 15098–15103. doi: 10.1002/anie.202002220

  37. [37]

      F. Van Bambeke, J.M. Michot, J. Van Eldere, P.M. Tulkens, Clin. Microbiol. Infect. 11 (2005) 256–280. doi: 10.1111/j.1469-0691.2005.01131.x

  38. [38]

      G. Liu, J. Liu, Y. Liu, et al., J. Am. Chem. Soc. 136 (2014) 590–593. doi: 10.1021/ja4102634

  39. [39]

      A. Karagianni, M. Malamatari, K. Kachrimanis, Pharmaceutics 10 (2018) 18. doi: 10.3390/pharmaceutics10010018

  40. [40]

      H. Liu, H.H.Y. Tong, Z. Zhou, J. Therm. Anal. Calorim. 147 (2022) 12947–12963. doi: 10.1007/s10973-022-11762-1

  41. [41]

      C.B. Romañuk, Y.G. Linck, A.K. Chattah, et al., Int. J. Pharmaceut. 391 (2010) 197–202.

  42. [42]

      D.M. Campoli-Richards, J.P. Monk, A. Price, et al., Drugs 35 (1988) 373–447. doi: 10.2165/00003495-198835040-00003

  43. [43]

      T.J. Gan, Curr. Med. Res. Opin. 26 (2010) 1715–1731. doi: 10.1185/03007995.2010.486301

  44. [44]

      G.G. Zhang, D. Law, E.A. Schmitt, et al., Adv. Drug Deliv. Rev. 56 (2004) 371–390.

  45. [45]

      H. Wang, L. Wang, G. Xie, et al., Cryst. Growth Des. 24 (2024) 3399–3409. doi: 10.1021/acs.cgd.4c00120

  46. [46]

      A.O. Surov, A.P. Voronin, A.N. Manin, et al., Mol. Pharmaceut. 11 (2014) 3707–3715. doi: 10.1021/mp5004652

  47. [47]

      T.S. Thakur, R. Thakuria, Cryst. Growth Des. 20 (2020) 6245–6265. doi: 10.1021/acs.cgd.0c00654

  48. [48]

      L. Yuan, E. Horosanskaia, F. Engelhardt, et al., Cryst. Growth Des. 19 (2019) 854–867. doi: 10.1021/acs.cgd.8b01425

  49. [49]

      S. Rajabnezhad, T. Ghafourian, A. Rajabi-Siahboomi, et al., Carbohydr. Polym. 238 (2020) 116208.

  50. [50]

      K. Greco, R. Bogner, J. Pharm. Sci. 101 (2012) 2996–3018. doi: 10.1002/jps.23025

  51. [51]

      H. Wikström, J. Rantanen, A.D. Gift, L.S. Taylor, Cryst. Growth Des. 8 (2008) 2684–2693. doi: 10.1021/cg070629e

  52. [52]

      R.J. Davey, N. Blagden, S. Righini, H. Alison, E.S. Ferrari, J. Phys. Chem. B 106 (2002) 1954–1959.

  53. [53]

      E. Arunan, G.R. Desiraju, R.A. Klein, et al., Pure Appl. Chem. 83 (2011) 1637–1641. doi: 10.1351/pac-rec-10-01-02

  54. [54]

      J. Weng, Y. Huang, D. Hao, Y. Ji, Chin. J. Chem. Eng. 28 (2020) 935–948.

  55. [55]

      L. Jia, M. Guo, J. Dai, et al., Cryst. Growth Des. 23 (2023) 832–841. doi: 10.1021/acs.cgd.2c01071

  56. [56]

      A. Berzins, D. Zvanina, A. Trimdale, Cryst. Growth Des. 18 (2018) 2040–2045. doi: 10.1021/acs.cgd.7b01457

  57. [57]

      S. Emamian, T. Lu, H. Kruse, H. Emamian, J. Comput. Chem. 40 (2019) 2868–2881. doi: 10.1002/jcc.26068

  58. [58]

      D. Yang, H. Wang, Q. Liu, et al., Chin. Chem. Lett. 33 (2022) 3207–3211. doi: 10.1016/j.cclet.2021.10.012

  59. [59]

      F. Li, Z. Zheng, S. Xia, L. Yu, J. Mol. Struct. 1219 (2020) 128480.

  60. [60]

      A.J. Aquino, D. Tunega, G. Haberhauer, M.H. Gerzabek, H. Lischka, J. Phys. Chem. A 106 (2002) 1862–1871. doi: 10.1021/jp013677x

  61. [61]

      V.A. Russell, M.D. Ward, Chem. Mater. 8 (1996) 1654–1666.

• 

  Figure 1  Chemical structures of (a) CIF and (b) DFA. CIF-DFA dihydrate: (c) Hydrogen bonding network, (d) layered structures, and (e) packing diagrams. CIF-DFA methanol solvate: (f) hydrogen bonding network, (g) layered structures, and (h) packing diagrams.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Water sorption (filled points) and desorption (empty points) curves of CIF-DFA anhydrate, dihydrate, and methanol solvate at 25 ℃. (b) PXRD patterns of CIF-DFA anhydrate and dihydrate at different RHs. Pink dashed circles represent anhydrate characteristic peaks, black dashed circles represent dihydrate characteristic peaks. (c) PXRD patterns of phase transitions of CIF-DFA anhydrate and dihydrate in methanol vapor. Black triangles indicate anhydrate characteristic peaks, black squares represent dihydrate characteristic peaks, and black asterisks represent methanol solvate characteristic peaks. (d) PXRD patterns of phase transitions of CIF-DFA anhydrate at different water activities, with a corresponding phase diagram showing the stability of anhydrate as a function of water activity. Blue dotted boxes indicate anhydrate characteristic peaks, red dotted boxes indicate dihydrate characteristic peaks.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Anhydrate to dihydrate/methanol solvate: (ⅰ) Anhydrate dissolution releases CIF and DFA, which recrystallize with water molecules to form dihydrate; (ⅱ) Anhydrate dissolution releases CIF and DFA, which recrystallize with methanol molecules to form methanol solvate. (b) Different forms of the three salts and phase transition processes in dihydrate, methanol solvate: (ⅲ) Dihydrate solubilization and transformation to methanol solvates; (ⅳ) Transformation of methanol solvate to dihydrate.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Void map of dihydrate after removal of water. (b) Void map of methanol solvate after removal of methanol. (c) Hydrogen bond energy of dihydrate. (d) Hydrogen bond energy of methanol solvate.

  下载: 全尺寸图片 幻灯片
•