Synthesis, crystal structure and charge-distribution validation of a new alluaudite-type phosphate, Na2.22Mn0.87In1.68(PO4)3

The title compound crystallizes in the alluaudite structure type. Its three-dimensional framework includes channels in which partially occupied sodium cations are situated.

Na 2.22 Mn 0.87 In 1.68 (PO 4)3 , sodium manganese indium tris(phosphate) (2.22/0.87/ 1.68), was obtained in the form of single crystals by a flux method and was structurally characterized by single-crystal X-ray diffraction. The compound belongs to the alluaudite structure type (space group C2/c) with general formula X(2)X(1)M(1)M(2) 2 (PO 4 ) 3 . The X(2) and X(1) sites are partially occupied by sodium [occupancy 0.7676 (17) and 1/2] while the M(1) and M(2) sites are fully occupied within a mixed distribution of sodium/manganese(II) and manganese(II)/indium, respectively. The three-dimensional anionic framework is built up on the basis of M(2) 2 O 10 dimers that share opposite edges with M(1)O 6 octahedra, thus forming infinite chains extending parallel to [101]. The linkage between these chains is ensured by PO 4 tetrahedra through common vertices. The three-dimensional network thus constructed delimits two types of hexagonal channels, resulting from the catenation of M(2) 2 O 10 dimers, M(1)O 6 octahedra and PO 4 tetrahedra through edge-and corner-sharing. The channels are occupied by Na + cations with coordination numbers of seven and eight.

Structural commentary
The principal building units ( Fig. 1) of the three-dimensional framework structure of Na 2.22 Mn 0.87 In 1.68 (PO 4 ) 3 are mixedoccupancy (Mn, Na) [= M(1); site symmetry 2] and (Mn1, In) [= M(2)] sites with distorted octahedral environments and two phosphate tetrahedra (P1 and P2); the two sites associated with Na + cations (Na1; Na2 with site symmetry 2) are partially occupied and are situated in the resulting voids. By edgesharing, the (Mn,In)O 6 octahedra form (Mn,In) 2 O 10 dimers, which are linked by highly distorted (Mn,Na)O 6 octahedra into infinite zigzag chains along [101] (Fig. 2). The connection of these chains through vertices belonging to P1O 4 and P2O 4 tetrahedra gives layers perpendicular to [010] (Fig. 3), which, in turn, are linked into the three-dimensional framework by sharing corners with phosphate tetrahedra. This framework accommodates two types of channels extending parallel to [001] in which the Na + cations are located (Fig. 4).
The mean <M1-O> distance of 2.329 Å is between those of 2.23 and 2.42 Å predicted by the sums of the ionic radii (Shannon, 1976) for Mn 2+ and Na + cations in an octahedral environment. The mean <M2-O> distance of 2.150 Å is between the mean distance of 2.142 Å observed for In 3+ in an octahedral environment in NaCuIn(PO 4 ) 2 (Benhsina et al., 2020) and 2.238 Å for Mn 2+ in the same coordination in K 0.53 Mn 2.37 Fe 1.24 (PO 4 ) 3 (Hidouri & Ben Amara, 2011). The PO 4 tetrahedra show a slight distortion, as indicated by the range of P-O bond lengths [1.538 (2)-1.550 (2) Å for P1O 4 and 1.520 (3)-1.566 (2) Å for P2O 4 ], with mean bond lengths of <P1-O> = 1.544 (2) Å and <P2-O> = 1.546 (2) Å , consistent with 1.537 Å as calculated by Baur (1974) for the orthophosphate group. The coordination spheres of the two crystallographically distinct Na sites ( Fig. 1) in the channels were defined under the assumption of a maximum Na-O distance L max = 3.13 Å , suggested by Donnay & Allmann (1970). The environment around Na1 consists of seven O atoms with distances varying from 2.35 (3)     The principal building units of the alluaudite-type phosphate Na 2.22 Mn 0.87 In 1.68 (PO 4 ) 3 with displacement ellipsoids drawn at the 50% probability level. [Symmetry codes: The refined structure model is confirmed by (i) the bondvalence method (Brown & Altermatt, 1985;Brown, 2002) and (ii) the charge-distribution (Chardi) method (Nespolo, 2015(Nespolo, , 2016. The Chardi method is a development of Pauling's concept of bond strength (Pauling, 1929). Instead of the empirical parameters used in the bond-valence approach, it exploits the experimental bond lengths deduced from the structural study to compute a non-integer coordination number (effective coordination number = ECoN) around a PC atom (atom placed at the center of a polyhedron, q > 0), which is coordinated by V atoms (atoms located at the vertices; q < 0); q is the formal oxidation number. ECoN takes into account not only the number of V atoms around a given PC atom, but also their weight in terms of relative distances. Calculated charges Q(i) and valences V(i) are in good agreement with the formal oxidation number (q) multiplied by occupancy rates. The dispersion factor MAPD, MAPD ¼ 100 , which measures the mean absolute percentage deviation, is 1% for the calculated cationic charges. The variation of the ECoN value with respect to the traditional coordination number indicates the degree of distortion. The results of the two validation models are compiled in Table 1.

Synthesis and crystallization
Commercially available NaNO 3 , Mn(NO 3 ) 2 Á6H 2 O, In 2 O 3 , MoO 3 and (NH 4 ) 2 HPO 4 were mixed in stoichiometric ratios of 2:1:1:1:2 and dissolved in aqueous nitric acid. The resulting solution was then evaporated by heating at 353 K. The obtained dry residue was ground in an agate mortar, and then heated increasingly in an open platinum crucible up to 873 K. The sample was then reground and mixed with sodium dimolybdate Na 2 Mo 2 O 7 in the molar ratio P:Mo = 2:1. The mixture was heated for 1 h at 1243 K to give a melt that was subsequently cooled down to room temperature at a rate of 10 K h À1 . Brown hexagonally shaped crystals were obtained by washing the final product with hot water in order to dissolve the flux.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 2 (17), and free refinement of the occupancy of Na1 resulted in a value very close to 0.5. For the final refinement, this value was fixed at 0.5, and all other occupancies were refined to ensure electrical neutrality of the compound. The remaining maximum and minimum electron densities are located 0.74 Å from P2 and 1.07 Å from O24, respectively. Computer programs: KappaCCD Server Software (Nonius, 1997), HKL SCALEPACK and DENZO (Otwinovski & Minor, 1997), SIR92 (Altomare et al., 1993), SHELXL2018/ 3 (Sheldrick, 2015), ORTEP-3 for Windows and WinGX (Farrugia, 2012) and DIAMOND (Brandenburg, 1999).

Computing details
Data collection: KappaCCD Server Software (Nonius, 1997); cell refinement: HKL SCALEPACK (Otwinovski & Minor, 1997); data reduction: HKL DENZO (Otwinovski & Minor, 1997); program(s) used to solve structure: SIR92 (Altomare et al., 1993); program(s) used to refine structure: SHELXL2018/3 (Sheldrick, 2015); molecular graphics: ORTEP-3 for Windows (Farrugia, 2012) and DIAMOND (Brandenburg, 1999); software used to prepare material for publication: WinGX (Farrugia, 2012). Special details Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.
Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )