Recent Advances in the Synthesis of Chiral Tetrahydroisoquinolines via Asymmetric Reduction

Abstract

Enantiopure tetrahydroisoquinolines (THIQs), recognized as privileged skeletal structures in natural alkaloids, have attracted considerable attention from chemists due to their biological and pharmacological activities. Synthetic strategies for optically active THIQs have been rapidly and extensively developed in the past decades. In view of simplicity and atom economy, asymmetric reduction of N-heteroaromatics, imines, enamines, and iminium salts containing an isoquinoline (IQ) moiety should be the preferred approaches to obtain chiral THIQs. This review focuses on recent advances in the catalytic asymmetric synthesis of enantiopure THIQs via asymmetric reduction, including asymmetric hydrogenation, transfer hydrogenation, reductive amination, and deracemization. Highly enantioselective synthesis of THIQs was achieved via transition-metal-catalyzed asymmetric reduction and organocatalytic asymmetric reduction utilizing either catalyst activation or substrate activation strategy. Despite much progress in the enantioselective synthesis of THIQs, there still remain considerable opportunities and challenges for progress and developments in this field of research, particularly in the development of asymmetric catalytic systems for the direct reduction of IQs.

1. Introduction

Optically active tetrahydroisoquinolines (THIQs), recognized as privileged scaffolds, are prevalent in natural alkaloids (Figure 1) [1,2,3,4]. Notably, (S)-norcoclaurine derived from dopamine and L-tyrosine, serves as a significant building block for the construction of various benzyl THIQ alkaloids through enzymatic processes found in a range of plant species, including (S)-reticuline, (S)-scoulerine, (S)-norlaudanosine [3]. THIQ alkaloids represent an important family of the biologically active molecules whose potential pharmaceutical activities, including anti-tumor, anti-depressant, anti-HIV, anti-amoebic, anti-virus activities, have attracted considerable interest over the past decades [2]. For instance, (R)-salsolinol [5] acts as a potent competitive inhibitor of human monoamine oxidase A. Meanwhile (S)-scoulerine [6] functions as a sedative and muscle relaxing agent. Notably, several commercially available pharmaceutical drugs, such as (+)-quinapril [7] and (+)-solifenacin [8,9], have been successfully utilized as active pharmaceutical ingredients in clinical settings for the treatment of hypertension and overactive bladder, respectively, for decades. The growing demand within the pharmaceutical market to access chiral drugs containing a THIQ skeleton in optically pure form has the spurred rapid development of enantioselective synthetic methodologies.

As a result of the flourishing development of organic chemistry, various synthetic approaches to obtain chiral THIQs have emerged over the past decades, including kinetic resolution, dynamic kinetic resolution/transformation, asymmetric Pictet-Spengler cyclization, asymmetric nucleophilic addition, asymmetric cross-dehydrogenative coupling, asymmetric reduction, etc. [1]. Among these methodologies, catalytic asymmetric reduction of unsaturated bond (C=N bonds) undoubtedly represents one of the most straightforward and atom-economical strategies for the enantioselective synthesis of optically pure THIQs (Scheme 1) [10]. The catalytic reduction of imines, enamines, and N-heteroaromatics bearing an isoquinoline (IQ) core enables the stereoselective synthesis of a wide range of chiral THIQs. Hydrogen or hydride sources are the most commonly used reductants in asymmetric reduction, including hydrogen gas (H₂), formic acid 1, isopropanol 4, Hantzsch esters (HEHs) 5, dihydrophenanthridines (DHPDs) 8, etc. From the perspective of reductants, asymmetric reduction encompasses both asymmetric hydrogenation and asymmetric transfer hydrogenation. Asymmetric hydrogenation has been achieved by employing transition metals as catalysts, along with hydrogen gas as a green and clean hydrogen source [11,12,13,14,15,16,17,18,19,20,21]. This process entails generating catalytically active chiral metal-hydride (M-H) species followed by hydride addition to unsaturated bonds to facilitate the production of chiral amines. However, asymmetric transfer hydrogenation generally employs organic hydride donors as hydrogen sources, such as formic acid, isopropanol, HEHs, or DHPDs [14,20,22,23,24,25,26,27,28,29]. The hydride transfers from hydrogen sources to prochiral substrates through hydride addition involving a chiral active metal-hydride species or via hydrogen bonding with an organocatalyst. Furthermore, asymmetric reductive amination and deracemization generally yield enantiopure THIQs via the asymmetric hydrogenation or transfer hydrogenation of the in situ generated imines or iminium salts.

