Iridium-Catalyzed Asymmetric Ring-Opening of Oxabenzonorbornadienes with N-Substituted Piperazine Nucleophiles

Iridium-catalyzed asymmetric ring-opening of oxabenzonorbornadienes with N-substituted piperazines was described. The reaction afforded the corresponding ring-opening products in high yields and moderate enantioselectivities in the presence of 2.5 mol % [Ir(COD)Cl]2 and 5.0 mol % (S)-p-Tol-BINAP. The effects of various chiral bidentate ligands, catalyst loading, solvent, and temperature on the yield and enantioselectivity were also investigated. A plausible mechanism was proposed to account for the formation of the corresponding trans-ring opened products based on the X-ray structure of product 2i.


Results and Discussion
The substrates 1a-1b were readily prepared by Diels-Alder reactions of benzynes with furan according to literature procedures [75]. To understand the nature of the catalytic ring-opening and optimize the reaction conditions, we first chose different chiral bisphosphine ligands, including (S)-BINAP, (R)-(S)-PPF-P t´B u 2 , (S)-p-Tol-BINAP, and (S)-(R)-NMe 2 -PPh 2 -Mandyphos, to validate the catalytic activity of the iridium complexes. Consequently, a more efficient iridium catalyst system for the ring-opening reaction was explored. The different types of chiral ligands reacted with [Ir(COD)Cl] 2 to form iridium complexes to determine the viability of the enantioselectivity (Scheme 1). To probe the iridium-catalyzed asymmetric ring-opening of oxabicyclic alkene 1a with 1-(2-fluorophenyl)piperazine, chiral bisphosphine ligand (S)-p-Tol-BINAP was used and 1 equivalent of NH 4 I was added as the additive. We found that the ring-opening product 2a was obtained in high yield (up to 99%) with moderate enantioselectivity (54% ee) ( Table 1, entry 4). However, the enantiomeric excess value was low (2%-58% ee) when (S)-BINAP and ferrocene bisphosphine ligands were used as the chiral ligands (Table 1, entries 1-3). The enantiomeric excess value was 55% ee when (S)-p-Tol-BINAP was used as the ligand in the presence of 1.25% mol [Ir(COD)Cl] 2 (Table 1, entry 5). Therefore, (S)-p-Tol-BINAP was chosen as the optimized ligand.

Results and Discussion
The substrates 1a-1b were readily prepared by Diels-Alder reactions of benzynes with furan according to literature procedures [75]. To understand the nature of the catalytic ring-opening and optimize the reaction conditions, we first chose different chiral bisphosphine ligands, including (S)-BINAP, (R)-(S)-PPF-P t-Bu2, (S)-p-Tol-BINAP, and (S)-(R)-NMe2-PPh2-Mandyphos, to validate the catalytic activity of the iridium complexes. Consequently, a more efficient iridium catalyst system for the ring-opening reaction was explored. The different types of chiral ligands reacted with [Ir(COD)Cl]2 to form iridium complexes to determine the viability of the enantioselectivity (Scheme 1). To probe the iridium-catalyzed asymmetric ring-opening of oxabicyclic alkene 1a with 1-(2-fluorophenyl)piperazine, chiral bisphosphine ligand (S)-p-Tol-BINAP was used and 1 equivalent of NH4I was added as the additive. We found that the ring-opening product 2a was obtained in high yield (up to 99%) with moderate enantioselectivity (54% ee) ( Table 1, entry 4). However, the enantiomeric excess value was low (2%-58% ee) when (S)-BINAP and ferrocene bisphosphine ligands were used as the chiral ligands (Table 1, entries 1-3). The enantiomeric excess value was 55% ee when (S)-p-Tol-BINAP was used as the ligand in the presence of 1.25% mol [Ir(COD)Cl]2 (Table 1, entry 5). Therefore, (S)-p-Tol-BINAP was chosen as the optimized ligand. Scheme 1. The proposed mechanism for asymmetric ring-opening of 1a with N-substituted piperazines. With the catalyst system consisting of [Ir(COD)Cl]2 and (S)-p-Tol-BINAP in hand, other reaction parameters were further optimized. We screened several commonly used solvents ( Table 2, entries 1-9), the solvent effect on enantioselectivities of ring-opening reaction was remarkable, as seen from Table 2. Table 2. Effects of solvent on the ring-opening a .
