P-Stereogenic and Non P-Stereogenic Ir-MaxPHOX in the Asymmetric Hydrogenation of N -Aryl Imines. Isolation and X-ray Analysis of Imine-Iridacycles.

: A small library of Ir-MaxPHOX catalysts has been applied to the asymmetric hydrogenation of N -aryl imines. A structure-activity analysis of the three chiral center MaxPHOX ligand has been performed. Using complex 1b, the hydrogenation of N -aryl imines took place with up to 96% enantiomeric excess at atmospheric pressure of hydrogen and low temperature. The impact of the stereochemical information at the phosphorous center is small with respect the selectivity, but large with respect the catalyst activity. Non P-stereogenic analogs of MaxPHOX were also synthesized and tested but provided lower selectivity. The selectivity observed could be explained by taking into account that the actual catalysts were cyclometalated imine complexes formed in situ . [IrHCl(MaxPHOX)(imine)] complexes 9 and 10 were synthesized and characterized by X-ray crystallography. These complexes, via chloride abstraction, provided the active catalytic species with the same levels of selectivity. Finally, the influence of the counter ion on the catalyst performance was also studied.


INTRODUCTION
Many active pharmaceutical ingredients and agrochemicals contain chiral amines. In this regard, there is considerable interest in developing efficient methods that provide single enantiomers of such compounds, thus avoiding inefficient racemate resolutions.
Metal-catalyzed asymmetric hydrogenation of imines is one of the methods of choice in terms of industrial applicability. 1 While ruthenium has provided excellent results in transfer-hydrogenation reactions, iridium has shown better performance for the direct hydrogenation of imines. 2 In this field, Pfaltz and others have shown that Ir-P,N catalysts provide an excellent platform for the reduction of N-aryl imines. 3 We recently developed the MaxPHOX ligand system, which is built from three fragments-an amino alcohol, an amino acid and a P-stereogenic phosphinous acid ( Figure 1). 4 The Ir-MaxPHOX complexes have shown outstanding selectivity in the hydrogenation of cyclic enamides. 5 One of the key advantages of the MaxPHOX ligand is the structural diversity arising from its possible configurations and substitution patterns, which can be adapted to a specific reaction. To demonstrate this versatility, here we report on the asymmetric hydrogenation of N-aryl imines using the Ir-MaxPHOX catalyst system. To study the influence of the stereogenic center of the phosphorus atom, non-Pstereogenic MaxPHOX ligands were also synthesized and evaluated in the asymmetric hydrogenation. We spotted a catalyst that hydrogenates N-aryl imines with up to 96% enantiomeric excess (ee) at atmospheric pressure of hydrogen (balloon). The putative imine-cyclometalated catalyst derivatives were also isolated, characterized by X-ray analysis and tested in the reaction.

