Purpose: Approximately one-third of organic materials show crystalline polymorphism, with a further third capable of forming hydrates or solvates . The exact polymorphic state of an active pharmaceutical ingredient may influence several physicochemical properties. In particular, the hydrate/solvate state of crystalline substances is of particular concern as different hydrate/solvate forms can affect the material solubility, dissolution rate, flowability, and compressibility.
Methods: Carbamazepine (CBZ) is an anti-convulsant used in the treatment of epilepsy has often been used as a model material when studying polymorphs. It is known to form a dihyrate  and a 1:1 solvate with acetone . Amorphous CBZ was prepared by soaking crystalline CBZ in water overnight. The resulting slurry was then dried at 0% relative humidity. This has been previously proven to produce 100% amorphous carbamazepine . The hydration/solvation and dehydration/desolvation transitions are affected by both temperature and solvent concentration. These can often be investigated using vapor sorption techniques. Dynamic Vapor Sorption (DVS) was used to measure the water vapor and acetone vapor sorption isotherms of CBZ at 25 °C. This was then compared to a competitive sorption experiment where both vapors are present. In-situ Raman spectroscopy was used to confirm the ultimate hydrate/solvate form and transition points.
Results: Figure 1 displays water vapor isotherm (a.) and in-situ Raman spectra (b.) results for CBZ at 25 °C. At 80% RH and above, there is clear evidence of stoichiometric hydrate formation. This is also supported by the wide hysteresis gap between sorption and desorption isotherms. The corresponding Raman results show little shift in the spectra due to hydrate formation. For acetone vapor sorption, the isotherm (Figure 1b) indicates CBA forms a 1:1 solvate upon exposure to 10% P/Po. It remains a solvate on the desorption phase until it is fully dried. The Raman results (Figure 2b) show more pronounced shifts in the spectra. In particular, between 2900 and 3100 cm-1, there are unique features present for the solvate form that are not present in the anhydrous form.
In a co-adsorption experiments, the sample was first exposed to acetone vapor at 10% P/Po, where it would form a solvate (see Figure 1b). Then, the sample was exposed to humidity steps from 0 to 90% RH. Upon exposure to 40% RH, there was a significant mass loss as water vapor appears to ‘kick out’ the acetone solvate. By 90% RH, the system forms a hydrate. These results indicate that the hydrate is thermodynamically favored.
Conclusion: Whether formed from the liquid or vapor phase, both solvation-desolvation processes are thermodynamically equivalent. Therefore, gravimetric vapor sorption studies can be a powerful tool in characterizing solvates over a wide range of solvent concentrations and environmental temperatures. CBZ forms both a hydrate and acetone solvate when exposed to the corresponding vapor. Co-adsorption studies indicate that the hydrate is more thermodynamically favored. The combination of vapor sorption and Raman spectroscopy can provide additional information in determining vapor-induced hydrate or solvate formation.
 J.O. Henck, U.J. Griesser and A. Burger, Pharm. Ind., 59 (1997) 165-169.
 L.E. McMahon, P. Timmins, A.C. Williams and P. York, J. Pharm. Sci., 85 (1996) 1064-1069.
 C.F. Terrence, M. Sax, G.H. From, C.H. Chang and C.S. Yoo, Pharmacol., 27 (1983) 85-94.
 R. Surana, A. Pyne and R. Suryanarayanan, AAPS PharmSciTech, 4 (2003) Article 68.
Majid Naderi– Surface Measurement Systems, Ltd.
Damiano Cattaneo– Surface Measurement Systems, Ltd.
Jerry Heng– Reader, Imperial College London, London, England
Frank Thielmann– Novartis Pharma AG
Srinivasulu Aitipamula– Institute of Chemical and Engineering Sciences
Jin Wang Kwek– Institute of Chemical and Engineering Sciences