Computational Design of SCS Nickel Pincer Complexes for the Asymmetric Transfer Hydrogenation of 1-Acetonaphthone

: Inspired by the active site structures of lactate racemase and recently reported sulphur– carbon–sulphur (SCS) nickel pincer complexes, a series of scorpion-like SCS nickel pincer complexes with an imidazole tail and asymmetric claws was proposed and examined computationally as potential catalysts for the asymmetric transfer hydrogenation of 1-acetonaphthone. Density functional theory calculations reveal a proton-coupled hydride transfer mechanism for the dehydrogenation of ( R )-(+)-1-phenyl-ethanol and the hydrogenation of 1-acetonaphthone to produce ( R )-(+)-1-(2-naphthyl)ethanol and ( S )-( − )-1-(2-naphthyl)ethanol. Among all proposed Ni complexes, 1 Ph is the most active one with a rather low free energy barrier of 24 kcal/mol and high enantioselectivity of near 99% enantiomeric excess ( ee ) for the hydrogenation of prochiral ketones to chiral alcohols.


Introduction
The synthesis of chiral compounds by metal-catalyzed asymmetric hydrogenation reactions has been widely used in the pharmaceutical [1], agrochemical [2], fragrance [3], and other fine chemical industries [4]. The catalytic asymmetric reduction of prochiral ketones and imines, especially asymmetric hydrogenation (AH) and asymmetric transfer hydrogenation (ATH), is one of the most efficient and versatile tools to produce chiral alcohols and amines. In both academic and industrial operations, catalysts used for AH and ATH are typically based on noble metals, such as Rh, Ir, and Ru [5]. The replacement of such high-cost and toxic precious metals with abundant and environmentally benign base metals, such as Fe, Co, Ni, etc., for catalytic AH and ATH reactions has attracted increasing attention in recent years [6][7][8][9][10][11][12]. Gao [9] and Morris [10][11][12] groups reported tetradentate PNNP iron catalysts for the ATH of acetophenone with high enantioselectivities. Morris and co-workers [13] reported unsymmetrical iron P-NH-P' complexes for the asymmetric hydrogenation of aryl ketones with enantiomeric excess (ee) values greater than 90%. In contrast to the encouraging progress archived in iron catalysts, cobalt catalysts for catalytic ATH of ketones have rather low enantioselectivities [9], and only a few Ni catalysts were reported so far [14][15][16]. In 2008, Hamada et al. [14] applied nickel bisphosphine complexes for the asymmetric hydrogenation of α-amino-β-keto ester hydrochlorides through dynamic kinetic resolution and achieved high diastereoand enantioselectivities (88−93% ee) for the production of anti-β-hydroxy-α-amino esters. Later on, they applied the same catalyst for the asymmetric hydrogenation of substituted aromatic α-aminoketone hydrochlorides to produce β-aminoalcohols with excellent diastereo-and enantioselectivities [15]. In 2012, Dong et al. [16] reported Ni(II) complexes chelated by PNO ligands for the ATH of a series of aromatic ketones using 2-propanol as the hydrogen source and obtained corresponding optical alcohols up to 84% ee under mild conditions. Inspired by the structures of the active site of lactate racemase [17] and experimentally reported sulphur-carbon-sulphur (SCS) palladium pincer complexes [18], we recently proposed a series of scorpion-like SCS nickel pincer complexes with an imidazole tail as potential catalysts for lactate racemization [19]. Our density functional theory (DFT) calculations revealed a weak enantioselectivity in the hydrogenation of pyruvate catalyzed by those SCS nickel pincer complexes with unsymmetrical ligands. Based on those findings, we proposed and computationally examined in this work a series of potential unsymmetrical catalysts with enhanced steric effect by adjusting the size of functional groups in the SCS pincer ligand for the more challenging asymmetric transfer hydrogenation of naphthyl ketone.  [18], our previously proposed scorpion-like SCS nickel pincer complex [19], and the unsymmetrical SCS nickel pincer complexes (1 Ph , 1 Me , 1 Et , and 1 tBu ) proposed in this study. Scheme 1 shows the catalytic cycles for the ATH of 1-acetonaphthone to (R)-(+)-1-(2-naphthyl)ethanol and (S)-(−)-1-(2-naphthyl)ethanol catalyzed by catalyst 1 Ph . (R)-(+)-1-phenyl-ethanol was used as the hydrogen source because the phenyl group in it could have π-π stacking interaction with the phenyl groups in 1 Ph (see Supplementary Materials) and reduce the reaction energy barriers. Actually, both (R)-(+)-1-phenyl-ethanol and (S)-(−)-1-phenyl-ethanol in racemic (±)-1-phenyl-ethanol could act as hydrogen sources with similar barriers for dehydrogenation. Figure 2 shows the calculated free energy profile of the reaction described in Scheme 1. The optimized structures of some important intermediates and transition states for hydrogen transfer are displayed in Figure 3. The free energy profiles of the same reaction catalyzed by 1 Me , 1 Et , and 1 tBu are shown in Figures 4-6, respectively. α-aminoketone hydrochlorides to produce β-aminoalcohols with excellent diastereo-and enantioselectivities [15]. In 2012, Dong et al. [16] reported Ni(II) complexes chelated by PNO ligands for the ATH of a series of aromatic ketones using 2-propanol as the hydrogen source and obtained corresponding optical alcohols up to 84% ee under mild conditions. Inspired by the structures of the active site of lactate racemase [17] and experimentally reported sulphur-carbon-sulphur (SCS) palladium pincer complexes [18], we recently proposed a series of scorpion-like SCS nickel pincer complexes with an imidazole tail as potential catalysts for lactate racemization [19]. Our density functional theory (DFT) calculations revealed a weak enantioselectivity in the hydrogenation of pyruvate catalyzed by those SCS nickel pincer complexes with unsymmetrical ligands. Based on those findings, we proposed and computationally examined in this work a series of potential unsymmetrical catalysts with enhanced steric effect by adjusting the size of functional groups in the SCS pincer ligand for the more challenging asymmetric transfer hydrogenation of naphthyl ketone.  [18], our previously proposed scorpion-like SCS nickel pincer complex [19], and the unsymmetrical SCS nickel pincer complexes (1Ph, 1Me, 1Et, and 1tBu) proposed in this study. Scheme 1 shows the catalytic cycles for the ATH of 1-acetonaphthone to (R)-(+)-1-(2-naphthyl)ethanol and (S)-(−)-1-(2-naphthyl)ethanol catalyzed by catalyst 1Ph. (R)-(+)-1-phenyl-ethanol was used as the hydrogen source because the phenyl group in it could have π-π stacking interaction with the phenyl groups in 1Ph and reduce the reaction energy barriers. Actually, both (R)-(+)-1-phenyl-ethanol and (S)-(−)-1-phenyl-ethanol in racemic (±)-1-phenyl-ethanol could act as hydrogen sources with similar barriers for dehydrogenation. Figure 2 shows the calculated free energy profile of the reaction described in Scheme 1. The optimized structures of some important intermediates and transition states for hydrogen transfer are displayed in Figure 3. The free energy profiles of the same reaction catalyzed by 1Me, 1Et, and 1tBu are shown in Figures 4, 5, and 6, respectively.        The free energy difference of the enantio-determining steps, TS5,6-Ph and TS5',6'-Ph, is 3.5 kcal/mol, which could lead to an approximate ee value of 99.5%. The asymmetric transfer hydrogenation of 1-acetonaphthone catalyzed by 1Ph, 1Me , 1Et, and 1tBu has similar mechanisms but different relative energies in the catalytic cycle. Table 1 lists the free energy barriers of the reactions and the enantioselectivity of 1Ph, 1Me, 1Et, and 1tBu. As shown in Table 1, the free energy barrier of the dehydrogenation of (R)-(+)-1-phenyl-ethanol (ΔG1) catalyzed by 1Ph is 25.1 kcal/mol (5Ph'→TS2,3-Ph), which is 0.5, 2.3, and 2.8 kcal/mol lower than the ΔG1 of 1Me (6Me'→TS2,3-Me) , 1Et (6Et'→TS2,3-Et), and 1tBu (1tBu→TS2,3-tBu), respectively. The free energy barriers of enantio-determining steps (ΔG2 and ΔG3) of 1Ph, 1Me, 1Et, and 1tBu have the same trend. Although 1Me, 1Et, and 1tBu have stronger steric effects than 1Ph, there is a trade-off between their free energy barriers and enantioselectivities. 1tBu has the strongest steric effect, which leads to the highest free energy barrier and the highest enantioselectivity. 1Ph has an enantioselectivity close to 1tBu, and the lowest free energy barriers for hydrogen transfers because of the π-π stack interaction between the naphthyl group in the reactant and the phenyl group in 1Ph.   The free energy difference of the enantio-determining steps, TS5,6-Ph and TS5',6'-Ph, is 3.5 kcal/mol, which could lead to an approximate ee value of 99.5%. The asymmetric transfer hydrogenation of 1-acetonaphthone catalyzed by 1Ph, 1Me , 1Et, and 1tBu has similar mechanisms but different relative energies in the catalytic cycle. Table 1 lists the free energy barriers of the reactions and the enantioselectivity of 1Ph, 1Me, 1Et, and 1tBu. As shown in Table 1, the free energy barrier of the dehydrogenation of (R)-(+)-1-phenyl-ethanol (ΔG1) catalyzed by 1Ph is 25.1 kcal/mol (5Ph'→TS2,3-Ph), which is 0.5, 2.3, and 2.8 kcal/mol lower than the ΔG1 of 1Me (6Me'→TS2,3-Me) , 1Et (6Et'→TS2,3-Et), and 1tBu (1tBu→TS2,3-tBu), respectively. The free energy barriers of enantio-determining steps (ΔG2 and ΔG3) of 1Ph, 1Me, 1Et, and 1tBu have the same trend. Although 1Me, 1Et, and 1tBu have stronger steric effects than 1Ph, there is a trade-off between their free energy barriers and enantioselectivities. 1tBu has the strongest steric effect, which leads to the highest free energy barrier and the highest enantioselectivity. 1Ph has an enantioselectivity close to 1tBu, and the lowest free energy barriers for hydrogen transfers because of the π-π stack interaction between the naphthyl group in the reactant and the phenyl group in 1Ph. The free energy difference of the enantio-determining steps, TS 5,6-Ph and TS 5',6'-Ph , is 3.5 kcal/mol, which could lead to an approximate ee value of 99.5%. The asymmetric transfer hydrogenation of 1-acetonaphthone catalyzed by 1 Ph , 1 Me , 1 Et , and 1 tBu has similar mechanisms but different relative energies in the catalytic cycle. Table 1 lists the free energy barriers of the reactions and the enantioselectivity of 1 Ph , 1 Me , 1 Et , and 1 tBu . As shown in Table 1, the free energy barrier of the dehydrogenation of (R)-(+)-1-phenyl-ethanol (∆G 1 ) catalyzed by 1 Ph is 25.1 kcal/mol (5 Ph '→TS 2,3-Ph ), which is 0.5, 2.3, and 2.8 kcal/mol lower than the ∆G 1 of 1 Me (6 Me '→TS 2,3-Me ), 1 Et (6 Et '→TS 2,3-Et ), and 1 tBu (1 tBu →TS 2,3-tBu ), respectively. The free energy barriers of enantio-determining steps (∆G 2 and ∆G 3 ) of 1 Ph , 1 Me , 1 Et , and 1 tBu have the same trend. Although 1 Me , 1 Et , and 1 tBu have stronger steric effects than 1 Ph , there is a trade-off between their free energy barriers and enantioselectivities. 1 tBu has the strongest steric effect, which leads to the highest free energy barrier and the highest enantioselectivity. 1 Ph has an enantioselectivity close to 1 tBu , and the lowest free energy barriers for hydrogen transfers because of the π-π stack interaction between the naphthyl group in the reactant and the phenyl group in 1 Ph .  Catalysts It is noteworthy that the special role of the imidazole groups in those newly proposed SCS pincer complexes are proton reservoirs facilitating proton-coupled hydride transfer in the dehydrogenation and hydrogenation reactions. The ethylene group connecting the pyridinium ring and the imidazole group ensures the adjustability of the imidazole group's position for easy accepting or donating of protons. The substituents on the arm of the SCS pincer ligand balance the free energy barriers and the enantioselectivity of the ATH reactions.

