Exploring Equilibria between Aluminium(I) and Aluminium(III): The Formation of Dihydroalanes, Masked Dialumenes and Aluminium(I) Species

Abstract The design of new reductive routes to low oxidation state aluminium (Al) compounds offers the opportunity to better understand redox processes at the metal centre and develop reactivity accordingly. Here, a monomeric AlI compound acts as a stoichiometric reducing agent towards a series of AlIII dihydrides, leading to the formation of new low oxidation state species including symmetric and asymmetric dihydrodialanes, and a masked dialumene. These compounds are formed by a series of equilibrium processes involving AlI, AlII and AlIII species and product formation can be manipulated by fine‐tuning the reaction conditions. The transient formation of monomeric AlI compounds is proposed: this is shown to be energetically viable by computational (DFT) investigations and reactivity studies show support for the formation of AlI species. Importantly, despite the potential for the equilibrium mixtures to lead to ill‐defined reactivity, controlled reactivity of these low oxidation state species is observed.


General Experimental Section
All manipulations were carried out using standard Schlenk-line and glovebox techniques under an inert atmosphere of argon or dinitrogen. A MBraun Labmaster glovebox was employed, operating at < 0.1 ppm O2 and < 0.1 ppm H2O. Solvents were dried over activated alumina from an SPS (solvent purification system) based upon the Grubbs design and degassed before use. Glassware was dried for 12 h at 120 °C prior to use. Benzene-d6 was stored over activated 3Å molecular sieves. NMR-scale reactions were conducted in J. Young's tap tubes and prepared in a glovebox. All heating mentioned was done using a DrySyn NMR tube heating block. 1 H (tetramethylsilane; 0 ppm) and 13 C (tetramethylsilane; 0 ppm) spectra were obtained on BRUKER 400 MHz or 500 MHz machines unless otherwise stated; all peak intensities are derived from internal standard peaks with values quoted in ppm. Data was processed using the MestReNova or Topsin software. Variable temperature measurements were ran by Dr Abil Aliev. C IV refers to quaternary carbons. Mass spectrometry was ran on a Waters G2-XS Xevo QToF instrument in positive mode by Oscar Ayrton. Elemental analysis was conducted by Elemental Microanalysis Ltd. and have been obtained to the best of our abilities given the extremely air and moisture sensitive nature of the compounds.
Trimethylamine alane and compound A were synthesised according to literature procedures. 1,2 The synthesis of 1 was previously reported in our research group. 3 Diphenyl acetylene was purchased from Sigma Aldrich and dried under high vacuum prior to use. Hexafluorobenzene was purchased from Sigma Aldrich and used without purification. Other chemicals were purchased from Sigma Aldrich, Fluorochem or Alfa Aeser. S6 8: Compound A (23.3 mg, 0.05 mmol) in benzene-d6 (0.6 mL) was added to 6 (41 mg, 0.1 mmol) and the solution was transferred to a JY NMR tube. The reaction was monitored by 1 H NMR spectroscopy and after 1 hour complete consumption of the starting materials and intermediate was observed. The solvent was removed in vacuo and hexane (1 mL) was added. The solution was left to stand at 25 °C and crystals of B formed. The reaction was filtered and the filtrate stored in a freezer at -35 °C, where the product precipitated as orange crystals (16 mg, 40%). Anal. Calc. (Al2C50H68N4): C,78.08;H,8.80;N,7.19. Found: C,77.03;H,8.81;N,7.07. 10: Compound A (23 mg, 0.05 mmol) in benzene-d6 (0.6 mL) was added to 9 (14.4 mg, 0.05 mmol) and the solution was transferred to a JY NMR tube. The reaction was monitored by 1 H NMR spectroscopy and after <15 minutes complete consumption of the starting material was observed. The solvent was removed in vacuo and the orange solid was recrystallised from a mixture of toluene and hexane at -35 °C (34 mg, 91%).
12 a-c: Compound 7 (11 mg, 0.015 mmol) was dissolved in benzene-d6 (0.6 mL) and the solution was transferred to a JY NMR tube. Hexafluorobenzene (5.2 μL, 0.045 mmol) was added and the reaction was heated at 80 °C for 30 minutes. The sample was analysed by 1 H and 19 F NMR spectroscopy and found to contain a mixture of products which have been assigned as 12 a-c and 5. Other minor unidentified products are also found to be present. Attempts to isolate single crystals have thus far been unsuccessful.

Figure S6:
The reaction of 10 with 1 equiv. 9 in benzene-d6 at 25 °C (bottom) and after 1h at 80 °C (top).     Figure S9: An expanded spectrum of the reaction of 2 with excess A at 80°C for 6 days in cyclohexane-d12.