RESULTS AND DISCUSSION
Ligand Synthesis. The iridium catalysts used in the present study are shown in Figure   2. Twelve P-stereogenic MaxPHOX catalysts were employed. These arise from the four possible diastereomeric configurations 1, 2, 3 and 4, plus three distinct oxazoline substitutions, R = Ph, (a), R = iPr (b), and R = tBu (c). The synthesis of these ligands (1ac, 2a-c, 3a-c and 4a-c) involved the coupling of amino acid, amino alcohol and Pstereogenic phosphinous acid fragments and final complexation with an iridium precursor. 5 The corresponding non-P-stereogenic 5a and 5b ligands with a dicyclohexylphosphine moiety were also synthesized in a similar fashion, as shown in Scheme 1. N-Boc protected valine was coupled to either D-or L-valinol. The unprotected amino alcohols 6a-b were then coupled to a phosphinyl mesyl mixed anhydride derived from dicyclohexylphosphinous acid borane. 6 This reaction is highly selective for amine nucleophiles and it provided borane-protected aminophosphines 7a-b. Reaction with SOCl2 and treatment with NaHCO3 gave the protected phosphine-oxazoline ligands 8a-b. Finally, borane deprotection with pyrrolidine, treatment with [Ir(cod)Cl]2, and counter ion exchange with NaBArF afforded non P-stereogenic catalysts 5a and 5b (Scheme 1).
Asymmetric hydrogenation of imines. With the Ir-MaxPHOX catalysts in hand, we proceeded to study their performance in the hydrogenation of acetophenone N-phenyl imine with 1 mol% catalyst loading in DCM at 55 bar of H2 (Table 1). The best results were obtained for catalysts with configurations 1 and 4, the isopropyl group on the oxazoline providing the best results. Thus, complex 1b gave the highest enantiomeric excess (89% ee) and 4b the second best (85% ee) but with the opposite configuration.
Catalysts 1b and 4b would be enantiomers of each other, except that the chirality at phosphorus is inverted. This result suggests that the stereochemical configuration at phosphorus had a limited impact on the selectivity. On the other hand, we observed that the configuration of the chiral center on the backbone of the ligand had a major effect on the selectivity. Catalysts 2b and 3b, while having the same configuration at the P-center and oxazoline with respect 1a and 4b, provided much lower selectivity. We called this phenomenon the "tail effect". We believe this behavior is associated with conformational changes within the Ir-P,N six-membered ring chelate. Finally, non P-stereogenic catalysts 5a and 5b were also tested, providing 57% and 81% ee, respectively. These results demonstrate that although the chiral oxazoline plays the most important role for our system, the presence of additional stereogenic centers both at the phosphine and the tail position is beneficial in terms of selectivity. With the best catalyst in hand (1b), we then studied the influence of hydrogen pressure, solvent and temperature on the reaction outcome (Table 2). Selectivity was relatively insensitive to the reaction pressure. In this regard, a reduction in the pressure from 55 bar to 3 bar produced only a slight increase in enantiomeric excess (from 89 to 90% ee, Table   R P-stereogenic 4b that only differs from the configuration at the P-center and which provided the second best selectivity (85% ee) in the initial catalysts screening at 55 bar, was also tested at atmospheric H2 pressure (Table 2, entry 9). We observed that 4b was considerably less active; since, after 72h, only 8% conversion was achieved. This demonstrates that while the chirality on phosphorous has a small influence on the selectivity it has a great impact on the catalyst activity. Once the reaction parameters had been optimized, we proceeded to assess the scope of the hydrogenation with catalyst 1b with various aryl imines. Reduction of p-and mmethoxyacetophenone N-phenyl imine proceeded with 95% and 92% ee, respectively (Table 3, entries 2 and 3). However, o-methoxyacetophenone N-phenyl imine was hydrogenated with only 28% ee (Table 3, entry 4). This decrease in selectivity is most likely due to a less favorable E/Z ratio of the starting imine. 7 N-Phenyl imines derived from p-chloroacetophenone and methyl 2-naphthyl ketone were both reduced, with 94% ee (Table 3, entries 5 and 6). Imine derived from ethyl phenyl ketone was also reduced, showing a lower enantiomeric excess (74%, Table 3, entry 7). 7 Finally, acetophenone imines derived from other anilines, such as p-anisidine, p-toluidine and 2,3dimethylaniline, were again reduced with excellent selectivity and with 90-95% ee. a) Unless specified all reactions were conducted overnight with 1 mol% catalyst loading in DCM at -20°C with atmospheric H2 pressure (balloon). b) Reaction carried out with 2 mol% of catalyst loading at 3 bar of H2. c) Reaction conducted at room temperature.
Effect of the counter ion on the reaction rate, catalyst intermediates and mechanistic considerations. Since catalyst 1b showed activity and selectivity that paralleled the best P,N-Ir systems reported for the hydrogenation of N-phenyl ketimines, we studied the rate of the reaction and how the counter ion influences the performance of the catalyst. 8 The hydrogenation of acetophenone N-phenyl imine using 1 mol% of 1b-BAr F and 1b-BF 4 at atmospheric pressure of hydrogen was monitored by GC analysis, as shown in Figure 4. Catalyst 1b-BAr F proved extremely active, and 100% conversion was observed after 50 min using only atmospheric H2 pressure. The use of a smaller counter ion, such as BF4, resulted in a slower reaction and full conversion was only achieved after 150 min (2.5 h). Furthermore, the use of BF4 as counter ion had a significant deleterious effect on selectivity since the product amine was obtained with only 41% ee (90% ee using BArF, Figure 3).  Pfaltz and coworkers have recently shown that an imine iridacycle is the actual catalyst when using Ir-P,N complexes. 9 These cyclometallated Ir (III) species form spontaneously upon reaction between the iridium complex and the starting imine under hydrogen pressure. With this in mind, and to further increase knowledge of the true catalytic system, we questioned whether these intermediates are also involved in our Ir-MaxPHOX system.
Complexes 1b   probability. Only the hydrogen atom attached to iridium has been drawn.
To verify whether the isolated imine iridacycles can be used as precatalysts, compound 9 was tested in the hydrogenation of imines. Thus, 1 mol% of 9 was treated with 2 equivalents of NaBArF in order to produce the active catalyst by counter ion exchange (Scheme 3). The use of this mixture as catalyst afforded the reduced product with 60% conversion and 91% ee. Despite the lower conversion observed, that can be justified by an inefficient chloride abstraction reaction, this result strongly suggests that the cationic cyclometallated complex arising from chloride abstraction of 9 is an active species in the catalytic cycle.