DFT Calculation Details
All DFT calculations in this study were performed using the Gaussian 09 (Revision C.01, Gaussion, Inc., Wallingford, CT, USA) suite of Ab Initio programs [20] for the ωB97X-D [21] functional with the all-electron 6-31+G(d,p) basis set for H, C, N, O, S, and P atoms, [22][23][24] and the Stuttgart relativistic effective core potential basis set for Ni (ECP10MDF) [25]. All structures in this paper were optimized in acetonitrile by using the integral equation formalism polarizable continuum model (IEFPCM) [26] with solvation model based on density (SMD) [27] atomic radii solvent corrections. The ground states were confirmed as singlet through comparison with optimized high-spin analogs. An ultrafine integration grid (99,590) was used for numerical integrations. Thermal corrections were calculated within the harmonic potential approximation on optimized structures under T = 298.15 K and 1 atm pressure. Unless otherwise noted, the relative energies reported in the text are Gibbs free energies with solvent effect corrections. The calculated structures were verified to have no imaginary frequency (IF) for all intermediates and only one IF for each transition state. All transition states were also confirmed to connect proper reactants and  It is noteworthy that the special role of the imidazole groups in those newly proposed SCS pincer complexes are proton reservoirs facilitating proton-coupled hydride transfer in the dehydrogenation and hydrogenation reactions. The ethylene group connecting the pyridinium ring and the imidazole group ensures the adjustability of the imidazole group's position for easy accepting or donating of protons. The substituents on the arm of the SCS pincer ligand balance the free energy barriers and the enantioselectivity of the ATH reactions.

