Reversible Dihydrogen Activation and Catalytic Transfer Hydrogenation with Iminophosphinoyl‐Tethered Ruthenium Carbene Complexes

In the last decades, metal ligand cooperativity has intensively been used as tool in homogeneous catalysis for the development of potent catalysts for a variety of applications, above all hydrogenation and transfer hydrogenation reactions. Herein, we report on the synthesis of iminophosphinoyl‐tethered ruthenium carbene complexes and their use in the activation of dihydrogen and the catalytic transfer hydrogenation of p‐fluoroacetophenone. Variation of the N‐substituent to an electron‐withdrawing p‐nitrophenyl group enabled the reversible activation of elemental hydrogen via a 1,2‐addition across the metal carbon double bond both in solution and in solid state. In comparison to the thiophosphinoyl‐tethered carbene complex the presented iminophosphinoyl systems showed a considerably increased catalytic activity in transfer hydrogenation reactions, thus highlighting the tunability of the cooperative behavior of nucleophilic carbene complexes.


Introduction
Transfer hydrogenation is an important tool for the synthesis of alcohols from ketones and is used also in the chemical and pharmaceutical industry as an alternative to traditional hydrogenation catalysis, since the use of hydrogen transfer reagents such as 2-propanol circumvents the necessity of hydrogen gas at higher pressures, often unwanted in large scale applications. [1]Many catalysts applied in transfer hydrogenations make use of so called non-innocent ligands, which -in contrast to typical spectator ligands used to control the steric and electronic properties of the metal center -actively take part in the chemical transformation.This interplay between metal and ligand is also referred to as metal ligand cooperativity. [2]Hereby, the ligand usually takes part in the dehydrogenation process by accepting the proton of the hydrogen source, while the hydride is delivered to the metal center.Such a mechanism is for example observed in the wellknown chiral ruthenium catalyst A (Figure 1), which has been reported by Noyori and coworkers in the 1990 s [3] and constantly been improved in the following years. [4,5,6]Here, the dehydrogenation leads to the transformation of the amido to an amino ligand upon protonation, a concept which was utilized in several other catalysts afterwards. [7]Another important mechanism of metal ligand cooperativity applied in transfer hydrogenations has been established by Milstein and coworkers with PNN pincer complexes such as B. [8] Here, an aromatization-dearomatization mechanism enabled a variety of bond activation reactions and a number of catalytic transformations.
Besides these landmark examples for bifunctional catalysts, various other types of complexes have been reported, which make use of different cooperation mechanisms, including direct additions across metal ligand bonds (e. g. on the FeÀ N linkage in C [9] ) or via active participation of ligand sites more distal to the metal center. [10]Also carbene ligands have been found to actively engage in bond activation reactions through splitting of EÀ H bonds by addition across the M=C bond, thus leading to a transition from a carbene to an alkyl species (M=CR 2 !MÀ C(H)R 2 ). [11]The first successful bond activation reactions with carbene ligands were realized by Shaw and coworkers with aliphatic PCP pincer ligands. [12]This system was later further improved by Ozerov, [13] Piers [14] (D) and others by using aromatic linkers to prevent β-hydride elimination processes in the ligand backbone. [15]ur group has focused on the use of methandiide-derived late transition metal carbene complexes such as 1 and 2, which feature a highly polarized M=C carbon linkage due to the strong electron withdrawing groups at the carbenic carbon center. [16,17,18]This enabled the facile activation of numerous element hydrogen bonds and small molecules.First investigations on transfer hydrogenation with methandiide-derived carbene complexes focused on the use of the thiophosphinoyltethered carbene complex 1, [16,19] which however showed only low activity in the transfer hydrogenation of ketones with 2propanol as hydrogen source.This could be explained by the limited stability of the complex at elevated temperatures and the high polarity of the M=C bond, which results in highly exergonic activations limiting reversibility and the transfer of the activated substrates. [20]In the past, iminophosphinoylfunctionalized carbene complexes of type 2 (Figure 1) have been shown to enable more selective and also reversible activation reactions compared to carbene complex 1 due to an overall less polarized and therefore more stable Ru=C bond. [18,21,22]For example, complex 2 b showed selective PÀ H bond activation with secondary phosphines by addition across the M=C bond, while compound 1 only gave rise to complex reaction mixtures.We envisioned that this stabilization of the Ru=C linkage would also influence the performance in hydrogen activations and transfer hydrogenation reactions, as the release or transfer of dihydrogen should be facilitated by a more stable Ru=C double bond.Herein, we report on our results in the transfer hydrogenation of p-fluoroacetophenone with iminophosphinoyl-tethered carbene complexes.

