The Reactivity of Phosphanylphosphinidene Complexes of Transition Metals Toward Terminal Dihaloalkanes

The reactivities of phosphanylphosphinidene complexes [(DippN)2W(Cl)(η2-P–PtBu2)]− (1), [(pTol3P)2Pt(η2-P=PtBu2)] (2), and [(dppe)Pt(η2-P=PtBu2)] (3) toward dihaloalkanes and methyl iodide were investigated. The reactions of the anionic tungsten complex (1) with stochiometric Br(CH2)nBr (n = 3, 4, 6) led to the formation of neutral complexes with a tBu2PP(CH2)3Br ligand or neutral dinuclear complexes with unusual tetradentate tBu2PP(CH2)nPPtBu2 ligands (n = 4, 6). The methylation of platinum complexes 2 and 3 with MeI yielded neutral or cationic complexes bearing side-on coordinated tBu2P–P-Me moieties. The reaction of 2 with I(CH2)2I gave a platinum complex with a tBu2P—P—I ligand. When the same dihaloalkane was reacted with 3, the P—P bond in the phosphanylphosphinidene ligand was cleaved to yield tBu2PI, phosphorus polymers, [(dppe)PtI2] and C2H4. Furthermore, the reaction of 3 with Br(CH2)2Br yielded dinuclear complex bearing a tetraphosphorus tBu2PPPPtBu2 ligand in the coordination sphere of the platinum. The molecular structures of the isolated products were established in the solid state and in solution by single-crystal X-ray diffraction and NMR spectroscopy. DFT studies indicated that the polyphosphorus ligands in the obtained complexes possess structures similar to free phosphenium cations tBu2P+=P−R (R = Me, I) or (tBu2P+=P)2.

Data collection and data reduction were controlled by using the X-Area 1.75 program (STOE, 2015). An absorption correction was performed using integrated reflections by a combination of frame scaling, reflection scaling and a spherical absorption correction. The structures were solved using intrinsic phasing implemented in SHELXT and refined anisotropically using the program packages Olex2 1 and SHELX-2015 2,3 .Positions of the C-H hydrogen atoms were calculated geometrically taking into account isotropic temperature factors. All H-atoms were refined as riding on their parent atoms with the usual restraints.
Special treatment. Tungsten complexes. Structure 1b was refined with substitutional disorder of Br1/Cl1 atoms with site occupation factors of 0.650(5)/0.350 (5), respectively. Several intensities (namely 11 reflections) were affected by beamstop and were excluded from the refinement. The same kind of substitutional disorder was found in 1c where Br1/Cl1 atoms have site occupation factors of 0.308(5)/0.692 (5), respectively, and in 1e where Br1/Cl1 atoms have site occupation factors of 0.319(7)/0.681 (7), respectively. Structure 1e contains the main molecule sitting on the rotational 2-fold axis, with the centre of the alkyl chain disordered. Additionally, disordered eight solvent molecules of pentane were excluded from the refinement using the SQUEEZE procedure.
Platinum complexes. Structure 2a did not require any special conditions. Structure 2b also was solved and refined by routine methods, however it contained a strong electron density peak between P1 and Pt1 atoms. In our opinion it is an artefact (e.g. due to cut of Fourier series), since no atom can be present at that location. Hydrogen atom presence must be excluded based on 1 H and 31 P NMR spectra. Structure 2c required to keep three carbon atoms C12, C13 and C14 in isotropic model to avoid abnormal expansion of the displacement ellipsoids. As a side effect two electron holes were created in neighbourhood of bromine atoms, which are hard to remove and most probably are just artefacts. Structural analysis for 3a produced a result with electron density map affected by minor disorder/twinning. As a result six strong residual electron density peaks are present in vicinity of I2 and carbon atoms rising A-level alerts by the checkcif procedure. We repeated the diffraction experiment thrice with different crystal specimens with the same result. Probably the twinning or disorder has minor contribution so it is hard to model, nevertheless the heaviest atoms, platinum and iodine, give relatively high distortion of the resultant electron density map at their new positions. Structure 3c was solved and refined using classical methods with four reflections omitted from the refinement as being affected by beamstop and being outliers not reproducing the actual reciprocal space.
Crystallographic data for all structures reported in this paper have been deposited with the Cambridge Crystallographic Data Centre as supplementary publication No. CCDC 1963412-1963419. Copies of the data can be obtained free of charge on application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (Fax: (+44) 1223-336-033; E mail: deposit@ccdc.cam.ac.uk).

IR SPECTROSCOPIC DATA
The FTIR spectra of crystalline products were recorded using a Nicolet iS50 FT-IR spectrometer equipped with the Specac Quest single-reflection diamond attenuated total reflectance (ATR) accessory. Spectral analysis was carried out by using the OMNIC software package.        9 . All atoms were described by a Slater-type triple-ζ quality basis set with two polarization functions, corresponding to TZ2P basis set 10 in the ADF package. Relativistic effects were included using scalar Zeroth Order Regular Approximation (scalar ZORA) model [11][12][13] . The functional and basis set were chosen according to our experience. [14][15][16] Moreover, results of this work may be compared with the one previously published by our group as the same basis set and functional were used.
Starting geometries for all compounds were taken from experimental crystallographic data. Starting geometries for non-coordinated molecules tBu2PPMe, tBu2PPI and tBu2PPPPtBu2 were taken from complexes 2a, 2b and 3c, respectively. Then the geometries were optimized for each spin state (singlet and triplet) and charge (+1 and -1). Geometries of complexes 1a, 2a, 2b were not optimized and experimental coordinates were used for further calculations.
As we were not able to perform calculations for 3c due to the size of the system, we have replaced Ph substituents with methyl groups (3c') in phosphine ligands and then optimized such obtained structure. On optimized geometries of tBu2PPMe, tBu2PPI, tBu2PPPPtBu2 and 3c' and non-optimized geometries of 1a, 2a, 2b a series of other calculations were conducted -Natural Bonding Orbitals (NBO), Hirshfeld population analysis 17 , Mayer Bond Order analysis 18 . Natural Bonding Orbitals (NBO, version 6.0) 19 analysis was performed for all systems including calculations of Natural Localized Molecular Orbitals (NLMO) 20 and Natural Population Analysis (NPA) 21 . Condensed Fukui functions were obtained using Hirshfeld Population Analysis to quantitatively assign properties of atoms.
To elucidate diversified reactivity of phosphanylphosphinidene tungsten complexes towards dihalogenalkanes, we propose simple model estimating energetic effects associated with formation of four-, five-, six-and seven-membered rings products as presented in Figure  S49. To this end, we have optimized structures of respective reagents and performed harmonic frequency calculations to obtain values of free energy of reactions. The result are presented in Table S4. All calculations for compounds c-CNP and n-CNP (N = number of carbon atoms in aliphatic chain) presented in the paper were performed using the Gaussian 09 22 program package. Molecular geometries were optimized by using the ωB97XD functional by Head-Gordon 23 with 6-31+G(d,p) basis set. Nature of the final gas phase geometries as a local minima (no imaginary frequencies) on the potential energy surface was then validated by harmonic frequency calculations at the same level of theory. Values of calculated energies, enthalpies and free energies derived from thermochemical calculations were corrected for the zero-point energy (ZPE).