DFT Calculation Details
All DFT calculations in this study were performed using the Gaussian 09 (Revision C.01, Gaussion, Inc., Wallingford, CT, USA) suite of Ab Initio programs [20] for the ωB97X-D [21] functional with the all-electron 6-31+G(d,p) basis set for H, C, N, O, S, and P atoms [22][23][24] and the Stuttgart relativistic effective core potential basis set for Ni (ECP10MDF) [25]. All structures in this paper were optimized in acetonitrile by using the integral equation formalism polarizable continuum model (IEFPCM) [26] with solvation model based on density (SMD) [27] atomic radii solvent corrections. The ground states were confirmed as singlet through comparison with optimized high-spin analogs. An ultrafine integration grid (99,590) was used for numerical integrations. Thermal corrections were calculated within the harmonic potential approximation on optimized structures under T = 298.15 K and 1 atm pressure. Unless otherwise noted, the relative energies reported in the text are Gibbs free energies with solvent effect corrections. The calculated structures were verified to have no imaginary frequency (IF) for all intermediates and only one IF for each transition state. All transition states were also confirmed to connect proper reactants and products by intrinsic reaction coordinate calculations. The JIMP2 molecular visualizing and manipulating program was employed to draw the 3D molecular structures [28].

Quantitative Estimation of Enantiomeric Excess
The enantiomeric excess can be quantitatively estimated from the free energy barrier difference of enantio-determining steps (∆∆G = ∆G 2 − ∆G 3 ) based on the transation state theory [29]: where R is the universal gas constant and T is the absolute temperature of the reaction in calculation (298.15 K in this case).

Conclusions
Inspired by the structures of the active center of lactate racemase and some experimentally reported base metal SCS pincer complexes, we proposed a series of scorpion-like SCS pincer nickel complexes 1 Ph , 1 Me , 1 Et , and 1 tBu and computationally predicted their potentials as catalysts for the ATH of ketones. Our computational study reveals a proton-coupled hydride transfer dehydrogenation and hydrogenation mechanism, in which the proton and hydride on the hydroxymethyne group in (R)-(+)-1-phenyl-ethanol simultaneously transfer to the imidazole nitrogen and the sp 2 carbon coordinated to nickel and then transfer to the carbonyl group in 1-acetonaphthone for the formation of (R)-(+)-1-(2-naphthyl)ethanol or (S)-(−)-1-(2-naphthyl)ethanol. Among all SCS pincer nickel complexes we proposed, 1 Ph has well-balanced catalytic activity (∆G 2 = 23.8 kcal/mol) and enantioselectivity (∆∆G = −3.5 kcal/mol). We believe the π-π stacking effect between the phenyl groups in reactants and 1 Ph helps to stabilize intermediates and reduce energy barriers in the reaction. Our computational design not only provides prototypical SCS nickel pincer complexes as promising catalysts for the asymmetric hydrogenation of 1-acetonaphthone, but also sheds a light on the development of efficient base metal catalysts with high chiral selectivity. The further design of SCS nickel pincer complexes for the asymmetric hydrogenation of various ketones and imines using different hydrogen sources is underway.
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4344/9/1/101/s1, Table S1: Relative free energies between 1 tBu and key transition states calculated by using different density functionals, Table S2: Relative free energies between 1 tBu and key transition states in different solvents, Figure S1: Optimized structures and solvent corrected absolute free energies of 1 Ph and two conformers with the imidazole group far away from the metal center (1 Ph ') and on the other side of the pincer ligand close to the PPh 2 group (1 Ph ").
Funding: This work was supported by the National Natural Science Foundation of China (21673250, 21703256, and 21873107). The authors are also grateful to the Institute of Chemistry, Chinese Academy of Sciences for funding.

Conflicts of Interest:
The authors declare no conflict of interest.