The Continuum Between Hexagonal Planar and Trigonal Planar Geometries

Abstract New heterometallic hydride complexes that involve the addition of {Mg−H} and {Zn−H} bonds to group 10 transition metals (Pd, Pt) are reported. The side‐on coordination of a single {Mg−H} to Pd forms a well‐defined σ‐complex. In contrast, addition of three {Mg−H} or {Zn−H} bonds to Pd or Pt results in the formation of planar complexes with subtly different geometries. We compare their structures through experiment (X‐ray diffraction, neutron diffraction, multinuclear NMR), computational methods (DFT, QTAIM, NCIPlot), and theoretical analysis (MO diagram, Walsh diagram). These species can be described as snapshots along a continuum of bonding between ideal trigonal planar and hexagonal planar geometries.


General Experimental
Unless otherwise specified, all manipulations were carried out using standard Schlenk and glovebox techniques, under inert atmosphere (nitrogen or argon). A MBRAUN Labmaster glovebox was employed operating with concentrations of H2O and O2 below 0.1 ppm.
Anhydrous solvents were obtained from a Grubbs type SPS system and stored over activated 3Å molecular sieves under inert atmosphere. Alternatively, they were dried using molecular sieves and degassed by freeze-pump-thaw procedures. Stable liquid organic reagents were dried over 3Å molecular sieves and degassed by freeze-pump-thaw cycles before use. All other reagents were obtained from commercial suppliers (Sigma-Aldrich, Alfa Aesar, Fluorochem) and used without further purification. The synthesis and characterisation of the β-diketiminate ligand, 1 compounds 2b and 3a, 2 3b 3 and 6a 4 have been described elsewhere.
[1]2 was prepared by modified literature procedures as detailed below. 13 CH3I 5 and 13 CH3Li 6 were synthesised sequentially from 13 CH3OH (99 atom % 13  Single crystal X-ray data for compounds 2a, 4, 5 and 6b were collected using an Agilent Xcalibur PX Ultra A diffractometer, and the structures were refined using the SHELXTL and SHELX-2013 program systems. 8

X-ray Diffraction Study of Side-Product
Two distinct products can be crystallized from the reaction between [PdMe2(κ 2 -TMEDA)] and 3a. Larger pale yellow block crystals of 4 and smaller colourless crystals can be separated by hand under a microscope. X-ray diffraction experiments undertaken with the colourless crystals were resolved to show a dimeric structure with two [{(DippNCMe)2CH}Zn] units, but further analysis was inconclusive due to unassignable electron density (max peak 8.3, R1 = 18.55%, GooF = 1.581) between the two Zn centers. This structure was persistent across several crystals that were analysed, and appears to be a mixture of different species. The analogous [{(MesNCMe) 2 CH}ZnMe] species was isolated from the reaction between [PdMe2(κ 2 -TMEDA)] and 3b, and afforded crystals suitable for X-ray diffraction which confirmed the structure. The [{(MesNCMe)2CH}ZnMe] complex has been previously reported and gives identical data to the crystals obtained here. 12 S17 5. Crystallographic Data -X-ray Diffraction X-ray crystal structure of 2a. The C61-based included hexane solvent molecule in the structure of 2a was found to be disordered across a centre of symmetry, and two unique orientations were identified of ca. 26 and 24% occupancy (with two further orientations of the same occupancies being generated by operation of the inversion centre). The geometries of the two unique orientations were optimised, the thermal parameters of adjacent atoms were restrained to be similar, and all the atoms of both unique orientations were refined isotropically.
The Pd-H-Mg bridging hydrogen atom was located from a ΔF map and refined freely.
X-ray crystal structure of 4. The structure of 4 was found to sit across a centre of symmetry at the middle of the central C-C bond of the bridging N006-based TMEDA ligand. The C01based included hexane solvent molecule was found to be disordered across a centre of symmetry, with one unique orientation identified. The geometries of the hexane solvent was optimised and refined anisotropically. The Pd-H-Zn bridging hydrogen atoms were located from a ΔF map and refined freely.
X-ray crystal structure of 5. The C21-and C33-based iso-propyl groups in the structure of 5 were both found to be disordered and in each case two orientations were identified, of ca.
66:34 and 57:43%, respectively. The geometries of each pair of orientations were optimised, the thermal parameters of adjacent atoms were restrained to be similar, and both pairs of orientations were refined anisotropically. The C86-based included solvent molecule was found to sit in one position and was refined anisotropically. Pd-H-Zn bridging hydrogen atoms were located from a ΔF map and refined freely.
X-ray crystal structure of 6b. The C12-, C27-, and C42-based iso-propyl groups in the structure of 6b were all found to be disordered and in each case two orientations were identified, of ca. 51:49, 54:46 and 87:13% occupancy respectively. The geometries of each pair of orientations were optimised, the thermal parameters of adjacent atoms were restrained to be similar, and only the non-hydrogen atoms of the major occupancy orientations were refined anisotropically (those of the minor occupancy orientations were refined isotropically).
The included solvent was found to be highly disordered, and the best approach to handling this diffuse electron density was found to be the SQUEEZE routine of PLATON. 13 16 The bonding in these ternary hydrides is likely dominated by an ionic interaction between the {PdHn} nand M + or M 2+ fragments with limited or no interaction between the metals themselves. As such it is reasonable to conclude that the structures of 2, 5 and 6 are not identical to those known from ionic metal salts as both the geometries and short metal---metal distances are inconsistent with those established for ternary hydrides.