Synthesis of the iminophosphinoyl-substituted carbene complex 2 c
From our previously reported iminophosphinoyl-substituted carbene complexes 2 a and 2 b, only the nitro-phenyl derivative 2 b appeared to be a suitable candidate for transfer hydrogenation.The silylated complex 2 a was shown to be susceptible to alcoholysis of the NÀ Si bond and therefore would not be sufficiently stable under the catalysis conditions. [17]To gain further insight into the influences of different iminophosphinoyl moieties on the reactivity of the corresponding carbene complexes, we thus decided to additionally prepare the new carbene complex 2 c (Scheme 1) with a less electron withdrawing phenyl substituent.Complex 2 c was synthesized via a stepwise protocol as outlined in Scheme 1.At first, the phosphonium salt 4 was synthesized by reacting phosphine bromide 3 [23] with aniline in the presence of trihexyl amine, giving 4 as colorless solid in 88 % yield.The bromide was subsequently deprotonated using n-butyllithium delivering iminophosphorane 5 as colorless solid in an excellent 97 % yield.Interestingly, the 31 P{ 1 H} NMR spectrum of 5 in dichloromethane shows two peaks in a ratio of 0.12 : 1 at δ P = 23.1 and À 10.1 ppm.Careful analysis of the corresponding 1 H and 13 C{ 1 H} NMR spectra revealed these signals to originate from an equilibrium between the iminophosphorane 5 and the ylide tautomer 5' present in solution (Scheme 2).Such an equilibrium between different tautomers has also been reported for bis(iminophosphinoyl)methanide systems. [24]The signal at δ P = À 10.1 ppm in the 31 P{ 1 H} NMR spectrum can be attributed to the iminophosphorane tautomer 5, while the downfield shifted signal corresponds to ylide 5'.The protons at the carbon bridge of 5 give rise to a doublet in the 1 H NMR spectrum at δ H = 6.50 ppm ( 2 J HP = 9.6 Hz), while two separate signals for the CH and NH groups can be observed for the aminophosphonium tautomer 5' at δ H = 3.19 ppm (doublet; 2 J HP = 20.0Hz) and 5.50 ppm, respectively.
To conveniently transfer the ligand via salt metathesis onto a metal center, we next targeted the isolation of its alkali metal salt.Deprotonation of 5 with sodium hydride gave methanide 6 as colorless solid in 98 % yield.Successful deprotonation is evidenced by a single new signal in the 31 P{ 1 H} NMR spectrum at δ P = 0.10 ppm and a highfield shifted signal for the methanide proton in the 1 H NMR spectrum at δ H = 2.71 ppm with a coupling constant of 2 J HP = 11.7 Hz.Iminophosphorane 5 and its sodium salt 6 could additionally be characterized by X-ray crystallography (Figures S43 and Figure 2; for crystallographic details of the iminophosphorane see chapters 3 and 3.1 in the supporting information).
