Catalytic CO2 Reduction with Boron‐ and Aluminum Hydrides

Abstract The previously reported dimeric NHI aluminum dihydrides 1 a,b, as well as the bis(NHI) aluminum dihydride salt 9 +[OTs]−, the bis(NHI) boron dihydride salt 10 +[OTs]−, and the “free” bis(NHI) ligand 12 were investigated with regard to their activity as a homogenous (pre)catalyst in the hydroboration (i. e. catalytic reduction) of carbon dioxide (CO2) in chloroform under mild conditions (i. e. room temperature, 1 atm; NHI=N‐heterocyclic imine, Ts=tosyl). Borane dimethylsulfide complex and catecholborane were used as a hydride source. Surprisingly, the less sterically hindered 1 a exhibited lower catalytic activity than the bulkier 1 b. A similarly unexpected discrepancy was found with the lower catalytic activity of 10 + in comparison to the one of the bis(NHI) 12. The latter is incorporated as the ligand to the boron center in 10 +. To elucidate possible mechanisms for CO2 reduction the compounds were subjected to stoichiometric reactivity studies with the borane or CO2. Aluminum carboxylates 4, 6, and 7 + with two, four, and one formate group per two aluminum centers were isolated. Also, the boron formate salt 11 +[OTs]− was characterized. Selected metal formates were subjected to stoichiometric reactions with boranes and/or tested as a catalyst. We conclude that each type of catalyst (1 a,b, 9 +, 10 +, 12) follows an individual mechanistic pathway for CO2 reduction.

. 1 H NMR spectrum (400.1 MHz, CDCl3) of 4. Figure S2. Expansion of the 1 H NMR spectrum from Figure S1 to show selected signals of the assumed cis-isomer as a minor component (NCH at 6.26 ppm and CO2H at 7.01 ppm). A reaction reaction mixture as for the isolation of 4 was continuously exposed to CO2 over 8 days with frequent repressurizing of the CO2 atmosphere (1.0-1.1 bar) within the otherwise sealed flask. After the indicated period the 1 H NMR spectrum in CDCl3 (i.e. the reaction medium) revealed complete conversion of 4 to 5 and the latter is marked by a characteristic formate HCO2 signal at 7.47 ppm. Figure S4. 1 H NMR spectrum (400.1 MHz, CDCl3) after prolonged exposition of 4 to CO2 atmosphere showing presumed 5 as the major component.