Scheme 3:
The use of iridacycle 9 as precatalyst in the hydrogenation of imines.
For several years, the mechanism of imine reduction by iridium catalysts has been the subject of debate. 10 During the preparation of this manuscript, Wiest and coworkers published a theoretical DFT mechanistic study of this process. 11 The authors took into account the findings of Pfaltz and considered cyclometalated iridium(III)-hydride species as the intermediates in the catalytic cycle. The study concluded that the hydrogenation with the cyclometalated precursors proceeds through an outer sphere mechanism, as depicted in a simplified manner in Scheme 4. According to this proposal, the active species in the catalytic cycle would be the cationic hydride-dihydrogen complex I, which, upon proton transfer, activates the amine to form the neutral dihydride complex II. The stereodetermining step is the hydride transfer trans to phosphorus to the activated imine to yield complex III.
Scheme 4: Outer sphere mechanism for the iridium-catalyzed hydrogenation of imines.
Our results are in agreement with this proposal. In this system, the imine ligand and the oxazoline isopropyl substituent are the fragments that surround the hydride transferred to the protonated imine in the stereodetermining step ( Figure 5). Analyzing the selectivity observed with the four stereoisomers of MaxPHOX ligands (1b, 2b, 3b and 4b, Table 1), we can conclude that the chirality on phosphorus and the tail position is highly relevant.
The intriguing question now is that since phosphine and the tail moiety are placed far away from the reacting center, why does the chirality on these fragments have such an impact on selectivity?
Two concurrent effects could explain this behavior. First, small conformational differences on the reacting site may provide a selectivity bias. Observing the pseudoenantiomeric reacting sites ( Figure 5) in complex 9, the N-phenyl of the cyclometalated imine is tilted upwards, while for 10, the N-phenyl ring is almost perpendicular to the Irimine metallacycle. These conformational differences of the N-phenyl group are ultimately governed by the configuration of the phosphine and tail positions and should lead to a difference in selectivity between 9 and 10. 12 A second effect would be the relative stability of complexes IVa/IVb with a vacant coordination site (Scheme 5). It is easy to envisage that the coordination of H2 to either complex IVa or IVb would provide a different selectivity outcome. Therefore, the stereochemical stability of the transient iridium complexes appears to be essential to guarantee high selectivity during hydrogenation. This notion is in agreement with the fact that in our case 9 and 10 have the stereochemistry of IVa, while Pfaltz and co-workers, reported the X-ray crystal structure of an [IrH(THF)(PHOX)(imine)] + complex, where the cyclometalated phenyl group is trans to phosphine and which, upon removal of the solvent molecule, provides a complex with the same stereochemistry at the metal as IVb. 9 We believe that the chirality on phosphorous and the tail positions in the MaxPHOX ligand system are important for stabilizing the arrangement of the ligands around the metal center in the true catalytic species.

Scheme 5:
Equilibrium between complexes IVa and IVb with a vacant coordination site.

