In order to clarify the symmetry problem along the stannite -kësterite join [Cu 2 FeSnS 4 -Cu 2 ZnSnS 4 ], a structural study of synthetic Cu 2 Fe 1-x Zn x SnS 4 single crystals was performed (x = 0, 0.2, 0.5, 0.7, 0.8 and 1, respectively). The metal distribution among the tetrahedral cavities was determined by refining different models in both the I4 and I42m space groups. The best agreement was obtained in I42m, even for the Zn-rich members of the series. However, two different mechanisms of incorporation take place along the stannite-kësterite join. For pure stannite and zincian stannite (x = 0, 0.2, 0.5), the 2a position (0,0,0) is mainly occupied by (Fe,Zn), whereas Cu is the dominant species at 4d (0,½,¼). For ferroan kësterite and pure kësterite (x = 0.7, 0.8, 1), the 2a position is fully occupied by Cu, whereas (Zn,Fe) and the remaining Cu are disordered at 4d. On the basis of the structural results, pure Me-S bond-distances are proposed for Fe, Cu, Zn in both 2a and 4d sites, and the metal distribution among the tetrahedral sites is obtained accordingly. For x ≥ 0.7, the Me-S distance found for the atom located at 2a closely approaches that found for the atom located at 4d, thus producing a more regular framework. Accordingly, distortion parameters and 2 of the S(Me 3 Sn) tetrahedron decrease with increasing Zn. This feature, in turn, is the reason for the pseudocubic symmetry of the lattice observed in the Zn-rich region (2a close to the c parameter). The unit-cell volume linearly increases with increasing Zn, thus confirming the mainly covalent character of the bonds in these compounds. The previously noted inversion of slope in the unit-cell parameters at x = 0.7 corresponds to the point of the series wherein Cu becomes predominant at the 2a site. The proposed model accounts for the structural and geometrical variations observed along the stannite-kësterite series, even if no change of space group is assumed.