X-ray Crystallographic Data
All crystals were ran on a Agilent Oxford Diffraction SuperNova equipped with a microfocus Cu Kα Xray source and an Atlas CCD detector. Full spheres of data were collected to 0.84 Å resolution with each 1° scan frame in ω collected twice. Total collection time varied depending on size and quality of crystal, and sample temperature. The Cryojet5® used for these measurement is the original prototype device developed by Oxford Instruments and the Pt-resistance sensor is located in the copper-block heat exchanger and not in the nozzle of the instrument close to the sample (in contrast to the CryojetHT® used in the PXRD experiments). Thus the temperatures quoted in these SXD experiments should be treated as nominal (despite stability to much better than 0.1 °C). Using Olex2, 5 the structure was solved with the ShelXT 6 structure solution program using Intrinsic Phasing and refined with the ShelXL 7 refinement package using Least Squares minimisation.
The X-ray crystal structure of 4 Figure S18: The X-ray crystal structure of 4. Hydrogen atoms and solvent molecules omitted for clarity.
Single crystals of 4 were grown by slow evaporation from hexane solution. 4 was found to crystallise in a P-1 space group. The unit cell contained a disordered molecule of hexane that was modelled using a solvent mask (SQUEEZE).

S18
The X-ray crystal structure of 7 Figure S19: The X-ray crystal structure of 7. Select hydrogen atoms and solvent molecules omitted for clarity.
Single crystals of 7 were grown by slow evaporation from hexane solution.
The X-ray crystal structure of 8 Figure S20: The X-ray crystal structure of 8. Select hydrogen atoms and solvent molecules omitted for clarity.
Single crystals of 8 were grown by slow evaporation from hexane solution.
The X-ray crystal structure of 10 Figure S21: The X-ray crystal structure of 10. Select hydrogen atoms and solvent molecules omitted for clarity.
Single crystals of 10 were grown from hexane solution at -35 °C.
The structure of 10 was found to have a small amount of unresolved residual electron density near the hydride (H). This was interpreted as co-crystallised 10-OH and the crystal was modelled as a mixture of 10 (90%) and 10-OH (10%). It was not possible to locate the H atom of the OH group in the minor component. These atoms were refined isotropically.

S20
The X-ray crystal structure of 11 Figure S22: The X-ray crystal structure of 11. Select hydrogen atoms and solvent molecules omitted for clarity.
Single crystals of 11 were grown from benzene solution at 25 °C.

Computational methods
DFT calculations were run using Gaussian 09. Al centres were described with Stuttgart SDDAll RECPs and associated basis sets and the 6-31G** basis sets were used for all other atoms. 8-10 The functions ωB97X, M062X, M06L and B3PW91 were investigated.
Geometry optimisation calculations were performed without symmetry constraints. Free energies reported within the main text are corrected for the effects of benzene solvent (ε=2.2706) using the using the polarizable continuum model (PCM). 11 In addition, single point dispersion corrections were applied to the ωB97X optimised geometries (dispersion corrected ωB97X-D functional). 12 The graphical user interface used to visualise the various properties of the intermediates was GaussView 5.0.8.

Functional Testing
The stuctures of compounds 3 and 4 were optimized using a split 6-31G**(C, H, N)/SDDAll (Al) basis set and series of different fuctionals. Key bond lengths were compared with single crystal X-ray diffraction data (SCXRD). The models all slightly over estimate the bond lengeths between the Al metal centre and heteroatoms, but in both cases the M06L functional was found to provide the best agreement between experimental and theoretical Al-Al bond lengths. The M06L functional was used to further investigate the equilibrium reactions and reported energies include a single point solvent correction using the PCM solvent model (solvent=benzene).   Figure S23: Proposed reaction pathway for the formation of 2-4 and B (M06L; Al (SDDAll), C H N (6-31G**)+ ∆Esolv (PCM, benzene)). Gibbs free energies relative to starting materials A+1 (kcal mol -1 ).

Figure S24:
The relative Gibbs free energies of products from the reaction of 1 and A (kcal mol -1 ).

Figure S25:
The relative Gibbs free energies of products for two different reaction pathways from the proposed intermediate AmAl(I) (kcal mol -1 ).

Figure S27
: Proposed reaction pathway for the formation of 7 and B. Gibbs free energies relative to starting materials A+5 (kcal mol -1 ).

Figure S28:
The relative Gibbs free energies of products from the reaction of 5 and A (kcal mol -1 ).

Figure S29
: Proposed reaction pathway for the formation of 10. The relative Gibbs free energies of products from the reaction of 9 and A (kcal mol -1 ).  Figure S30: The relative Gibbs free energies of products from the reaction of 9 and A (kcal mol -1 ).      Figure S45: 13 C{ 1 H} NMR spectrum of compound 11.