The crystal structure of a clinopyroxene with composition (Na0.75Ca0.25)(Cr0.75Mg0.25)Si2O6 was refined at 100, 150, 200, 250 and 298 K. The work was performed in the context of an investigation on the low-temperature behaviour of A+M3+Si2O6 (with A dominant in Na and M = transition elements) pyroxenes in order to provide new insights concerning the phase transition and anomalies recently found in the low-temperature behaviour of NaTiSi2O6 and NaGaSi2O6 compounds. The refinements were done in the C2/c space group (wR2 between 0.048 and 0.068), and no change of symmetry was observed down to 100 K. Highly-anisotropic axial thermal expansion occurs with the scheme αb ≥ αb > αc.

The M2, M1 and T polyhedra expand with αM2 > αM1 αT as generally observed in pyroxenes. A discontinuity in the M1 polyhedral volume is observed between 200 and 250 K, similar to the one observed in NaGaSi2O6 between 190 and 235 K.

The atomic displacement parameters are scaled according to the following pattern: UM2 > UO2 > UO3 ≥ U01 > UT ≥ UM1 Comparison with previous data along the CaMgSi2O-NaCrSi2O6 join suggests significant positional disorder for the O1 oxygen, due to repulsion of the 2p orbitals of O1 and the non-bondine 3d electrons of Cr.

The exact nature of tripuhyite remains controversial more than 100 years after the first description of the mineral. Different stoichiometries and crystal structures (rutile or tri-rutile types) have been suggested for this Fe-Sb-oxide. To address these uncertainties, we studied tripuhyite from Tripuhy, Minas Gerais, Brazil (type material) and Falotta, Grisons, Switzerland using single-crystal and powder X-ray diffraction (XRD), optical microscopy and electron microprobe analysis.

Electron microprobe analyses showed the Fe/Sb ratios to be close to one in tripuhyite from both localities. Single crystal XRD studies revealed that tripuhyite from the type locality and from Falotta have the rutile structure (P42mnm, a = 4.625(4) c = 3.059(5) and a = 4.6433(10) c= 3.0815(9) Å, respectively). Despite careful examination, no evidence for a tripled c parameter, characteristic of the tri-rutile structure, was found and hence the structure was refined with the rutile model and complete Fe-Sb disorder over the cationic sites in both cases (type material: R1 = 3.61%; Falotta material: R1 = 3.96%). The specular reflectance values of type material tripuhyite and lewisite were measured and the following refractive indices calculated (after Koenigsberger): tripuhyite nmin = 2.14, nmax = 2.27; lewisite (cubic) n = 2.04.

These results, together with those of 57Fe and 121Sb Mössbauer spectroscopy on natural and synthetic tripuhyites reported in the literature, indicate that the chemical formula of tripuhyite is Fe3+Sb5+O4 (FeSbO4). Thus, tripuhyite can no longer be attributed to the tapiolite group of minerals of general type AB2O6. A comparison of the results presented with the mineralogical data of squawcreekite suggests that tripuhyite and squawcreekite are identical. In consequence, tripuhyite was redefined as Fe3+Sb5+O4 with a rutile-type structure. Both the proposed new formula and unit cell (rutile-type) of tripuhyite as well as the discreditation of squawcreekite have been approved by the Commission on New Mineral and Mineral Names (CNMMN) of the International Mineralogical Association (IMA).