Crystal structures of two PCN pincer iridium complexes and one PCP pincer carbodiphosphorane iridium intermediate: substitution of one phosphine moiety of a carbodiphosphorane by an organic azide

The syntheses and crystal structures of two PCN pincer iridium complexes, prepared from the reaction of their respective PCP pincer carbodiphosphorane iridium precursors with an organic azide, are reported. Crystal data for one of the precursors is also discussed.


Chemical context
Carbodiphosphoranes (CDPs), also termed double ylides, consist of two tertiary phosphines connected to a central divalent carbon(0) atom. The P-C bonds are best described as donor-acceptor interactions (Petz & Frenking, 2010). Most of the chemistry associated with CDPs concerns compounds with Lewis acids. Since the central CDP carbon possesses two ISSN 2056-9890 lone electron pairs, it is therefore able to interact with either one or two Lewis acids (Chauvin & Canac, 2010;Petz & Frenking, 2010). Reactions involving the cleavage of the P-C bonds of the CDP functionality are less common in contrast to phosphorus ylides (Petz & Frenking, 2010;Kolodiazhnyi, 1999). We hereby report the non-innocent reactivity (Poverenov & Milstein, 2013) of a PCP pincer ligand, whose central carbon is part of a CDP functionality, with an organic azide in the coordination sphere of iridium.
We believe that the reaction is initiated by an interaction of the electrophilic organic azide with the central CDP carbon of the PCP ligand, which disposes of one lone electron pair (Petz & Frenking, 2010). In a related reaction, N-heterocyclic carbenes (NHCs) have been reported to form end-on adducts with organic azides to form triazenes (Khramov & Bielawski, 2005). The interaction of the CDP with the organic azide results in the formation of a double bond between the central carbon and the terminal nitrogen of the organic azide and is associated with the cleavage of one P-C bond of the CDP functionality, while the carbon-iridium bond remains intact. At this stage, a deeply coloured and presumably five-coordinate Ir I intermediate was detected by monitoring the reaction via 31 P-NMR spectroscopy. This intermediate features the triazenylidenephosphorane ligand (4-Cl-C 6 H 4 N 3 )C(dppm) and a monodentate dppm unit. The absence of a hydrido ligand is attributed to an antecedent reductive elimination of hydrochloric acid, which, according to NMR spectroscopic results, is absorbed by the CDP carbon of the starting complex 1a. This carbon atom turns out to be the strongest base of the system, apparently more basic than the nitrogen atoms and the central carbon of the PCN pincer ligand of 2. Consequently, only 50% of the educt is converted to 2. However, an almost quantitative and fast conversion into 2 was achieved upon addition of basic alumina. The formation of 2 is finalized via the dissociation of a chlorido ligand and the coordination of the displaced phosphine functionality to the Ir centre.
At first sight, the resulting PCN pincer ligand of 3 looks like a tautomer of the PCN pincer ligand of 2: while in the PCN pincer of 3, one proton is attached to C1 and C2 respectively, the PCN pincer of 2 carries two protons at C2 and none at C1. In contrast to the neutral ligand of 2, the PCN ligand in 3 carries a double negative charge, deduced as follows: first, in view of coordination number 6, 3 constitutes an Ir III complex. Second, the coordination compound 3 carries no charge. Third, the cyanido ligand contributes a À1 charge, and the iridium central atom a +3 charge. Since the dppm ligand is neutral, the charge of the PCN pincer ligand can be calculated to be À2. We suspect that the pathway of the reaction is similar to the formation of 2, except that the cyanido ligand permanently stays in the coordination sphere of iridium. The coordination of the displaced phosphine functionality to the Ir I centre is thought to induce a two-electron transfer from iridium to the PCN ligand related to an oxidative addition reaction, and to be followed by the transfer of one proton from C2 to C1.

Structural commentary
The structures of compounds 2, 1b and 3 are given in Figs. 1, 2 and 3, respectively. Selected bond lengths and angles for all three compounds are given in Table 1.
The structure of 1b (Fig. 2) displays an octahedral iridium(III) coordination compound with a meridional C(dppm) 2 PCP pincer ligand and one chlorido ligand situated trans to the central CDP carbon atom. The remaining sites are occupied by the hydrido and cyanido ligands positioned trans to each other. The structure is closely related to that of [Ir(Cl) 2 (H)(C(dppm) 2 )-3 P,C,P)] (1c) (Partl et al., 2018), which contains one chlorido ligand instead of the cyanido ligand trans to the hydrido ligand.  Structure of 3 with displacement ellipsoids drawn at the 30% probability level. Only the ipso carbon atoms of the phenyl groups are shown for clarity.

Supramolecular features
In the crystal of 2, supramolecular features appear to revolve around the chloride anion (Table 2): Cl1 interacts with the methylene group of one dppm unit (C2-H2BÁ Á ÁCl1 = 2.62 Å ) and to a proton of one dichloromethane molecule (C11-H11BÁ Á ÁCl1 i = 2.49 Å ). It must be mentioned, however, that due to the positional disorder of both the chloride anion and the dichloromethane solvate units, these 'bond' lengths are an estimation and may not necessarily reflect any actual intermolecular interactions.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 5  and C 7 H 8 show occupational disorder, with the toluene molecule exhibiting additional 1:1 positional disorder with some nearly overlying carbon atoms. We propose a correlation between CH 2 Cl 2 and C 7 H 8 , because of short intermolecular ClÁ Á ÁC contacts. Therefore, the two solvent molecules Cl3/ C10-Cl4 and Cl5/C11-Cl6 have an occupancy of 0.75 and the 'two' toluene molecules, C12-C18 and C19-C25, an occupancy of 0.25. Several bond restraints were used to refine the toluene carbon atoms reasonably isotropically. The hydride hydrogen of 1b was localized and refined isotropically without restraints.

Special details
Experimental. All data sets were measured with several scans to increase the number of redundant reflections. In our experience this method of averaging redundant reflections replaces in a good approximation semi-empirical absorption methods (absorption correction programs like SORTAV lead to no better data sets). 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.

sup-12
Acta Cryst. (2019). E75, 75-80 Refinement. 2:1 positional disorder of the counter anion Cl1:Cl1A. The solvent molecules CH 2 Cl 2 and C 7 H 8 show occupational disorder, whereas the toluene shows additional positional disorder of ratio 1:1 with some nearly overlying carbon atoms. We think, because of short intermolecular Cl···C contacts between CH 2 Cl 2 and C 7 H 8 , that there is a correlation between these molecules. Therefore, the two solvate molecules Cl3-C10-l4 and Cl5-C11-Cl6 have an occupancy of 0.75 and the `two′ toluene molecules, C12-C18 and C19-C25, an occupancy of 0.25. Several bond restraints must be used to refine the toluene carbon atoms reasonably isotropically.

Special details
Experimental. All data sets were measured with several scans to increase the number of redundant reflections. In our experience this method of averaging redundant reflections replaces in a good approximation semi-empirical absorption methods (absorption correction programs like SORTAV lead to no better data sets). 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. Refinement. Hydrogen atoms at C1 and C2 were localized and refined with isotropic displacement parameters.