Due to the overall low solubility of methanide 6, crystallization was conducted by dissolving 6 in hot THF with an equimolar amount of 15-crown-5 and slowly cooling the mixture to RT. 6 crystallizes in the orthorhombic space group P2 1 2 1 2 1 .The sodium cation shows no contact to the carbon bridge and is coordinated solely by the crown ether and one oxygen of the sulfonyl moiety.The sum of angles at the central carbon atom C1 amounts to 359(6)°, which is in line with a sp 2 hybridization and a planar geometry around the carbon atom.Reaction of methanide 6 with half an equivalent of [(pcymene)RuCl 2 ] 2 delivered chloro complex 7 as orange solid in 90 % yield.7 is characterized by a singlet in the 31 P{ 1 H} NMR spectrum at δ P = 45.5 ppm and by doublets at δ H = 3.89 ppm ( 2 J HP = 5.57Hz) and δ C = 34.4ppm ( 1 J CP = 67.4Hz) in the 1 H and 13 C{ 1 H} NMR spectrum, respectively, for the CH moiety.Chloro complex 7 crystallizes in the monoclinic space group P2 1 /c (Figure 3), with the ruthenium center being coordinated by the nitrogen of the imino moiety and the methanide carbon.The Ru1À N1 bond length of 2.127(2) Å is slightly shorter than in the p-nitroaniline substituted congener (2.148(2) Å), which is well in line with the increased electron donating ability of the aniline moiety. [18]he formation of the desired carbene complex 2 c was finally accomplished by dehydrohalogenation of 7 with sodium tert-butoxide, allowing for the isolation of 2 c as blue solid in 94 % yield (absorption maximum in the UV-VIS spectrum at 607.0 nm; see Figure S42 in the supporting information).2 c is characterized by a singlet in the 31 P{ 1 H} NMR spectrum at δ P = 60.1 ppm and the absence of a signal for any proton at the carbon bridge in the 1 H NMR spectrum.Furthermore, a strongly low field shifted doublet at δ C = 134.0ppm ( 1 J CP = 71.1 Hz) could be assigned to the carbenic carbon atom.The crystal structure of 2 c (monoclinic space group P2 1 /n; Figure 3) reveals a planar coordination geometry around the central carbon atom C1 (sum of angles 359.3(2)°), which is well in line with the formation of a ruthenium carbene species with a metal carbon double bond.Since the reported structure of the 4-nitroaniline substituted carbene complex 2 b was not of sufficient quality for a discussion of bond length and angles only a comparison with the thiophosphinoyl-tethered complex 1 can be made (Figure 1). [16,18]The Ru1-C1 bond length in 2 c amounts to 1.956(2) Å, which is slightly shorter than in 1 (1.965(2)Å), arguing for a more pronounced double bond character.

Cooperative activation of elemental hydrogen and dehydrogenation of 2-propanol
With the iminophosphinoyl-tethered ruthenium carbene complexes at hand, we next addressed their ability in hydrogen activation.To this end, both complexes 2 b and 2 c were subjected to an atmosphere of dihydrogen at room temperature (Scheme 3) and the reaction progress monitored by 31 P { 1 H} and 1 H NMR spectroscopy (see Figures S31 for 2 b and Figures S34 and S35 for 2 c).In the reaction of carbene complex 2 b, the selective and full conversion to a new compound characterized by a 31 P{ 1 H} NMR signal at δ P = 37.3 ppm could be observed over the course of 7 days.This compound could be unambiguously identified as the hydrido complex 8 b formed by the expected 1,2-addition of H 2 across the former Ru=C double bond.In the 1 H NMR spectrum, a broad singlet at δ H = 3.71 ppm could be assigned to the proton at the carbon bridge, while the ruthenium bound hydrogen atom gives rise to a singlet in the hydride region at À 5.37 ppm.The methanide carbon atom is characterized by a doublet at δ C = 26.7 ppm ( 1 J CP = 67.2Hz) in the 13 C{ 1 H} NMR spectrum.
To our surprise, removal of the solvent under reduced pressure led to the partial back reaction to carbene complex 2 b.This reversibility of the H 2 activation was even observed in the solid state.Thus, isolation of compound 2 b by means of precipitation and drying of the obtained solid only gave a mixture of 2 b and 8 b in a ratio of 1.22 : 1.To further prove the existence of an equilibrium, a series of NMR experiments was performed.At first, carbene complex 2 b was reacted with elemental hydrogen in a J-Young NMR tube until the hydrogenation reaction was almost complete.Excess elemental hydrogen was then removed, and NMR spectra were taken at once and after shaking for 16 h at room temperature.The obtained 31 P{ 1 H} NMR and 1H NMR spectra showed a slight shift of the ratio between carbene complex 2 b and hydride complex 8 b in favor of the carbene complex (See Figures S32 and S33 in the Supporting Information).Further heating of the reaction mixture for 2 hours to 30, 40 and 50 °C respectively, shifted the equilibrium even further back to the unreacted carbene complex (to a final ratio of 2 b to 8 b of 0.3 : 1 based on the integration of the 1 H NMR spectrum).