Procedure for the Isolation of {I Mes NAl(CO2H)2}2 (6)
In a 50 mL Schlenk flask 495 mg of 1a (0.71 mmol) were dissolved in CDCl3. The solution is frozen at liquid nitrogen temperature and the flask was evacuated. It was pressurized with CO2 (1.0-1.1 bar) while being thawed in a water bath (fresh from "non-warm" tap). The reaction mixture was stirred in the sealed flask and after 18 h it was repressurized with CO2 atmosphere (1.0-1.1 bar). After stirring at room temperature for a total period of 38 h the volatiles were removed in vacuo. The crude product was suspended in Et2O (5 mL) and THF was added until a clear solution was obtained (ca. 13 mL). The product was crystallized at −30 °C. The cold supernatant was withdrawn from the sedimented solid and the crystalline material was dried in vacuo. From the glass vessel 356 mg of product (6) were collected (0.41 mmol, 58%). Crystals suitable for SCXRD analysis were grown by slow diffusion of solvent vapor from a pentane/toluene mixture into a solution of 6 in CDCl3.   A 25 mL Schlenk flask was charged with 234 mg of 2 (0.26 mmol) and 253 mg BCF (0.49 mmol, BCF = tris(pentafluorophenyl)borane). The mixture of solids was cooled to 0 °C and 3.9 mL CDCl3 were added by syringe. The reaction mixture was stirred for 30 min at 0 °C, the cooling was removed and it was stirred over night at room temperature. An oil separated upon addition of pentane to the reaction mixture.
Separation of the phases and removal of volatiles from the oil leads to formation of an off-white foam that is crushed into a fine powder using a spatulum.  1975.19]: C 55.94, H 3.88, N 4.25; found: C 55.19, H 3.76, N 4.19. The low value found for carbon is explained by the formation of incombustible boron carbides.   Crude 4 (55 mg, 0.06 mmol) and Ph3C + [Al(OC(CF3)3)4] − were mixed in 1 mL toluene and stirred for 24 h.
The resulting suspension was mixed with pentane (1.5 mL). The solid was sedimented and the supernatant withdrawn. The solid was dissolved in 0.7 mL 1,2-difluorobenzene and the vial containing this solution was placed in a reservoir with a mixture of toluene/pentane (2 mL / 4 mL) for gas phase diffusion exchange of solvent. After three days the vial was separated from the reservoir and the solvent slowly allowed to evaporate in a glovebox workstation atmosphere to afford crystals suitable for SCXRD analysis.
Procedure for the synthesis of crude (I Dip NCH2)2•(HOTs)2 (8•(HOTs)2) The compound was synthesized and isolated similar to a procedure for (L Mes NCH2)2•(HOTs)2 [S4] : A Schlenk vessel equipped with a J. Young's PTFE tap and a magnetic stirrer bar was charged with L Dip NH (4.456 g, 11.04 mmol) and 1,2-bistosylethane (2.05 g, 5.5 mmol). To the stirring mixture of solids was added toluene (35 mL) and the pressure within the vessel was reduced (a negligible amount of toluene evaporated or it was cooled to prevent solvent loss). The tap was closed to seal the reaction vessel before placing it into an oil bath heated to 150 °C (the top level of the oil should be a little above the surface level of the reaction mixture), and the reaction was driven for 45 h. The solid was collected on a frit and washed with 20 mL toluene before compressing the filter cake by applying a pressure gradient along the frit. The powder was dried in vacuum after which 5.683 g (4.83 mmol, 88%) of crude product were transfered from the frit.   of Li[AlH4] in THF (0.89 g ≡ 0.98 mL, 0.98 mmol) at −78 °C with stirring. The reaction was continued for 10 min at low temperature then the cooling was removed and it was stirred for 5 h at room temperature (beware: evervescence!). A slightly turbid solution was obtained and the volatiles were evaporated under reduced pressure (30 min). The residue was dried in dynamic vaccum for 3 h to yield a foamy solid. The product was extracted into 4.5 mL CDCl3 and mixing the filtrate with 6 mL pentane resulted in phase separation (liquid/liquid, lower phase with smaller volume). After storage at −30 °C over night no solid had formed. However, upon briefly keeping the biphasic mixture at room temperature few crystal seeds spontanously formed. Crystal growth was continued at −30 °C for one day to yield a major solid fraction.
The cold supernatant was decanted and the solid fraction was homogenized by manipulation with a spatulum before it was extensively dried in dynamic vacuum for one day. From the glass vessel 649 mg of a coloreless powder were transfered that analyzed to 9 + [OTs] − with no lattice solvent included (0.63 mmol, 64%).
Crystals suitable for SCXRD analysis were grown from a mixture of THF, Et2O, pentane and 1,2-difluorobenzene at −30 °C over a period of 10 days.

Procedure for the Isolation of [(I Dip NCH2)2B(H)HCO2] + [OTs] − (11 + [OTs] − )
In a 50 mL Schlenk flask 10 + [OTs] − (377 mg, 0.44 mmol) was dissolved in CDCl3 (4 mL). The solution was frozen at liquid nitrogen temperature and the vessel evacuated. While thawing in a water bath (fresh from "non-warm" tap) the flask was pressurised with CO2 (1.0-1.1 bar) and the reaction driven for 18 h at room temperature. The solvent was evaporated in vacuo to yield a solid foam which was crushed to a fine powder using a spatulum. From the reaction vessel 350 mg of a colorless powder were transferred that analyzed to 10 + [OTs] − containing 2 /3 equivalent of CDCl3 as lattice solvent (0.36 mmol, 82%).
Crystals suitable for SCXRD analysis were obtained from a product fraction containing stoichiometric amounts of CH2Cl2 dissolved in a mixture of CDCl3, THF, and Et2O (1:1:1) that had been stored at −30 °C for one day.

2.) Experimental Details -Catalysis Study
General procedure for the catalytic CO2 reduction with borane dimethylsulfide complex In a glovebox workstation a Schlenk flask was charged with the (pre)catalyst, the naphthalene standard (only in case of 9 + [OTs] − ) and CDCl3 (the conversions were driven on a 2-4 mL scale with regard to the reaction-mixture volume, the relative amounts can be taken from Table 1 of the main article). At the Schlenk line borane dimethylsulfide complex was added and the resulting mixture frozen at liquid nitrogen temperature before setting the flask to vacuum. The liquid nitrogen bath was removed, the flask was pressurized with CO2 (1.0-1.1 bar) and a water bath (fresh from "non-warm" tap) was applied for controlled thawing of the mixture with stirring. After temperature accomodation the water bath was removed and the reaction driven at room temperature with constant CO2 pressure.               . This confirms the sufficient stability of 1b in CDCl3. Reference for the compound's proton NMR shifts in C6D6 can be found in the literature [S2] .