Computational Methods
DFT calculations were run using Gaussian 09 (Revision D.01) 22 and Gaussian 16 23 using the B97X hybrid exchange-correlation functional. 24 NBO analysis was performed using NBO 6.0. 25 QTAIM analysis was conducted using the AIMAll package. 26 Data are presented from veryfine mesh calculations but there are no differences in the appearance of bcps when calculations were run with a superfine or ultrafine mesh. Standard cut-offs for plotting the data with AIMAll were used. Non-covalent interactions were analysed using the NCIPLOT 3.0 program. 27 Geometry optimisations were performed without symmetry constraints, unless otherwise specified, and the nature of the stationary points was confirmed as minima by frequency calculations (no imaginary frequencies). The default numerical integration grid was improved using a pruned grid with 99 radial shells and 590 angular points per shell (int=ultrafine). It should be noted that for the geometry and bonding analysis performed in this work, the use of dispersion or solvent corrections was deemed unnecessary.
The level of theory used has previously been benchmarked in our group and shown to reproduce accurately the experimental results. 2,4 Different basis sets (BS1-4) were used as detailed. Geometry optimisations and population analyses on the complexes were performed mainly using BS1. Geometry optimisation of complex 6b with BS1 did not accurately reproduce the experimentally observed Mg-H distances, likely due to the flat nature of the PES. Therefore, the optimisation and population analysis of 6b was performed using a slightly modified basis set, BS2. Optimisation and bonding analysis of the model systems were performed using the larger basis set, BS3, which includes quadruple-ξ functions for Mg and Zn. Finally, wavefunction generation for QTAIM inputs were generated using BS4. Initially single-point calculations for population analyses for all full structures were performed with this basis set as well, but the systems were found to be too big. Bonding analysis was carried out for 1 and 3a using BS1, BS3 and BS4, and for 2a and 2b using BS1 and BS4, with no significant change (±0.03) in values observed (see Table S2). Bonding analysis using BS1 is discussed in the paper.
BS1 was built as follows. 28 The SDD effective core potential was used for all metals (SDDAll).
The split-valence 6-31G(d) basis set was used for C and H atoms. The basis set for metal hydrides was expanded by adding one extra set of diffuse functions and three sets of p-and one set of d-polarisation functions, i.e. formally [6-31++G(d,3pd)]. The triple-ξ 6-311+G* basis set was used for heteroatoms.

S24
BS2 was built as follows. The SDD effective core potential was used for all metals (SDDAll).
The split-valence 6-31G(d,p) basis set was used for C and H atoms, and the triple-ξ 6-311+G* basis set was used for heteroatoms. The metal hydrides used the same expanded basis set as BS1, formally [6-31++G(d,3pd)].
BS3 was built as follows. Pd and Pt were described with the SDD effective core potential, while the other atoms (C, H, N and P) including metals (Mg or Zn) were described using Ahlrichs quadruple-ξ basis set def2-QZVPP. 29 BS4 was built as follows. Pd and Pt were described with the SDD effective core potential, while the other atoms (C, H, N and P) including metals (Mg or Zn) were described using Ahlrichs triplet-ξ basis set def2-TZVPP. 29

Molecular Orbital Analysis
Analysis of 2a' (model) Figure S13. Qualitative MO diagram (idealised to Cs symmetry) of a model of 2a constructed from interaction of a bent PdL2 fragment with a {MgH} + unit.