Since release of elemental hydrogen was possible from solid 2 b, also the activation of hydrogen was attempted directly with the solid material, i. e. in the absence of any additional solvent.Indeed, exposure of carbene complex 2 b as a solid material to an atmosphere of dihydrogen led to the selective formation of hydrido complex 8 b as evidenced by the NMR spectra of the obtained compound and a color change from dark green to dark brown.Both, the solid hydrido complex 8 b as well as solutions in toluene also showed fast dihydrogen release upon heating to 75 °C.Due to this high sensitivity of hydrido complex 8 b towards temperature and pressure changes, its NMR characterization could only be conducted on an in situ prepared sample.Overall, this reversibility furthermore confirms the stability of the RuÀ C double bond in carbene complex 2 b.None of the prior synthesized methandiide-derived carbene complexes showed reversibility of the hydrogen activation even at elevated temperatures.Instead, the highly polar M=C bond resulted in a high thermodynamic preference of the activation products with no back reaction being possible at mild conditions. [16]he reaction of the phenyl-substituted carbene complex 2 c with elemental hydrogen likewise led to the formation of the desired hydrido complex 8 c over the course of 1 week.Hereby, facile isolation of the product as a yellow solid in 82 % yield was possible by means of solvent removal and subsequent workup, since hydrido complex 8 c underwent no release of dihydrogen at reduced pressure or elevated temperatures.Therefore, 8 c could be characterized by multi nuclear NMR spectroscopy, as well as elemental and X-ray diffraction analysis (Figure 4).In the 31 P{ 1 H} NMR spectrum, 8 c exhibits a singlet at δ P = 30.7 ppm, while the proton at the carbon bridge and the ruthenium bound hydridic hydrogen atom can be detected in the 1 H NMR spectrum as doublet of doublets at δ H = 3.97 ppm ( 2 J HP = 2.19 Hz, 3 J HH = 0.88 Hz) and as singlet at δ H = À 4.93 ppm, respectively.
8 c crystallizes in the monoclinic space group P2 1 /c and the asymmetric unit contains two molecules of the hydrido complex as well as an additional THF solvent molecule.The successful formation of the hydrido complex could unambiguously be confirmed by the presence of the ruthenium bound  hydrogen atom which was localized in the difference Fourier map and refined independently (Ru1-H71 1.54(5) Å).The Ru1-C1 bond length amounts to 2.195(3) Å, which is much longer than the Ru=C bond in carbene complex 2 c (1.9561(19) Å), thus corroborating well with the change from a RuÀ C double bond to a single bond.Likewise, the RuÀ N bond in 8 c slightly elongates upon hydrogenation from 2.118(2) Å in 2 c to 2.144(3) Å.
Interestingly, the hydrogen activation with the phenylsubstituted carbene complex 2 c was less selective than that of 2 b.Upon approaching complete consumption of carbene complex 2 c a new compound started to form, accompanied by the consumption of hydrido complex 8 c.This new complex is characterized by a 31 P{ 1 H} NMR shift at δ P = 29.1 ppm very close to the signal of hydrido complex 8 c (see Figure S34 for time dependent overlay of 31 P{ 1 H} spectra).The transformation proceeds very slowly and reached a ratio of 1:0.16 between hydrido complex 8 c and the new compound after three weeks of reaction time at RT. Stirring the reaction at 40 °C for additional 3 weeks shifted the ratio to 1 : 1.2, but at that time additional decomposition to an unidentified product started to occur and prolonged reaction time did not lead to further changes in the product ratio.In depth analysis of the obtained 31 P{ 1 H} and 1 H NMR spectra as well as additional 2D NMR spectra (Figures S36 and S37 in the Supporting Information) led to the conclusion, that the newly formed compound is the corresponding trans-isomer 8 c-trans of hydrido complex 8 c (Scheme 4).The formation of such a trans-isomer has never been observed in any EÀ H bond activation reaction with carbene complexes 1 or 2. [25] A similar isomerization was observed by Piers and coworkers during the synthesis of rhodium PCP pincer complexes, where the equilibrium is achieved through a reductive elimination and activation of a CÀ H bond at the rhodium center. [26]ydrido complex 8 c-trans gives rise to a new signal for the hydridic hydrogen atom in the 1 H NMR spectrum at δ H =-3.07 ppm.The signal for the hydrogen atom at the methylene bridge appears at the same shift as the corresponding proton in 8 c (δ H = 3.97 ppm) leading to an overlap of both signals. 