Thanks to the indefatigable efforts of chemists, significant developments have been made in the enantioselective synthesis of enantiopure THIQs [1,4,30]. Although several reviews have been published on the enantioselective synthesis of THIQ alkaloids, none has provided a comprehensive and timely overview specifically addressing enantioselective synthesis of THIQs via asymmetric reduction [1,4,31]. This paper focuses on recent advances in the synthesis of chiral THIQs via an asymmetric reduction process by employing both chiral organometallic catalysts and organocatalysts.

2. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Transition-Metal-Catalyzed Hydrogenation

Hydrogen gas (H₂) is widely acknowledged as the most efficient and cost-effective reductant in catalytic hydrogenation. Asymmetric hydrogenation of N-heteroaromatics, imines, enamines, and iminium salts featuring an IQ core has proved to be one of the most straight and effective strategies for enantioselective synthesis of THIQs in terms of atom economy and environmental sustainability. Various synthetic strategies have been developed for the asymmetric synthesis of THIQs by employing chiral organometallic catalysts, including chiral titanium, ruthenium, rhodium, iridium complexes, etc.

2.1. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Titanium-Catalyzed Hydrogenation

The titanium-catalyzed hydrosilylation of unsaturated compounds, such as olefins, alkynes, ketones, represents an effective approach to introduce silyl groups into substrate. In 1991, the Buchwald group employed Cp*₂TiCl₂ as a catalyst, ⁿBuLi as an activator, and triethoxysilane as a reductant for the reduction of esters (Scheme 2) [32]. This reduction was achieved through the hydrosilylation of esters, followed by in situ hydrolysis under either acidic or alkaline conditions.

Inspired by research on the titanium-catalyzed reduction of esters, Buchwald and co-workers disclosed the first titanocene-catalyzed hydrogenation of imines. With IQ-type imine 13a as starting material, hydrogenation was performed under high H₂ pressure (2000 psi), furnishing chiral 6,7-methoxy-1-phenyl-THIQ 14a at an 82% yield and 98% ee [33]. By employing chiral titanocene complex 15 as a catalyst precursor, the catalytically active species titanium^III-hydride 16 was formed in the presence of ⁿBuLi and phenylsilane. Subsequently, the asymmetric reduction of imines proceeded effectively through a process of titanium-hydride addition with hydrogen gas as a stoichiometric reductant.

2.2. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Ruthenium-Catalyzed Hydrogenation

Heterogeneous asymmetric hydrogenation presents a promising option in industrial synthesis, primarily due to the competitive advantages in the recyclability and reusability of catalysts. In 2012, Bartók and co-workers developed the first heterogenous chiral ruthenium catalysts for the asymmetric hydrogenation of IQ-type imine 13b and enamine 19a (Scheme 3) [34]. The chiral catalyst ([RuCl₂(p-cymene)]₂/(S,S)-TsDPEN) was immobilized on aminophosphane-functionalized silica SiO₂-NPPh-Ru 17 and 18 to form chiral heterogenous catalysts which exhibited excellent activity and enantioselectivity, providing (R)-salsolidine 14b and its derivative 20a in good yields and ee values. However, a significant decrease in catalytic activity was observed when using the ruthenium immobilized catalyst during a second recycling process. By adding additional [RuCl₂(p-cymene)]₂ and (S,S)-TsDPEN for each hydrogenation cycle, the desired product (R)-salsolidine was successfully obtained with a retention of reactivity and enantioselectivity during the third recycle run. This strategy provides a straightforward and scalable synthetic technique for the application of heterogeneous asymmetric synthesis in industry.