The influence of temperature was also investigated in the iridium-catalyzed asymmetric ring-opening reaction of oxabicyclic alkene 1a with 1-(2-fluorophenyl)piperazine. No product was obtained when the reaction mixture was stirred at 25˝C for 12 h (Table 3, entry 1). It was further found that the temperature had little effect on the enantioselectivity (Table 3, entries 2-4). The product 2a was obtained in 50% yield with 40% ee when the reaction mixture was stirred at 50˝C for 12 h ( Table 3, entry 2). Furthermore, the product 2a was obtained in 99% yield with 54% ee when the reaction mixture was stirred at reflux (80˝C) ( Table 3, entry 3). Consequently, the optimum reaction conditions were determined to be as follows: 2.5 mol % [Ir(COD)Cl] 2 , 5.0 mol % (S)-p-Tol-BINAP, 2 equiv. of 1-(2-fluorophenyl)piperazine, and 1 equiv. of NH 4 I to oxabicyclic alkene 1a as additive in THF at 80˝C. Table 3. Effects of the temperature on the ring-opening a .
To further extend the scope of this transformation, the reaction of dimethoxy substituted oxabenzonorbornadiene 1b with various N-substituted piperazines were also examined. It was found that the reactions of 1,4-dihydro-6,7-dimethoxy-1,4-epoxynaphthalene (1b), a less reactive substrate, with N-substituted piperazines offered the desired products in good yields with moderate enantioselectivity ( Table 5, entries 1-9).
Unfortunately, the reaction of 1,4-dihydro-6,7-dimethoxy-1,4-epoxynaphthalene (1b) with 1-(3,4-dichlorophenyl)piperazine afforded the corresponding ring-opening product 3e in a lower yield (47%) with poor enantioselectivity (16% ee) ( Table 5, entry 5).                      (1 equiv.) was then added and stirred for another 10-20 min. Substrate 1a (0.3 mmol, 1 equiv.) was added and the mixture was heated to reflux. N-Substituted piperazine nucleophiles (2 equiv.) were added at the first sign of reflux; b Isolated yield; c ee was determined by HPLC with a Chiralcel OD or AD column.   (1 equiv.) was then added and stirred for another 10-20 min. Substrate 1a (0.3 mmol, 1 equiv.) was added and the mixture was heated to reflux. N-Substituted piperazine nucleophiles (2 equiv.) were added at the first sign of reflux; b Isolated yield; c ee was determined by HPLC with a Chiralcel OD or AD column.   (1 equiv.) was then added and stirred for another 10-20 min. Substrate 1a (0.3 mmol, 1 equiv.) was added and the mixture was heated to reflux. N-Substituted piperazine nucleophiles (2 equiv.) were added at the first sign of reflux; b Isolated yield; c ee was determined by HPLC with a Chiralcel OD or AD column.   (1 equiv.) was then added and stirred for another 10-20 min. Substrate 1a (0.3 mmol, 1 equiv.) was added and the mixture was heated to reflux. N-Substituted piperazine nucleophiles (2 equiv.) were added at the first sign of reflux; b Isolated yield; c ee was determined by HPLC with a Chiralcel OD or AD column.   (1 equiv.) was then added and stirred for another 10-20 min. Substrate 1a (0.3 mmol, 1 equiv.) was added and the mixture was heated to reflux. N-Substituted piperazine nucleophiles (2 equiv.) were added at the first sign of reflux; b Isolated yield; c ee was determined by HPLC with a Chiralcel OD or AD column.               (1 equiv.) was then added and stirred for another 10-20 min. Substrate 1b (0.3 mmol, 1 equiv.) was added and the mixture was heated to reflux. N-Substituted piperazine nucleophiles (2 equiv.) were added at the first sign of reflux; b Isolated yield; c ee was determined by HPLC with a Chiralcel OD or AD column.
The stereochemistry of 1,2-trans ring-opened product 2i was unambiguously confirmed by X-ray crystallography. The single crystal of 2i was achieved by solvent evaporation from a mixture of dichloromethane, petroleum ether and ethyl acetate. Its configuration was assigned as (1S, 2S) and confirmed as 1,2-trans configuration, as shown in Figure 1 (See Supplementary Materials for details). It is obvious that the ring-opening reaction favors the formation of trans-2-N-substituted piperazine 1,2-dihydro-naphthalen-1-ol products.