1 H-1 H-COSY as well as 13 C-1 H-HSQC spectra (Figures S36 and S37 in the Supporting Information) validated this assignment.Unfortunately, no confirmation via X-ray diffraction analysis could be conducted, as no suitable crystals could be obtained from the reaction mixture.Therefore, to further solidify the claim of a cis trans isomerization, 1D selective gradient NOESY spectra with selective excitation of the corresponding frequencies of the hydridic hydrogen atoms were performed (Figures S38 and S39 in the Supporting Information).These showed a strong interaction of the hydridic hydrogen atom in 8 c with the cisproton at the methylene bridge, while in 8 c-trans only a very weak interaction could be observed, which is well in line with the assignment of the corresponding signals to a cis and a trans isomer.DFT studies on the PBE0/MWB28/def2tzvp level of theory revealed that 8 c-trans is only 1.66 kJ/mol more thermodynamically stable than 8 c which is in agreement with the shift of the equilibrium in slight favor of the trans-isomer after a prolonged reaction time.Thus, we conclude that the cisproduct 8 c is initially formed as a consequence of the kinetically favored cis addition, but isomerizes to its trans congener 8 c-trans until reaching equilibrium.Such an isomerization was not observed for the nitro-phenyl compound 8 b, presumably also because of the reversibility of H 2 activation with 8 b.This furthermore highlights the impact of the electronic properties of the iminophosphinoyl tether on the hydrogenation reactivity.
To evaluate a possible use of both carbene complexes 2 b and 2 c in transfer hydrogenation catalysis, dehydrogenation reactions with an excess of 2-propanol in toluene at 75 °C were performed (Scheme 5).
The reactions were monitored by in situ 31 P{ 1 H} and 1 H NMR spectroscopy and confirmed the formation of the corresponding hydrido complexes 8 (see Figures S40 and S41 in the Supporting Information) along with acetone.Prolonged reaction times led to decomposition into multiple unknown compounds.It is important to note, that, due to the need for elevated reaction temperatures the hydrido complex 8 c transformed into its trans isomer.After 4 h, a similar ratio between 8 c and 8 c-trans as obtained by the reaction of 2 c with elemental hydrogen at higher temperatures and prolonged reaction times was recorded.

Catalytic transfer hydrogenation of p-fluoroacetophenone
With the new carbene complexes 2 b and 2 c in hand and the preliminary investigations on the activation of elemental hydrogen and the dehydrogenation of 2-propanol showing promising results, the catalytic transfer hydrogenation of p-fluoroacetophenone was investigated, particularly focusing on the influence of the ligand substitution pattern on the reactivity of the carbene complexes.p-fluoroacetophenone was chosen as substrate as it enables determination of yields via NMR spectroscopy as well as GC/FID analysis.Reaction conditions were adapted from our previous report on the transfer hydro- genation with the thiophosphoryl-tethered carbene complex 1. [16] In a general protocol, transfer hydrogenations were performed at 75 °C in pure 2-propanol which acts as solvent as well as hydrogen source.At first, various catalyst loadings without an additional alkali metal base as additive were applied (Table 1, entries 1-6).Here, yields between 16 and 40 % conversion depending on the catalyst loading could be achieved with the phenyl substituted carbene complex 2 c, while the p-nitroanilin tethered carbene complex 2 b reached conversions between 12 % and 37 %.Further transfer hydrogenations were performed with alkali metal bases as additives (5 mol %) as these are commonly used as promoters in transfer hydrogenation reactions. [27]With potassium tert-butoxide, 43 % conversion could be reached with 0.5 mol % of carbene complex 2 c as catalyst, while carbene complex 2 b achieved 70 % yield.The results of the 4-nitroaniline substituted carbene complex 2 b are remarkable, as the conversion could nearly be doubled compared to the thiophosphinoyl-tethered catalyst 1 under comparable reaction conditions (36 % yield with carbene complex 1, albeit using acetophenone as substrate). [16]Similar yields could be achieved with KHMDS as additive (46 % and 72 % respectively).Surprisingly, the use of the sodium salts (NaOtBu and NaHMDS) led to a further increase in catalytic activity for the phenyl substituted carbene complex 2 c (67 % and 60 % respectively), while the results with carbene complex 2 b remained largely unaffected by the choice of alkali metal (72 % with NaOtBu and 68 % with NaHMDS).
