Efficient additive-free formic acid dehydrogenation with a NNN–ruthenium complex

A new ruthenium complex containing a pyridylidene amine-based NNN ligand was developed as a catalyst precursor for formic acid dehydrogenation, which, as a rare example, does not require basic additives to display high activity (TOF ∼10 000 h−1). Conveniently, the complex is air-stable, but sensitive to light. Mechanistic investigations using UV-vis and NMR spectroscopic monitoring correlated with gas evolution profiles indicate rapid and reversible protonation of the central nitrogen of the NNN ligand as key step of catalyst activation, followed by an associative step for formic acid dehydrogenation.


Synthesis of complex 2b
Ligand precursor 1a (400 mg, 0.84 mmol), was stirred in a biphasic mixture of aqueous KOH (1 M, 10 mL) and CH2Cl2 (20 mL) for 15 min. The CH2Cl2 phase was separated and washed with 0.1 M KOH solution (20 mL) followed by water (20 mL). The organic phase was dried over Mg2SO4, filtered, and all volatiles were removed under reduced pressure. The resulting oil was layered with pentane for 1 h to afford a solid, which was washed with pentane (2 ´ 20 mL) and dried in vacuo to yield a yellow waxy solid (272 mg, 99%). This solid was directly used for further synthesis without full characterisation.

Reversible deprotonation
Na2CO3 (ca. 8 mg, 75 µmol) was added and the suspension mixed. The color of the solution changed back from yellow to red. The 1 H NMR spectrum reverted to that of 2a.

General catalytic procedure
A stock solution of complex 2a (7.53 mg in 5 mL CH3CN, 10.6 mM) was prepared under exclusion of light and kept protected from light. An aliquot of the solution (1 mL, 2,13 µmol) was transferred into a 10 mL two-neck round-bottom flask and, under exclusion of light, all volatiles were evaporated under reduced pressure. The flask was equipped with a magnetic stirrer bar and a condenser. The condenser was connected to a three-way valve, connected to a Schlenk line and to a BlueV count volumetric gas measurement device. The round bottom flask and condenser were put under N2 atmosphere. Under N2 pressure the valve is opened to the volumetric counter and the tubing flushed with nitrogen. Degassed solvent (1.2 mL) was added to the round bottom flask, the valve to the Schlenk line was closed and the mixture submerged into a preheated oil bath for 10 min. Degassed formic acid (40 µl, 98% purity) was injected (time = 0), and gas evolution was monitored with the BlueV count device.

Gas analysis
Gas generated by FA dehydrogenation catalysis was analyzed by gas chromatography (GC 8610C device, SRI Instruments, USA; equipped with a packed Hayesep D column). Argon (99.9999 %, Carba Gas, Switzerland) was used as a carrier gas. CO2 and CO quantification was carried out with a flame ionization detector (FID) assembled with a methanizer (5 ppm detection limits for CO and CO2). H2 was quantified by a thermal conductivity detector (TCD).      80°C 70°C S19 Figure S20. Arrhenius plot (top) and Eyring plot (bottom), form rates deduced from temperature dependent catalysis (see slopes in figure S19) The activation energy Ea was calculated from the slope of the Arrhenius plot:

NMR investigation of FA dehydrogenation reaction
General considerations: The following experiments are prepared under N2 atmosphere. FA addition was however carried out in close proximity to the NMR spectrometer and the tube was opened to air during the addition.   Figure S26).        (2) K. Data reduction was performed using the CrysAlisPro 2 program. The intensities were corrected for Lorentz and polarization effects, and an absorption correction based on the multi-scan method using SCALE3 ABSPACK in CrysAlisPro 2 was applied. Data collection and refinement parameters are given in Table S2. The structures were solved by intrinsic phasing using SHELXT 3 , which revealed the positions of all non-hydrogen atoms. All non-hydrogen atoms were refined anisotropically. H-atoms were assigned in geometrically calculated positions and refined using a riding model where each Hatom was assigned a fixed isotropic displacement parameter with a value equal to 1.2Ueq of its parent atom (1.5Ueq for methyl groups). Refinement of the structures was carried out on F 2 using full-matrix least-squares procedures, which minimized the function Σw(Fo 2 -Fc 2 ) 2 . The weighting schemes were based on counting statistics and included a factor to downweight the intense reflections. All calculations were performed using the SHELXL-2014/7 4 program in OLEX2 5 . The structure of 2a was refined as a two-component twin, where the twin law corresponds to a rotation of -180 degrees around [-0.23 -0.00 0.97] (reciprocal) or [0.00 0.00 1.00] (direct). A disorder model was included for triflate, where the occupancy of the disorder components is refined using a free variable. The occupancies of both components together are restrained to 100%. For [2a-H] + , a disorder model was used for one of the PF6 units and for the CH2Cl2. There is another disordered CH2Cl2 molecule, which could not be modeled and therefore a mask was used to include the contribution of the electron density located in the void area into the calculated structure factors. Further crystallographic details are compiled in Table S2. Crystallographic data for both structures reported in this paper have been deposited with the Cambridge Crystallographic Data Centre (CCDC) as supplementary publication numbers 2251207 (2a), and 2251208 ([2a-H] + ).  (8)