In 2013, Fan and co-workers utilized heterogeneous chiral cationic ruthenium catalysts for the asymmetric hydrogenation of dihydroisoquinolines (DHIQs) 13 and quinolines in imidazolium ionic liquids (Scheme 3) [35]. The enantioselective hydrogenation of 1-alkyl 3,4-DHIQs was carried out utilizing a [Ru(hexamethylbenzene) ((R,R)-TsDPEN)OTf] complex 21 as a catalyst in [Bmim]NTf₂, providing a series of 1-alkyl THIQs 14 with excellent yields and ee values. The ruthenium catalyst exhibited enhanced stability when immobilized in ionic liquids compared to organic solvents. Furthermore, the catalyst could be recycled in the six consecutive runs without loss of reactivity and enantioselectivity.

2.3. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Rhodium-Catalyzed Hydrogenation

Rhodium complexes serve as effective catalysts for asymmetric transfer hydrogenation of C=N bonds but have been less explored in the research of the asymmetric hydrogenation of imines.

2.3.1. Asymmetric Hydrogenation of DHIQs with Rhodium Catalysts

In 2008, Xiao and co-workers successfully applied a chiral Rh/(R,R)-TsDPEN 22 complex to the asymmetric hydrogenation of DHIQs 13 and dihydro-β-carbolines with excellent yields and ee values (Scheme 4) [36]. Initially, the asymmetric hydrogenation failed to proceed with a [RhTsDPENCl] catalyst, which could be attributed to the strong coordination of the Rh-Cl bond impeding the formation of active cationic Rh-hydride species. Considering the necessity for the formation of Rh-hydride, an evaluation of the various commercially available silver salts, including AgPF₆, AgOTf, AgSbF₆, was carried out. Notably, dramatic anion effects on catalytic activity were observed with AgSbF₆ in conjunction with a small quantity of water (30 μL H₂O per 0.5 mmol substrates).

2.3.2. Asymmetric Hydrogenation of IQs with Rhodium Catalysts

Asymmetric hydrogenation of N-heteroaromatics is one of the most important synthetic methodologies for the synthesis of N-heterocycles, such as tetrahydroquinolines, indolines, and tetrahydroquinoxalines, etc. Despite considerable advancements in the asymmetric hydrogenation of N-heteroaromatics, the direct asymmetric hydrogenation of IQs remains a challenge due to their intrinsic problems. The following major obstacles need to be overcome: (1) the inherent aromatic stability of IQs; and (2) the strong coordination ability exhibited by both substrates and products, which can lead to metal catalysts poisoning. To overcome these difficulties, various synthetic strategies have been developed for the asymmetric hydrogenation of IQs, including catalyst activation and substrate activation [21,37].

With the activation of substrates and catalysts by strong Brønsted acid (HCl), Zhang and co-workers realized Rh-catalyzed asymmetric hydrogenation of IQs with a chiral thiourea-derived phosphine ligand, providing chiral THIQs 14 with high enantiomeric excesses and yields (Scheme 5a) [38]. Brønsted acid was essential for the hydrogenation, and the Rh-catalyzed hydrogenation failed to proceed in absence of Brønsted acid. The NMR studies revealed that an anion binding occurred between the chloride anions of the substrate and the N–H protons of the chiral ligand (Scheme 5b), which promoted effective stereoselective control over hydride transfer. To further investigate this mechanism, deuterium labeling experiments were conducted (Scheme 5c). The results from H–D exchange showed a remarkable 95% incorporation of deuterium at C-4 position of THIQs (d₂-14d) with IQ-type deuterium chloride CD₃OD being used as a solvent. The results revealed Brønsted acid-promoted enamine-iminium tautomerization of the partial hydrogenated intermediate 24a (Scheme 5d). Then, the hydrogenation proceeded by capturing the more active iminium intermediate 25a.

2.4. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Iridium-Catalyzed Hydrogenation

Iridium is one of the most effective and attractive catalysts for the asymmetric hydrogenation of imines and N-heteroaromatics [11,14,15,16]. The strong coordination with heteroatoms poisons the catalysts, resulting in deactivation. To improve the catalytic performance, various strategies have been developed in the past decades, including catalyst activation and substrate activation [37].