Additionally, a set of catalysis experiments with triphenylphosphine as additive was performed as our previous investigations revealed this additive to considerably increase the activity of the catalytic system. [16]Here, yields of 89 % (complex 2 c) and 90 % (complex 2 b) could be reached.To exclude any ligand exchange reactions in the complexes when treated with PPh 3 , complexes 2 b and 2 c were treated with an excess of PPh 3 in THF at 50 °C over 18 hours.In both cases, no exchange of the cymene ligand by triphenyl phosphine could be observed.
To compare the new catalysts with the previously reported thiophosphinoyl carbene complex 1 (Figure 1), a set of catalysis reactions with conditions adapted from the previous report were performed, [16] using acetophenone as substrate and KOtBu as base (Table 1, Entry 9-14, and Table S2 in the Supporting Information).To our delight, both carbene complexes achieved higher conversions in the catalysis reaction at 75 °C with especially complex 2 b showing significant improved yields.Thus, 2 b gave 80 % conversion with 0.53 mol % catalyst loading (55 % yield with complex 1) and 71 % yield at 0.34 mol % catalyst loading, while complex 1 only reached 55 and 51 % yield under the same conditions.Interestingly, increasing the reaction temperature to 90 °C did not further improve the yield for both iminophosphinoyl-tethered carbene complexes, but led to a slight decrease by approx.5 % (see Table S2).This (and the fact that prolonged reaction time did not lead to a significant improvement in yield; see Table S2) let us conclude that especially the iminophosphinoyl complex 2 b is faster in the transfer hydrogenation catalysis than carbene complex 1 but also less robust due to the overall weaker coordination of the iminophosphinoyl moiety to the metal center compared to the thiophosphinoyl moiety.
This impressively shows the potential of careful ligand design in the chemistry of methandiide-derived carbene complexes for generating potent catalysts operating via metal ligand cooperativity.Nonetheless, the iminophosphinoyl systems are not yet competitive with the state-of-the-art transfer hydrogenation catalysts confirming the need for further ligand optimization to reach higher conversions at shorter reaction times and lower catalyst loadings.

Conclusions
In summary, we have examined the use of iminophosphinoyltethered ruthenium carbene complexes in the activation of elemental hydrogen and the catalytic transfer hydrogenation of (p-fluoro)acetophenone.The phenyl and nitro-phenyl-substituted carbene complexes were both able to activate elemental hydrogen, with the p-nitroaniline-substituted system 2 b even showing reversible H 2 -activation by releasing hydrogen upon applying vacuum or elevated temperatures.Such a reversibility has not been reported before for the dihydrogen activation with carbene complexes of this type.Furthermore, both carbene complexes showed increased catalytic activities in the transfer hydrogenation reaction compared to the previously examined thiophosphinoyl-tethered carbene complex 1.This impressively demonstrates the influence of the ligand design on the reactivity of the corresponding carbene complexes.Manipulation of the electronics and hence the stability of the metal carbon double bond directly influences the thermodynamics of the activation process and hence a possible transfer and release.The obtained results offer valuable insights for the future design of such bifunctional catalysts, aiming to enhance reversible activation reactions and pave the way for more efficient catalytic applications.UV-VIS-spectra were recorded on a Shimadzu UV-1900i.To enable measurement of sensitive compounds inside a glovebox, the spectrometer was fitted with a fiberoptic interface (Hellma), optical fiber (Hellma, 6 m in total) and external cuvette holder (B&W Tek).Measurement and processing details for individual spectra can be extracted from the corresponding tables in the supporting information.GC analyses were carried out using an HP-5 capillary column (Phenyl methyl siloxane, 30 m×320×0.25, 100/2.3-30-300/3, 2 min at 60 °C, heating rate 30 °C/min, 3 or 10 min at 300 °C).Yields were determined by GC-FID using hexadecane as internal standard.All reagents were purchased from Sigma-Aldrich, ABCR, Acros Organics or TCI Chemicals and used without further purification.Phosphine bromide 3 [23] was synthesized according to previous reports.
