New data on the crystal-chemistry of arrojadite: an HT study

Arrojadite group minerals are very complex orthophosphates, with monoclinic symmetry. Ideally arrojadite has formula A2B2Ca1Na2+xM13Al(PO4)11(PO3OH1-x)W2, where A sites are occupied by Ba, Sr, Pb, Na, K plus vacancy; the B site is occupied by divalent Fe, Mn, Mg, (Zn, Li) plus vacancy; the W site can be occupied by OH or F [1]. Their classification was approved by IMA-CNMMN and published by [1]. Their structure was described as C2/c by [2] and [3], and later revised by [4] who showed that the correct space group is actually Cc on the basis of new single-crystal X-ray diffraction data (SCXRD) and Raman spectroscopy. Recently the Cc model has been questioned [5] although very recent spectroscopic evidences confirm the lower singony [6]. The collaboration with AMI led us to examine new samples of arrojadite with compositions never described so far. In particular, we found the occurrence of a specimen with Mg dominant at the M group cation site coming from the locally known "Costa Balzi Rossi" locality, close to the village of Isallo (Liguria, Italy). This sample has a 16.3944(4), b 9.9477(2), c 24.4435(7) Å, β 105.661(3)°, and V 3838.41(16) Å3 and SCRDX data confirms that Mg is dominant in most of the M sites. Another specimen from Hagendorf (Germany) revealed to be F and (Na,Ba) dominant at W and A sites respectively. The relative proposals for new mineral names are being prepared to be submitted soon to IMA-CNMNC. Both samples show structural models in Cc s.g. which have the presence of residual density reproducing the C2/c model. This has been observed for other specimens, among them the arrojadite-(NaFe) from Nickel Plate (USA) studied also by [3] and [5]. Due to the controversy related to the choice of the C2/c versus Cc space groups, we performed annealing experiment of an arrojadite-(KFe) sample from Rapid Creek, Yukon (Canada). The natural crystal has Cc symmetry without any residue of density in the Fourier difference maps. A single crystal (0.080 0.30 0.35mm radius) was placed in a sealed and vacuumed quartz vial together with a Pt crucible containing iron-wüstite buffer and annealed at 650 °C for 24 hours. The vial was dropped into water and the crystal was studied by SCXRD at the same conditions used for the natural crystal. The results of structure refinement in the Cc space group showed the presence of disorder in a similar fashion to that found for natural samples coming from Nickel Plate. Therefore, the annealing clearly promotes the disorder of cations leading to an apparent centre of symmetry. Acknowledgments: M.A. has been supported by ERC grant #307322 to F.N. [1] Chopin C., Oberti R. and Cámara F. (2006). Am. Mineral., 91, 1260-1270. [2] Krutik V.M., Pushcharovskii D.Y., Pobedimskaya E.A., Belov N.V. (1979). Kristallografiya, 24, 743-750. [3] Merlino S., Mellini M., Zanazzi P.F. (1981). Acta Crystallogr. B37, 1733-1736. [4] Cámara F., Oberti R., Chopin C., Medenbach O. (2006). Am. Mineral. 91, 1249-1259. [5] Kallfaß C., Hocha C., Schiera H., Simon A., Schubert H. (2010). Z. Naturforsch., 65b, 1427-1433. [6] Frost R.L., Xi Y., Scholz R., Frota Campos Horta L. (2013). Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 109, 138-145.