The monoclinic I2 structure of bassanite, calcium sulphate hemihydrate (CaSO 4 · 0.5H 2 O) P AOLO BALLIRANO 1 ,ADRIANA MARAS 1 ,SIMONE MELONI 2 and RUGGERO CAMINITI 2 1 Dipartimento di Scienze della Terra, Universit` a di Roma “La Sapienza”, P.le A. Moro 5, I-00185 Roma, Italy e-mail: paolo.ballirano@uniroma1.it 2 Dipartimento di Chimica, Istituto Nazionale per la Fisica della Materia, Universit` a di Roma “La Sapienza”, P.le A. Moro 5, I-00185 Roma, Italy Abstract: A structural analysis of CaSO 4 · 0.5H 2 O, a dehydration product of gypsum, has been carried out through the Rietveld method on X-ray powder diffraction data. A dehydrated powder of synthetic gypsum has been charged inside a non-hermetically sealed capillary in order to allow a slow rehydration. The starting material has been identified as -anhydrite, space group P6 2 22, cell parameters a = 6.9691(2) Å, c = 6.3033(2) Å. The final product of the rehydration of -anhydrite is CaSO 4 · 0.5H 2 O, space group I2 (unique axis b), cell parameters a = 12.0350(5) Å, b = 6.9294(3) Å, c = 12.6705(4) Å, = 90.266(3)°. The structure of the hemihydrate is strongly pseudo-trigonal, space group P3 1 21. The symmetry lowering arises from water molecules ordering inside the channels. Key-words: -anhydrite, CaSO 4 · 0.5H 2 O, bassanite, gypsum, X-ray powder diffraction, Rietveld method. Introduction The dehydration process of gypsum, CaSO 4 · 2H 2 O, has been investigated for many years. Thermogra- vimetric and IR spectroscopy studies (Putnis et al., 1990) indicated an apparently continuous water- loss mechanism characterised by an activation en- ergy of 90.3 kJ mol -1 . According to these authors the reaction proceeds from gypsum to hemihydrate (CaSO 4 · 0.5H 2 O, corresponding to the natural phase bassanite) to -CaSO 4 (soluble anhydrite) without the occurrence of any further intermediate- water-content form. These data were consistent with the opinion of Ball (1977) that the apparent deviations from stoichiometric bassanite were caused by physical sorption of water. The absence of subhydrates different from the hemihydrate was also indicated, from proton magnetic resonance da- ta, by Saito (1961). In contrast, other authors (e.g. Frik & Kuzel, 1982; Bushuev & Borisov, 1982; Abriel, 1983; Kuzel, 1987; Kuzel & Hauner, 1987; Abriel et al., 1988) have synthesised and described different forms of subhydrates (CaSO 4 · nH 2 O, 0.48 n 0.81) whose occurrence has been attributed to different condition of synthesis and in particular to the H 2 O partial pressure. The structural investigations of subhydrates and -anhydrite proceeded almost at the same rate as the similarities among the various structures were rapidly hypothesised and proved. For an early collection of relevant references see Deer et al. (1962). The first structural model for the hemihydrate was proposed by Gallitelli (1933). From Weissen- berg photographs he determined cell parameters (a = 11.94 Å, b = 6.83 Å, c = 12.70 Å, ca. 90°) and space group (C2) demonstrating that the monoclin- ic, pseudo-trigonal structure was characterised by channels containing the water molecules. Since then various subhydrates have been described Eur. J. Mineral. 2001, 13, 985–993 DOI: 10.1127/0935-1221/2001/0013/0985 0935-1221/01/0013/0985 $ 2.25 2001 E. Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart