Rational synthesis of normal, abnormal and anionic NHC–gallium alkyl complexes: structural, stability and isomerization insights

Using two alternative methodologies, new light has been shed on the stability and rational formation of abnormal NHC–gallium complexes.


Synthesis of [IPrLiGa(CH 2 SiMe 3 ) 4 ] (2)
Li(CH 2 SiMe 3 ) (1M in pentane, 1 mL, 1 mmol) was added to a solution of GaR 3 (0.33 g, 1 mmol in 10 mL hexane) and stirred for 1h at room temperature. To this suspension of [LiGaR 4 ] ∞ , 1 equivalent of IPr (0.39 g, 1 mmol) was added and the resulting orange suspension was stirred for another hour at room temperature. To the resulting orange suspension toluene was added dropwise with gentle heating until all of the visible solid has dissolved. Slow cooling of the resulting solution afforded X-ray quality crystals. The mixture was then concentrated and kept at -26 °C for a couple of days to yield a crop of colourless crystals (0.39 g, 48%). Anal. Calcd for C 43 13 C spectrum was not obtained. By switching to the donor solvent d 8 -THF it was evident from 1 H and 7 Li that the co-complex was broken and the free IPr and LiGaR 4 were identified.

Synthesis of [CH 3 C{[N(2,6-i Pr 2 C 6 H 3 )] 2 CHCGa(CH 2 SiMe 3 ) 3 }] (4)
A toluene solution of 3 (0.43 g, 0.5 mmol in 15 mL of toluene) was cooled down to -80 °C and stirred for 20 min. To this slurry, a toluene solution of MeOTf (0.08 g, 0.5 mmol in 3 mL of toluene) was added dropwise and stirred for an hour. The mixture was filtered through Celite to remove LiOTf and washed with more toluene (5 mL). The solvent was exchanged in vacuo to hexane (5 mL) to which 2 mL of fresh toluene were added. Obtained suspension was gently heated until a yellow solution was obtained which upon slow cooling afforded X-ray quality crystals. This mixture was then kept overnight at -30 °C to yield a crop of colourless crystals (0.25 g, 68%). Anal. Calcd for C 40 H 71 N 2 Si 3 Ga: C, 65.28; H, 10.00; N, 3.81. Found: C, 65.04; H, 9.91; N, 4.08.

Synthesis of [aIPrGa(CH 2 SiMe 3 ) 3 ] (5)
To a THF solution of 3 (0.43 g, 0.5 mmol in 10 mL of THF) IMesHCl (0.17 g, 0.5 mmol) was added from solid addition tube and stirred for 6h at room temperature. The mixture was filtered through Celite and washed with more THF (2 x 5 mL).
Clear filtrate was concentrated to ca. 5 mL in volume to which 2 mL of hexane was added and stored at -30 °C to afford colourless crystals of title compound (0.22 g, 61%

Synthesis of [IPrZn(CH 2 SiMe 3 ) 2 ] (6)
Zn(CH 2 SiMe 3 ) 2 (0.92 mL, 0.54 M in hexane, 0.5 mmol) was added via syringe to a suspension of IPr (0.19 g, 0.5 mmol) in hexane (10 mL) at room temperature to form a white suspension and stirred for 15 min at room temperature. The reaction mixture was then gently heated until all of the visible solid had dissolved. Slow cooling of the resulting solution afforded Xray quality crystals (0.22 g, 70%

Synthesis of [aI t BuGa(CH 2 SiMe 3 ) 3 ] (8)
Equimolar amounts of Ga(CH 2 SiMe 3 ) 3 (0.17 g, 0.5 mmol) and bis(tert-butyl)imidazol-2-ylidene (IBu) (0.09 g, 0.5 mmol) were suspended in hexane (10 ml) and stirred for one hour at room temperature. The resulting white suspension was gently heated until all of the visible solid had dissolved. Slow cooling of the resulting solution afforded a crop of colourless crystals (0.11 g, 43%).  Isolated compound 1 was used as a model system to study the influence of the solvent and additives. A 0.25 M solutions of pure crystalline compound 1 in deuterated solvent (C 6 D 6 or d 8 -THF) were prepared and sealed in Young's tap NMR tubes. Sealed tube was heated at 100 °C for a specific time followed by recording of 1 H NMR spectra (at room temperature) on a Bruker DPX 400 MHz spectrometer, operating at 400.13 MHz. Yields were calculated versus ferrocene which was used as an internal standard. We followed the isomerisation of pure 1 into 5 in C 6 D 6 ( Fig S2), d 8 -THF ( Fig S3) and then in the presence of excess of IPr ( Fig S4) and gallium regent ( Fig S5).