The stereochemistry of 1,2-trans ring-opened product 2i was unambiguously confirmed by X-ray crystallography. The single crystal of 2i was achieved by solvent evaporation from a mixture of dichloromethane, petroleum ether and ethyl acetate. Its configuration was assigned as (1S, 2S) and confirmed as 1,2-trans configuration, as shown in Figure 1 (See Supplementary Materials for details). It is obvious that the ring-opening reaction favors the formation of trans-2-N-substituted piperazine 1,2-dihydro-naphthalen-1-ol products. Based on our findings above, we propose a mechanism, outlined in Scheme 1. When [Ir(COD)Cl]2 was used as the iridium source, and reacted with (S)-p-Tol-BINAP to form the complex of [Ir(S)-p-Tol-BINAP)I]2 A in the presence of NH4I, which is then cleaved by solvent to give the monomeric iridium complex B. Reversible exo coordination of oxabenzonorbornadiene 1a leads to iridium complex C, followed by oxidative insertion with retention to form a bridgehead C-O bond and produce the π-allyl iridium alkoxide complex D. We further propose that the oxidative cleavage of the C-O bond is the enantioselectivity discriminating step in the catalytic cycle. Once iridium complex C is formed, the iridium alkoxide complex could be protonated by the N-substituted piperazine nucleophiles to generate cationic iridium complex E. This proton transfer has two effects. First, the iridium species are made more electrophilic as a result of the positive charge, and the nucleophile is rendered more nucleophilic by deprotonation. Second, the positioning of the iridium metal on the π-allyl moiety will influence the regioselectivity of nucleophilic attack. Nucleophilic attack with inversion is proposed to occur adjacent to the alkoxy group in an SN2 fashion relative to the iridium metal. Finally, product 2 is subsequently liberated and the iridium monomer is regenerated, which will either reform the dimer or continue the catalytic cycle. Based on our findings above, we propose a mechanism, outlined in Scheme 1. When [Ir(COD)Cl] 2 was used as the iridium source, and reacted with (S)-p-Tol-BINAP to form the complex of [Ir(S)-p-Tol-BINAP)I] 2 A in the presence of NH 4 I, which is then cleaved by solvent to give the monomeric iridium complex B. Reversible exo coordination of oxabenzonorbornadiene 1a leads to iridium complex C, followed by oxidative insertion with retention to form a bridgehead C-O bond and produce the π-allyl iridium alkoxide complex D. We further propose that the oxidative cleavage of the C-O bond is the enantioselectivity discriminating step in the catalytic cycle. Once iridium complex C is formed, the iridium alkoxide complex could be protonated by the N-substituted piperazine nucleophiles to generate cationic iridium complex E. This proton transfer has two effects. First, the iridium species are made more electrophilic as a result of the positive charge, and the nucleophile is rendered more nucleophilic by deprotonation. Second, the positioning of the iridium metal on the π-allyl moiety will influence the regioselectivity of nucleophilic attack. Nucleophilic attack with inversion is proposed to occur adjacent to the alkoxy group in an S N 2 fashion relative to the iridium metal. Finally, product 2 is subsequently liberated and the iridium monomer is regenerated, which will either reform the dimer or continue the catalytic cycle.

General
Solvents and solutions were transferred with syringes. 1 H-NMR spectra were recorded at 400 MHz using a Varian XL (Palo Alto, CA, USA) 400 spectrometer with CDCl 3 as reference standard (7.27 ppm). Spectral features are tabulated in the following order: Chemical shift (ppm); number of protons; multiplicity (s-singlet, d-doublet, t-triplet, q-quartet, m-multiplet, br-broad); coupling constants (J, Hz), 13 C-NMR spectra were recorded at 400 MHz with CDCl 3 as reference standard (77.23 ppm). IR spectra were obtained using a Nicolet DX (Madison, WI, USA) FT-IR spectrometer. High resolution mass was obtained from a VG 70-250S (double focusing) mass spectrometer at 70 ev (Waters, Milford, MA, USA). The enantiomeric excess value was measured by HPLC with CHIRALCEL OD or AD columns (Chiral Technologies, Minato-ku, Japan). Melting points were taken with a Tai-Ke melting point apparatus (Beijing, China). Analytical TLC was performed using EM separations percolated silica gel 0.2 mm layer UV 254 nm fluorescent sheets (Beijing, China). Column chromatography was performed as "Flash chromatography" as reported by using (200-300 mesh) Merck grade silica gel (Merck, Beijing, China). The THF, toluene, DME, and THP was distilled from sodium benzophenone ketyl immediately prior to use. DMF, CH 2 Cl 2 , CH 3 CN, ClCH 2 CH 2 Cl, and 1,4-dioxane were distilled from calcium hydride. Furan was distilled prior to use. All other reagents were obtained from Alfa Aesar (Shanghai, China) and J & K (Guangzhou, China) and used as received unless otherwise stated.