2.4.1. Asymmetric Hydrogenation of DHIQs with Iridium Catalysts

Catalyst Activation Strategy with Imides

In 1995, employing phthalimide as a co-catalyst, the Morimoto and Achiwa reported the iridium-catalyzed asymmetric hydrogenation of cyclic imines 13 bearing an IQ core, providing (S)-salsolidine 14b and its derivative in good to excellent enantioselectivities (Scheme 6) [39]. Remarkable improvement in enantioselectivity was observed in the presence of phthalimide, which may be coordinated to the vacant site of iridium center to improve the enantioselectivity of catalyst [40].

Catalyst Activation Strategy with Brønsted Acids/Oxidizing Halogen Reagents

In 2006, the Mashima group developed the first iridium-catalyzed asymmetric hydrogenation of cyclic imines via catalyst activation using Brønsted acid (Scheme 7) [41]. The iridium^III complexes, including the mononuclear halo-carboxylate iridium^III complexes 27 and cationic dinuclear triply halogen-bridged iridium^III complexes 28, were synthesized through the coordination of iridium^I precursor ([Ir(COD)Cl]₂) with chiral ligands, followed by oxidative addition of HX to chiral iridium^I complexes 27 or 28. The resulting iridium^III catalysts 27 and 28 displayed excellent activity and enantioselectivity in asymmetric hydrogenation.

Since the pioneering work of the Mashima group, catalyst activation using catalytic amounts of Brønsted acid has been widely applied in the iridium-catalyzed asymmetric hydrogenation of IQ-type imines, enamines, iminium salts, N-heteroaromatics, etc. (Scheme 8). In 2011, Zhang and co-workers achieved the asymmetric hydrogenation of DHIQs 13 by employing an iodine-bridged dimeric iridium complex as a catalyst [42]. The iridium^III complex [{Ir(H)[(S,S)-(f)-Binaphane]₂(μ-I)₃]⁺I⁻ was synthesized in the presence of [Ir(COD)Cl]₂, (S,S)-f-binaphane, and HI. The addition of I₂ and HI significantly improved both the activity and stereoselectivity in the hydrogenation process of DHIQs 13 via effective catalyst activation.

In 2012, Zhou and co-workers reported the iridium-catalyzed asymmetric hydrogenation of 1-alkyl DHIQs 13 utilizing a chiral spiro phosphoramidite ligand under low hydrogen pressure [43]. Hydrogenation failed to proceed in the absence of iodides. Remarkable improvements in catalytic performance were observed with various iodide sources, including I₂, KI, LiI or Bu₄NI, which could be ascribed to the in situ formation of Brønsted acid HI [44].

Scheme 8. Ir-catalyzed asymmetric hydrogenation of DHIQs via catalyst activation with Brønsted acids [42,43,45,46,47].

Scheme 8. Ir-catalyzed asymmetric hydrogenation of DHIQs via catalyst activation with Brønsted acids [42,43,45,46,47].

Next, the Zhang group employed HBr as a catalyst activator for the iridium-catalyzed enantioselective synthesis of various challenging sterically hindered THIQs 14, employing a chiral JosiPhos-type binaphane ligand [45,46]. This strategy demonstrated broad substrate applicability, excellent reactivity, and exceptional enantioselectivity in producing chiral THIQs. A gram-scale asymmetric hydrogenation was conducted smoothly, providing chiral 1-phenyl THIQ 14f with an impressive turnover number (TON) of 3000. Subsequently, by using a catalytic amount of trifluoroacetic acid (TFA) or methanesulfonic acid (MsOH), the asymmetric hydrogenation of 1-alkyl DHIQs 13 was also achieved in an iridium catalytic system [47]. The presence of TFA and MsOH contributed to the high reactivity and enantioselectivity observed during hydrogenation, due to the formation of mononuclear halo-carboxylate iridium^III catalysts.

