Crystal structure of di-μ-chlorido-bis[dichloridobis(methanol-κO)iridium(III)] dihydrate: a surprisingly simple chloridoiridium(III) dinuclear complex with methanol ligands

While attempting to synthesize a cyclopentadienyl iridium complex by the reaction between IrCl3·xH2O in methanol, several well-shaped crystals formed from the reaction mixture. Surprisingly, the crystals were of di-μ-chlorido-bis[dichloridobis(methanol-κO)iridium(III)] dihydrate, [Ir2Cl6(CH3OH)4]·2H2O. This is a surprising result in that, while many reactions of iridium chloride hydrate are carried out in alcoholic solvents, especially methanol and ethanol, this is the first structure of a chlorido-iridium compound with only methanol ligands.


Chemical context
The use of alcoholic solvents with IrCl 3 ÁxH 2 O for the formation of cyclopentadienyl or olefin iridium complexes is exceedingly common (Herde et al., 2007;Liu et al., 2008Liu et al., , 2011Morris et al., 2014). Lately, we have been investigating the syntheses of half-sandwich iridium complexes with varying tetramethylalkylcyclopentadienyl ligands (Morris et al., 2014). In all cases, the reaction takes place between IrCl 3 ÁxH 2 O and the tetramethylalkylcyclopentadiene in methanol, either under thermal or microwave conditions. In most cases, the yields of of Cp* R iridium chlorido-bridged dimers are good to excellent. Several reactions to synthesize Cp* R iridium complexes with R = long-chain alkyls such as n-hexyl, n-heptyl and n-octyl produced good yields of the desired [Cp* R IrCl 2 ] 2 compounds but, in one instance, only produced a few crystals which turned out to be those of the title compound. Given the number of reactions that are carried out with IrCl 3 ÁxH 2 O in methanol, that this is the first time this compound has been seen by us or by any others active in the field is surprising.

Structural commentary
The title structure ( Fig. 1) consists of two iridium-centered octahedra sharing one edge via chloride bridges. For each octahedron, there are two terminal chloride ligands in the same plane as the bridging chloride ligands. The axial positions that complete the octahedra are occupied by O-bonded methanol ligands. One of the methanol ligands on each iridium atom is hydrogen-bonded to a lattice water. The two iridium-centered octahedra are related by an inversion center. The Ir-Cl bridges are symmetrical with identical Ir-Cl bond lengths of 2.385 (1) Å with two of the Ir-Cl bonds equivalent by symmetry and the unique bonds coincidentally equivalent [2.3847 (10) and 2.3846 (11) Å ]. The only structure similar to the title compound currently in the Cambridge Structural Database (CSD version 5.35 with updates, Groom & Allen, 2014) is CCDC: CLESIR, bis(-chlorido)tetrachloridotetrakis(diethylsulfide)diiridium(III) (Williams et al., 1980). The structural similarities between CLESIR and the title compound are that both contain octahedrally coordinated iridium atoms with the octahedra sharing one edge via Ir-Cl-Ir bridges. There are also two terminal chloride ligands on each iridium for both compounds. In the case of CLESIR, however, the remaining ligands on the iridium are diethylsulfido ligands. An additional difference is that, for the title compound, all chloride ligands are in the equatorial plane with methanol ligands occupying axial positions. For CLESIR, the diethylsulfido ligands on one iridium atom occupy axial positions but occupy equatorial positions on the second iridium.

Supramolecular features
Each lattice water molecule forms four hydrogen bonds linking four different iridium-centered dimers. Table 1 lists the various parameters describing the hydrogen bonding. As a donor, the water participates in two O-HÁ Á ÁCl bonds to chloride ligands on different molecules while, as acceptor, the water participates in two O-HÁ Á ÁO bonds to methanol oxygen atoms on two additional molecules. A search of the CSD for O-HÁ Á ÁCl bonds between lattice water and chloride attached to any transition metal followed by analysis in Mercury (Macrae et al., 2008) show that the OÁ Á ÁCl distances have a mean of 3.151 Å with a mean deviation of 0.055 Å . The two OÁ Á ÁCl distances of 3.208 (4) and 3.285 (3) Å for this structure places the distances at the high end of the range. However, when acting as acceptors, the lattice water displays O(methanol)-O(water) distances of 2.752 (5) and 2.647 (5) Å . A search of the CSD with analysis by Mercury (Macrae et al., 2008) uncovers a mean O(donor)Á Á ÁO(acceptor) distance of 2.742 Å with a mean deviation of 0.085 Å , putting the donor-acceptor distances at the mean and slightly below the mean of these types of hydrogen bonds. Fig. 2 shows the hydrogen-bonding network that is created throughout the lattice of the title compound with the methanol methyl groups removed for clarity.

Figure 2
Packing diagram of the title compound showing the hydrogen-bonding (dashed lines) network. Displacement ellipsoids are drawn at the 50% probability level. the result). Analysis with Mercury (Macrae et al., 2008) found that Ir-O bonds in this small subset ranged from 2.185 to 2.317 Å with a mean of 2.251 Å and a standard deviation of 0.042 Å . The Ir-O bond lengths of the title compound of 2.066 (3) and 2.057 (3) Å are significantly smaller than the low end of this range. The small number of samples and the variety of structures available for comparison do not permit any clear conclusions as to the significance of these distances. All of the structures are of iridium(III) but the title compound is the only one with chloride as the sole other ligand set on each metal. While this structure determination was carried out at 100 K compared with room temperature for most of the other compounds with methanol ligands, such a significant bond shortening would not be expected based solely on temperature (Macchi & Sironi, 2004).

Synthesis and crystallization
IrCl 3 ÁxH 2 O and 1-heptyl, 2,3,4,5-tetramethylcyclopentadiene were mixed in a round-bottom flask with 15 mL of MeOH and the reaction mixture was refluxed for two days. This procedure has been successfully used to synthesize a number of pentaalkyliridium chloride compounds in the past. After cooling to room temperature, the round-bottom flask was placed into a freezer overnight. There was no evidence of any product crystallization. The reaction mixture was then evaporated to dryness, yielding a tarry mixture. The tarry mixture was dissolved in diethyl ether and allowed to evaporate slowly. After the ether had evaporated, the mixture was again very tarry in appearance, but this time with a few crystals obvious in the flask. The structure of the title compound was determined from one of those crystals. It is unclear why this reaction did not proceed normally.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 2