Reaction protocol for the dehydrogenation of iso-propanole.17.0 mg (25.6 μmol) of carbene complex 2 b or 18 mg (25.4 μmol) of carbene complex 2 c were dissolved in 0.5 mL toluene in a J-Young NMR tube.0.1 mL iso-propanol were added, and the mixtures heated to 75 °C. 31

Reaction protocol for the catalytic transfer hydrogenation of p-fluoroacetophenone
The reactions were carried out in 1 mL GC-Vials with teflon-coated stir bar (reactions without additive) or 8 mL screw cap vials with teflon-coated stir bar and septum cap (reactions with additive).
The reactions without additive were carried out in 1 mL of stock solution prepared from 5.65 g (40.9 mmol) p-fluoroacetophenone and 2.31 g (10.2 mmol) hexadecane filled to a volume of 25 mL (1 mL stock solution contained 1.64 mmol of p-fluoroacetophenone).
To prevent solubility issues with the applied additives, the reactions with additive were carried out in 4 mL of stock solution prepared from 5.65 g (40.9 mmol) p fluoroacetophenone and 2.31 g (10.2 mmol) hexadecane filled to a volume of 100 mL (4 mL stock solution contained 1.64 mmol of p-fluoroacetophenone).
The corresponding ruthenium complexes and (if present) additive (82.0 μmol, 5 mol %) were weight in the reaction vessel and 1 or 4 mL respectively of the appropriate stock solution were added.The reaction vessels were placed in a preheated heating block (75 °C) and stirred for 24 h.The reactions were then immediately cooled to 0 °C in an ice bath.Small aliquots were removed and filtered through silica and celite with ethyl acetate und subsequently analysed via GC/FID.The yield was determined by integration of the product peak in comparison to hexadecane as standard reagent.

Reaction protocol for the catalytic transfer hydrogenation of acetophenone
The reactions were carried out in 8 mL screw cap vials with tefloncoated stir bar and septum cap.
The reactions were carried out in 4 mL of stock solution prepared from 5.15 g (42.8 mmol) acetophenone and 2.31 g (10.2 mmol) hexadecane filled to a volume of 100 mL (4 mL stock solution contained 1.71 mmol of acetophenone).
The corresponding ruthenium complexes and KOtBu were weight in the reaction vessel and 4 mL of the stock solution were added.The reaction vessels were placed in a preheated heating block (75 °C) and stirred for the designated time.The reactions were then immediately cooled to 0 °C in an ice bath.Small aliquots were removed and filtered through silica and celite with ethyl acetate und subsequently analysed via GC/FID.The yield was determined by integration of the product peak in comparison to hexadecane as standard reagent.
In case of the catalysis experiments at 90 °C, a stock solution containing 1.71 mmol of acetophenone and 1.62 mmol of hexadecane in 2.18 mL ispropanole was used and additional 1.1 mL of toluene were added.
X-ray crystallography.Data collection of the compounds was conducted with an Oxford Synergy or Oxford SuperNova (Cuμsource, Atlas).The structures were solved using direct or dual space methods, refined with the Shelx software package and expanded using Fourier techniques. [28]The crystals of all compounds were mounted in an inert oil (perfluoropolyalkylether). Crystal structure determination was affected at 100 K. Crystallographic data (including structure factors) have been deposited with the Cambridge Crystallographic Data Centre as supplementary publication no.CCDC 2311352-2311356.Copies of the data can be obtained free of charge on application to Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; [fax: (+ 44) 1223-336-033; email: deposit@ccdc.cam.ac.uk].

Scheme 3 .
Scheme 3. Reaction of the carbene complexes 2 b and 2 c with elemental hydrogen.

Table 1 .
Catalytic transfer hydrogenation of p-fluoroacetophenone and acetophenone using carbene complexes 2 b and 2 c as catalysts.
[a] Yields determined via GC/FID (hexadecane as internal standard) and partly controlled by 19 F NMR spectroscopy.[b] Values taken from reference 16.
P{ 1 H} NMR spectra were recorded after 1 h, 4 h and 24 h.(SeeFiguresS40 and S41in the Supporting Information).