Oxidizing agents containing halogens, such as trichloroisocyanuric acid (TCCA), N-bromosuccinimide (NBS), N-iodosuccinimide (NIS), and 1,3-dichloro-5,5-dimethylhydantoin (DCDMH), also serve as excellent catalyst activators in iridium-catalyzed asymmetric hydrogenation. This exceptional performance could be ascribed to the following aspects. Firstly, halogen-containing oxidizing agents could elevate the valence state of the iridium^I center to catalytically active iridium^III species. Secondly, these oxidizing agents could be hydrogenated to the corresponding imides along with Brønsted acids HX to activate the catalysts under hydrogenation conditions.

In 2018, Zhou and co-workers reported the iridium-catalyzed hydrogenation of DHIQs 13 with a catalytic amount of NBS as an activator, giving the desired chiral THIQs 14 in moderate to good enantioselectivities (Scheme 9) [48]. The Facchetti and Rimoldi group then developed a heterobiaryl diphosphine ligand tetraMe-BITIOP bearing an axial chirality, which exhibited excellent catalytic performance in iridium-catalyzed enantioselective hydrogenation of DHIQs 13 via catalyst activation with NBS [49].

Substrate Activation Strategy of DHIQs with Brønsted Acids

Brønsted acids, as traceless activators, were found to be the most attractive and convenient reagents for substrate activation due to their ease of handling with both substrates and products. Brønsted acid could efficiently improve the reactivity of substrate via the in situ formation of iminium salts, while simultaneously avoiding the poisoning effect on iridium catalysts from the strong coordination of nitrogen atoms.

In 2012, the Zanotti-Gerosa group employed [Ir(COD)Cl]₂/(S)-P-Phos as a catalyst to the asymmetric hydrogenation of 1-Phenyl DHIQ hydrochloride salt (13f‧HCl) in the presence of equivalent amounts of Brønsted acid H₃PO₄ (Scheme 10) [50]. The addition of H₃PO₄ effectively improved both the activity and enantioselectivity of hydrogenation. The hydrogenation could occur smoothly on a multigram scale with a substrate-to-catalyst (S/C) molar ratio of 1060, providing over 200 g (S)-(1)-phenyl THIQ (14f) at 97% yield and 96% ee. Subsequently, Togni and co-workers developed a series of novel chiral P-trifluoromethyl ligands derived from JosiPhos which also exhibited excellent activity and enantioselectivity in the iridium-catalyzed hydrogenation of DHIQ hydrochloride salts (13‧HCl) [51]. Notably, when 1-phenyl DHIQ-type imine (13f) and imine hydrochloride (13f‧HCl) were subjected to the standard conditions, completely different results were obtained in terms of both reactivity and stereoselectivity. Therefore, Brønsted acid HCl was found to be responsible for the excellent activity and enantioselectivity of hydrogenation.

In 2018, the Zhou group realized the dual stereocontrol for enantioselective synthesis of chiral THIQs via the iridium-catalyzed asymmetric hydrogenation of DHIQs 13 by tuning the amounts of NBS (Scheme 11) [48]. An interesting reversal of enantioselectivity of products was observed when varying between conditions A and B. With a catalytic amount of NBS as an activator for catalyst, the (S)-enantiomer of THIQs ((S)-14) was obtained. However, when 1.5 equivalent amounts of NBS were used for both catalyst and substrate activation, the (R)-enantiomer of THIQs ((R)-14) was obtained. Mechanistic studies showed that NBS could be hydrogenated to the corresponding succinimide along with equivalent amounts of HBr in the presence of [Ir(COD)Cl]₂/(R)-BINAP. The iridium-catalyzed asymmetric hydrogenation DHIQ hydrobromide 13f‧HBr proceeded smoothly, providing THIQ 14f with a configuration that was distinctly different from that observed under condition A. Based on the experimental results and literature evidence [52,53], two completely different hydrogenation modes were proposed. With a catalytic amount of NBS, the addition of Ir-H species to C=N bond proceeded via a four-membered cyclic transition state TS₁ in an inner-sphere mechanism. In contrast, when using 1.5 equivalent amounts of NBS, DHIQs reacted with the in situ generated HBr derived from iridium-catalyzed hydrogenolysis of NBS to form iminium salts, followed by the asymmetric hydrogenation of the iminiums via a six-membered cyclic transition state TS₂.