Kinetic measurements
1. Kinetic isotopic effect 65 mg of pure crystalline compound 1 was dissolved in 0.4 mL of d 8 -THF in the glovebox. The reaction mixture was transferred into a sealed Youngs tap NMR tube. The reaction was heated at 100 °C and regularly monitored by 1 H NMR spectroscopy to determine yields which were calculated by integrating the iPr-methine protons of the product (5) versus the ferrocene standard. For kinetic isotopic experiment another sample was prepared in exactly the same fashion but with deuterium atoms incorporated in the starting material (1 D ). The data were plotted as molar concentration of the product versus time yielding straight lines, which were fitted by conventional linear regression (r 2 > 0.96) and k values were obtained from the corresponding slopes.

Initial rates
Isomerization of 1 into 5 in d 8 -THF at 323K was monitored using in situ NMR spectroscopy by following the appearance of the resonance assigned to the new C carbene -H bond (9ppm). The percentage of conversion was restricted to 5-7 % in order to calculate the initial rate (r o ) of the reaction. The data were plotted as molar concentration of the product versus time yielding

DFT calculations
Density Functional Theory (DFT) calculations 11 were performed using the Gaussian computational package G03. 12 In this series of calculations the geometries of the molecules and ions were optimised by employing the B3LYP density functionals 13,14 and the 6-311G** basis set. 15,16 The charge distributions were obtained from a Natural Bond Orbital analysis. 17

Electronic Supplementary Information
ESI22 Table S7. Optimized geometry of GaR 3 .

Electronic Supplementary Information
ESI28 Figure S27 Representation of molecular orbitals HOMO-1, HOMO and LUMO of VI IPr .

Optimized geometry of normal IMes (I IMes )
Table S12 Optimized geometry of I IMes . 9. Optimized geometry of abnormal IMes (II IMes )  11. Optimized geometry of abnormal IMes·GaR 3 complex (IV IMes )  12. Optimized geometry of normal I t Bu (I ItBu ) 13. Optimized geometry of abnormal ItBu (II ItBu ) 14. Optimized geometry of normal IBu·GaR 3 complex (III ItBu ) Several models of normal I t Bu•GaR 3 were constructed and their geometries were optimized. The resulting structures had a long C1-Ga bond and an energy value which showed no stabilisation over that of the separate species. One of the models is shown next.

Electronic Supplementary Information
ESI37 Figure S35 DFT study on the of the reaction of model systems I IPr and GaR 3 to afford intermediate A. Mg(CH 2 SiMe 3 ) 2 n/a n/a -0.11 -1.77

Electronic Supplementary Information
ESI38 Figure S36: 1 H NMR of 1 in C 6 D 6 solution.

Figure S37
1 H NMR of 1 in d 8 -THF solution.

Electronic Supplementary Information
ESI39 Figure S38 13 C{ 1 H} NMR of 1 in C 6 D 6 solution.

Figure S39
1 H DOSY of 1 in C 6 D 6 solution.

Electronic Supplementary Information
ESI41 Figure S42 1 H of 2 in C 6 D 6 solution.

Figure S43
7 Li of 2 in C 6 D 6 solution.

Electronic Supplementary Information
ESI42 Figure S44 1 H of 2 in d 8 -THF solution.

Electronic Supplementary Information
ESI43 Figure S46 1 H NMR of 3 in C 6 D 6 solution.

Figure S53
1 H NMR of 5 in C 6 D 6 solution.

Electronic Supplementary Information
ESI47 Figure S54 13 C NMR of 5 in d 8 -THF solution.

Figure S55
1 H NMR of 6 in C 6 D 6 solution.

Electronic Supplementary Information
ESI48 Figure S56 13 C{ 1 H} NMR of 6 in C 6 D 6 solution.

Figure S57
1 H NMR of 7 in C 6 D 6 solution. # traces of grease.

Electronic Supplementary Information
ESI49 Figure S58 13 C{ 1 H} NMR of 7 in C 6 D 6 solution. # traces of grease.

Figure S59
1 H NMR of 8 in C 6 D 6 solution.