General procedure for the catalytic CO2 reduction with catecholborane (HBcat)
In a glovebox workstation a Schlenk vessel was charged with the (pre)catalyst, the naphthalene standard (only in selected instances), catecholborane and CDCl3 in this order (in some instances the order of addition of catecholborane and CDCl3 was inverse; the conversions were driven on a 2-4 mL scale with regard to the reaction-mixture volume, the relative amounts can be taken from Table 2 of the main article).
At the Schlenk line borane dimethylsulfide complex was added and the resulting mixture frozen at liquid nitrogen temperature before setting the flask to vacuum. The liquid nitrogen bath was removed, the flask was pressurized with CO2 (1.0-1.1 bar) and a water bath (fresh from "non-warm" tap) was applied for controlled thawing of the mixture with stirring. After temperature accomodation the water bath was removed and the reaction driven at room temperature with constant CO2 pressure.  The catalytic conversion was carried out as detailed for the use of catecholborane as a reductant using the "free" bisNHI 12 as a (pre)catalyst and H-BBN as a hydride source. See the main article for further interpretation.

3.) X-ray Crystallographic Details
General Considerations: Data were collected on a single crystal x-ray diffractometer equipped with a CMOS detector (Bruker APEX III, κ-CMOS), an IMS microsource with MoKα radiation (λ = 0.71073 Å) and a Helios optic using the APEX3 software package (Compounds 4, 7 + [Al(OC(CF3)3)4] − , 9 + [OTs] − ) or on a single crystal x-ray diffractometer equipped with a CCD detector (Rigaku Oxford Diffraction, SuperNova, Atlas), a micro-focus sealed tube with CuKα radiation (λ = 1.54184 Å) and a mirror monochromator using the CrysAlisPro software package (Compounds 6, 11 + [OTs] − ). [S5,S6] The measurements were performed on single crystals coated with perfluorinated polyether oil. The crystals were fixed on top of a kapton micro sampler and frozen under a stream of cold nitrogen. A matrix scan was used to determine the initial lattice parameters. Reflections were corrected for Lorentz and polarisation effects, scan speed, and background using the CrysAlisPro software package or SAINT. [S6,S7] Absorption correction, including odd and even ordered spherical harmonics was performed using the CrysAlisPro software package or SADABS. [S6,S7] Space group assignment was based upon systematic absences, E statistics, and successful refinement of the structure. The structures were solved using SHELXT with the aid of successive difference Fourier maps, and were refined against all data using SHELXL-2014 in conjunction with SHELXLE. [S8,S9,S10] Hydrogen atoms -except if bound to Al or B -were calculated in ideal positions as follows: Methyl hydrogen atoms were refined as part of rigid rotating groups, with a C-H distance of 0.98 Å and Uiso(H) = 1.5·Ueq(C). Other H atoms were placed in calculated positions and refined using a riding model, with methylene and aromatic C-H distances of 0.99 Å and 0.95 Å, respectively, and other C-H distances of 1.00 Å, all with Uiso(H) = 1.2·Ueq(C). In the case of 7, the Al-H distances were constrained at 1.51 Å. Nonhydrogen atoms were refined with anisotropic displacement parameters. Full-matrix least-squares refinements were carried out by minimizing Σw(Fo 2 -Fc 2 ) 2 with the SHELXL weighting scheme. [S8] Neutral atom scattering factors for all atoms and anomalous dispersion corrections for the non-hydrogen atoms were taken from International Tables for Crystallography. [S11] A split layer refinement was used to treat with disordered groups and additional restraints on distances, angles and anisotropic displacements parameters were employed to achieve convergence within physically meaningful limits. The unit cell of 9 + [OTs] − contains two molecules of tetrahydrofuran and two molecules of pentane, the unit cell of 11 + [OTs] − contains four molecules of dichloromethane with partial tetrahydrofuran occupancy and the unit cell of 6 contains twelve molecules of chloroform; these were treated as a diffuse contribution to the overall scattering without specific atom positions using the PLATON/SQUEEZE procedure. [S12] Images of the crystal structure were generated with Mercury and PLATON. [S13,S14] The supplementary crystallographic data for this publication can be acquired free of charge by The Cambridge Crystallographic Data Centre via CCDC Reference Numbers: 1940504 to 1940508 Figure S39. Molecular structure of 7 + (in 7 + [Al(OR F )4] − ) in the solid state as derived from SCXRD analysis (thermal ellipsoids are depicted at the 30% level). Dip groups are depicted as wireframe model. Structure refinement afforded two independent molecules in the asymmetric unit and only one is shown. One Al center of each molecule bears an exocyclic formate group as a minor occupied site instead of a hydride (not shown, occupancy factor for exocyclic formate = ca. 20%). Hydrogen atoms omitted except at formate.