Substrate Activation Strategy of DHIQs with Organic Halides

Organic halides are undoubtedly efficient reagents for the activation of imines due to the rapid in situ formation of active iminium salts. In line with substate activation, Ratovelomanana-Vidal and co-workers described the iridium-catalyzed asymmetric hydrogenation of DHIQs 14 via iminium intermediates in the presence of TsCl and a proton sponge (Scheme 12) [54]. The addition of a proton sponge served to neutralize the in situ generated HCl and inhibit the direct hydrogenation of hydrogen chloride salts of imines, which also acted as active species in the asymmetric hydrogenation. Additionally, Zhou and co-workers utilized alkyl halide activated imines as substrates, realizing the highly enantioselective synthesis of tertiary amines 31 bearing a THIQ skeleton, which were the essential frameworks of some biologically active molecules, such as (R)-carnegine, (R)-harmicine [55].

2.4.2. Asymmetric Hydrogenation of IQ-Type Enamines

The asymmetric hydrogenation of IQ-type enamines is also a highly effective approach for the synthesis of tertiary amines bearing a THIQ core. Hydrogenation was achieved through an enamine-imine/iminium tautomerization, followed by asymmetric reduction by capturing the active imine or iminium tautomer.

Spiro chiral phosphoramidite scaffolds are prevalent ligands in transition-metal catalyzed asymmetric synthesis owing to their unique stereoselectivity. Zhou and co-workers developed a variety of spiro chiral phosphoramidite ligands, which were subsequently applied to the asymmetric hydrogenation of IQ-type enamines 32, furnishing tertiary amines 31 bearing a THIQ moiety in elegant yields and ee values (Scheme 13) [56]. When R’ group was alkyl, hydrogenation proceeded smoothly under low pressure H₂ (1 atm), which was completely different from the aryl substitute. The active iridium catalytic species was formed in the presence of iodine and characterized as Ir[(S_a,R,R)-L]₂I₂Cl [52]. A deuterium labeling experiment exhibited an enamine-iminium tautomerization in the asymmetric hydrogenation, which was promoted by the in situ generated DI through iminium intermediate 33.

2.4.3. Asymmetric Hydrogenation of IQs

The inherent aromatic stability, lower reactivity, and strong coordination ability of IQs 34 make asymmetric hydrogenation a challenge. The destruction of aromatic stability can significantly enhance the reactivity of substrates and simultaneously avoid the poisoning of catalysts from strong coordination to nitrogen. Substrate activation with organic halides and Brønsted acids generally serves as the most straightforward and powerful method (Scheme 14). The formation of salts 23 and 34‧HX reduces the aromaticity of the substrate and the strong coordination effects of the heteroatoms. In addition, the introduction of directing groups to the substrates enables a secondary coordination site to transition metals, resulting in elegant stereoselective control.

Substrate Activation Strategy of IQs with Organic Halides

In 2006, the Zhou group firstly employed chloroformates as substrate activators in the iridium-catalyzed asymmetric hydrogenation of IQs 34 (Scheme 15) [57]. Considering the formation of active species isoquinolinium salts 23, a series of lithium salts were investigated. Compared to the in situ generated LiCl, enantioselectivities were significantly enhanced in the presence of BF₄⁻ counterion. Hydrogenation provided the partially hydrogenated products 1,2-DHIQs in good yields and ee values, which could be converted into chiral THIQs through a two-step reduction [58]. Consequently, biologically active (S)-(−)-carnegine could also be obtained through this three-step transformation. Subsequently, those researchers successfully realized the iridium-catalyzed asymmetric hydrogenation of readily available isoquinolinium salts 23, which can be prepared from the reaction of IQs 34 with benzyl bromide [59]. The hydrogenation underwent a cascade process that involved [1,2]- or [1,4]-hydride addition, enamine-iminium isomerization, and asymmetric hydrogenation of iminium salts. The in situ generated HBr promoted the rapid enamine-iminium isomerization to form more reactive iminium salts.

