Crystal structure of a polymorph of μ-oxido-bis[(5,10,15,20-tetraphenylporphyrinato)iron(III)]

In the crystal structure of the new polymorph of of μ-oxido-bis[(5,10,15,20-tetraphenylporphyrinato)iron(III)], two FeIII tetraphenylporphyrin units are linked by μ2-oxido O atoms into dimers, leading to a square-pyramidal coordination for each of the cations.

The title compound, [Fe 2 (C 44 H 28 N 4 O) 2 O], was obtained as a by-product during the synthesis of Fe III tetraphenylporphyrin perchlorate. It crystallizes as a new polymorphic modification in addition to the orthorhombic form previously reported [Hoffman et al. (1972). J. Am. Chem. Soc. 94, 3620-3626; Swepston & Ibers (1985) Acta Cryst. C41, 671-673; Kooijmann et al. (2007). Private Communication (refcode 667666). CCDC, Cambridge, England]. In its crystal structure, the two crystallographically independent Fe III cations are coordinated in a square-planar environment by the four N atoms of a tetraphenylporphyrin ligand. The Fe III -tetraphenylporphyrine units are linked by a 2 -oxido ligand into a dimer with an Fe-O-Fe angle close to linearity. The final coordination sphere for each Fe III atom is square-pyramidal with the 2 -oxido ligand in the apical position. The crystal under investigation consisted of two domains in a ratio of 0.691 (3): 0.309 (3).
In a previous publication, we have reported the first lightcontrolled molecular spin switch based on Fe III tetraphenylporphyrin perchlorate (FeTPPClO 4 ) (Shankar et al., 2018). The starting material FeTPPClO 4 exists in the admixed-spin state (S = 3/2, 5/2). However, in a solution of acetone/dimethyl sulfoxide, a high-spin (S = 5/2) complex is formed (Shankar et al., 2018). The low-spin (S = 1/2) state can be induced by a photoswitchable azopyridine ligand and can be reversibly switched to the high-spin state by exposure to light (Shankar et al., 2018;Peters et al., 2019). This system is reversible by using dimethyl sulfoxide and is neither oxygen nor water sensitive, and no fatigue was observed after more than 1000 switching cycles (Shankar et al., 2018). Unfortunately, without dimethyl sulfoxide, the switching is not reversible and a by-product is formed as indicated from the shift of the pyrrol protons observed in an NMR experiment. The amount of this byproduct increases with increasing reaction time. To identify the nature of this by-product, we tried to obtain single crystals after very long reaction times, but without any success. If, however, 4-methylimidazole is used instead of a azopyridine ligand, dark red-coloured crystals of the same by-product were obtained. The crystals were subjected to single-crystal X-ray diffraction analysis, revealing that a dimer has formed where two Fe III cations are bridged by a 2 -oxido ligand. The source of oxygen is still unknown but it is likely that it possibly originates from water or from hygroscopic 4-methylimidazole. It is noted that a crystal structure of this compound has already been reported (Strauss et al., 1987) but this form crystallizes in the orthorhombic space group Aba2 (Hoffman et al., 1972;Swepston & Ibers, 1985;Kooijmann et al. 2007). Therefore, the new polymorph of the title compound was further investigated, and its crystal structure is reported in this communication.

Structural commentary
In the crystal structure of the triclinic polymorph of the title compound, the two crystallographically independent Fe III cations are each coordinated by the four N atoms of tetraphenylporphyrin ligands in a square-planar environment (Figs. 1 and 2). These complexes are linked into dimers via a 2oxido O atom, leading to a final square-pyramidal coordination for each of the Fe III cations (Fig. 2), with 5 values (Addison et al., 1984) of 0.04 (Fe1) and 0.01 (Fe2), indicating only slight deviations from the ideal geometry for which 5 = 0. For Fe1 the Fe-N bond lengths are very similar, whereas for Fe2 they are slightly different (Table 1). There are also small differences in the Fe-O distances, which shows that the bridge is not symmetrical [the Fe-O-Fe angle is 177.71 (18) ]. This is in contrast to the orthorhombic form where both Fe-O distances are identical because of symmetry restrictions as this complex is located on a twofold rotation axis (Hoffman et al., 1972;Swepston & Ibers, 1985;Kooijmann et al. 2007). Nevertheless, the orthorhombic form likewise shows a small distortion of the coordination polyhedron around Fe III , and in both modifications the Fe III cations are shifted out of the porphyrine plane in direction towards the O atoms [0.366 (1) Å for Fe1 and 0.399 (1) Å for Fe2 in the monoclinic structure of the title compound; Fig. 2]. The porphyrine ring planes in the title compound are rotated by 28.5 (5) against each other, whereas in the orthorhombic form they exhibit an almost staggered arrangement of the Fe-N bonds, close to D 4d symmetry.

Supramolecular features
In the crystal structure of the title compound, the dimers are arranged in columns that elongate parallel to the b axis (Fig. 3). There are no hydrogen bonds between the dimers, and there is also no hint of significantinteractions. Therefore, the packing appears to be dominated by non-directed van der Figure 1 Molecular structure of the title compound with atom labelling and displacement ellipsoids drawn at the 50% probability level. The H atoms are omitted for clarity; the disorder of one of the phenyl rings is shown with full and open bonds. Waals interactions. It is noted that the packing of the dimers is completely different in the two polymorphic forms. In the orthorhombic form, the dimers are also arranged in columns but neighbouring columns are shifted relative to each other; for comparison of the two polymorphs, see Figs. 3 and 4. The density of the triclinic polymorph is slightly higher than that of the orthorhombic form, indicating that the former most probably represents the thermodynamic stable form at absolute zero.

Synthesis and crystallization
FeTPPClO 4 was synthesized as reported (Shankar et al., 2018). The layering technique was used for crystallization. The lower layer consisted of FeTPPClO 4 dissolved in dichloromethane to which 50 ml 4-methylimidazole were added, and n-heptane was used as the upper antisolvent. Crystal structure of the orthorhombic form of the title compound in a view along the c axis.

Figure 3
Crystal structure of the title compound in a view along the b axis. The disorder of one of the phenyl rings is omitted for clarity.

Figure 2
Top and side view of the molecular structure of the title compound showing the coordination around the Fe III atoms. The disorder of one of the phenyl rings is omitted for clarity.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 2.
All crystals consisted of more than one domain, but the structure could be solved in space group P1 neglecting the presence of two domains. However, these refinement runs led to poor reliability factors and several electron density maxima were observed that could not be resolved. The TwinRotMat option in PLATON (Spek, 2009) suggested a twofold rotation axis as twin element with the matrix (1 0 0, 0 1 0, À0.389, À0.663 1). Several data sets in HKLF-5 format were generated using different sizes of the integration box in X-AREA (Stoe, 2008) and different overlap criteria in PLATON (Spek, 2009) until the best data set was obtained. The final refinement using this data set led to a ratio of the two domains of 0.691 (3): 0.309 (3) and acceptable reliability factors.

µ-Oxido-bis[(5,10,15,20-tetraphenylporphyrinato)iron(III)]
Crystal data  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.