CONCLUSIONS
From a small family of P-stereogenic Ir-MaxPHOX precatalysts, we pinpointed complex 1b, which shows activity and selectivity that matches the best Ir-P,N systems in the hydrogenation of acyclic imines. Catalyst 1b hydrogenated N-aryl imines with up to 96% ee at atmospheric pressure of hydrogen and low temperature. Catalyst 4b that only differs from the configuration at the P-center was significantly less active, demonstrating that the chirality on phosphorous has a great impact on the catalyst activity. Non-P- atmosphere. Et2O and CH2Cl2 were dried in a purification system. Other commercially available reagents and solvents were used with no further purification. Thin layer chromatography was carried out using TLC-aluminum sheets with silica gel. Flash chromatography was performed by using an automated chromatographic system with hexane/ethyl acetate or hexane/dichloromethane gradients as eluent unless otherwise stated. NMR spectra were recorded at 23°C on a 400 MHz, 500 MHz or 600 MHz apparatus. 1 H NMR and 13 C NMR spectra were referenced either to relative internal TMS or to residual solvent peaks. 31 P NMR spectra were referenced to phosphoric acid. Optical rotations were measured at room temperature (25°C) and concentration is expressed in g/100 mL. Melting points were determined using a Büchi apparatus and were not corrected. IR spectra were recorded in a FT-IR apparatus. HRMS were recorded in a LTQ-FT spectrometer using the Nanoelectrospray technique. Catalysts 1, 2, 3 and 4 were prepared as described previously. 5 Dicyclohexyl phosphinous acid borane. 6 In a N2 purged round bottom flask 2g of chlorodicyclohexylphosphine (8.59 mmol, 1 eq) were weighed. Anhydrous THF (8 mL) was added and the reaction was cooled to 0 ºC. Then, 815 µL ( ppm.

General procedure for the synthesis of aminophosphines (7a-b). A solution of
dicyclohexylphosphinous acid borane (1 eq) and methansulfonic anhydride (1.2 eq) in CH2Cl2 (0.2M) was cooled to 0 ºC. To this solution, anhydrous NEt3 (2.5 eq) was slowly added, and the mixture was stirred 1 h at 0 ºC. The corresponding amine (1.5 eq) was then added and the solution was stirred overnight at room temperature. Water was added and the mixture was allowed to warm to room temperature. The organic layer was separated and the aqueous phase was extracted twice with CH2Cl2. The combined extracts were washed with brine and concentrated under reduced pressure. Purification by flash chromatography (SiO2, hexanes:EtOAc) yielded the corresponding products as white solids.   General Procedure for the hydrogenation at high H 2 pressure. The corresponding imine (1 eq) and the corresponding catalyst (0.01 eq) were placed along with stirring bar in a glass tube inside a stainless steel high pressure reactor. The reactor was entered into a glove box and deoxygenated anhydrous solvent was added (0.17 M). The reactor was closed, removed from the glove box and connected to a hydrogen manifold. While stirring, the reactor was purged with vacuum-hydrogen cycles and then it was charged at the designated pressure. The hydrogen manifold was unplugged and the mixture was left to stir overnight at room temperature. The reactor was depressurized and the reaction was concentrated under vacuum. The conversion was determined by 1 H NMR. Catalyst was removed by filtration of the crude reaction through a short silica pad with CH2Cl2 and concentrated under vacuum. At this point, yield and enantiomeric excess were determined.

1,1-dicyclohexyl-N-((S)-1-((S)-4-isopropyl-4,5-dihydrooxazol-2-yl
General Procedure for the hydrogenation at atmospheric H 2 pressure. The corresponding imine (1 eq) and the corresponding catalyst (0.01 eq) were placed in a round bottom flask. The flask was purged with N2 and deoxygenated anhydrous solvent was added (0.17 M). The reaction was then set at the desired temperature. While stirring, a H2 filled balloon was connected. Using another needle as a gas exit the round bottom flask was purged with H2 until the solution went from orange to yellow (usually less than 3-4 minutes). The gas exit was removed and the reaction was left stirring overnight. The  13