Substrate Activation Strategy of IQs with Brønsted Acids

Brønsted acids, serving as traceless substrate activators, were proved to be the most attractive and convenient reagents for substrate activation, due to the associated easy treatment for substrates and products. In 2013, Mashima and co-workers utilized chiral iridium complexes A in the asymmetric hydrogenation of isoquinolinium salts (34‧HCl), which are derived from IQs and HCl (Scheme 16) [53,60]. The dual activation of both catalyst and substrate with HCl contributed significantly to the excellent reactivity and enantioselectivity. A range of enantiomerically pure THIQs 14 containing 1-alkyl/aryl substituent, 3-alkyl/aryl substituent, or 1,3-dialkyl/diaryl substituents could be obtained with excellent activity and enantioselectivity [53]. Subsequently, the Zheng and Zhang group described the enantioselective synthesis of 3-substituted THIQs 14 with a binaphane-derived JosiPhos as a chiral ligand through iridium-catalyzed asymmetric hydrogenation of quinolinium salts (34‧HCl) [61].

In 2013, Zhou and co-workers reported the iridium-catalyzed asymmetric hydrogenation of 3,4-disubstituted quinolinium salts (34‧HCl), furnishing chiral THIQs bearing a unique C-F stereogenic center with excellent yields and enantioselectivities [62]. A DCDMH-mediated catalyst activation strategy was utilized to improve the catalytic performance. The asymmetric hydrogenation of quinolinium salts may proceed through a dynamic kinetic resolution process, including the addition of an Ir-H to C=N bond to form 36, Brønsted acid-promoted rapid enamine–imine isomerization, and further addition of Ir-H to imine intermediate 37. Rapid enamine–imine isomerization would inhibit the hydrodefluorination pathways efficiently.

Employing the oxidizing halogenide TCCA as an activator, Zhou and co-workers realized the iridium-catalyzed asymmetric hydrogenation of IQs 34 (Scheme 17) [63]. To gain more insight into the mechanism, the halogen-bond complex 39, derived from 1-phenyl THIQ and NIS, was synthesized and subsequently confirmed by single-crystal X-ray crystallography, which suggested a halogen-bond interaction between the substrate and NIS. The hydrogenolysis of a halogen-bond complex proceeded in the presence of catalytically active Ir-H species to form HX, followed by the asymmetric hydrogenation of the in situ generated active isoquinolinium halide 34‧HX.

Substrate Activation Strategy with Directing Groups

Introducing a directing group to a substrate can enhance its coordination ability with the iridium center, thereby increasing the catalytic activity and stereoselectivity. In 2020, Stoltz’s group demonstrated the asymmetric hydrogenation of IQs by introducing a tethering auxiliary group at the C-1 position of IQs, such as hydroxyl, alkoxyl, amino groups, etc. This modification facilitated the cis-1,3-disubstituted THIQs (cis-14) with excellent diastereoselectivity and enantioselectivity (Scheme 18) [64]. Brønsted acid (AcOH) and additive TBAI ((ⁿBu)₄NI) were also crucial for the hydrogenation in terms of reactivity and stereoselectivity, which might be ascribed to the following aspects: (1) catalyst activation with Brønsted acid and iodide by elevating the valence state of Ir^I to Ir^III; (2) substrate activation with Brønsted acid; (3) Brønsted acid promoted the rapid tautomerization of enamine and imine; and (4) Brønsted acid mediated the dissociation of the product from iridium complex. However, the mechanism remains poorly understood.

Very recently, Stoltz’s group modified the experimental conditions by switching the solvent THF and additive TBAI to 1,2-dichloroethane (1,2-DCE) and TBABr ((ⁿBu)₄NBr), respectively. This adjustment facilitated a trans-selective hydrogenation of 1,3-disubstituted IQs [65]. This interesting result could be ascribed to the non-coordination of 1,2-DCE and a smaller halide Br, which changed the stereoselective control of hydrogenation. The strategy was also applied to the unique bis-THIQ alkaloids 41 with excellent reactivity and enantioselectivity.

Direct Hydrogenation Strategy

With a BCDMH activating iridium catalyst, Zhou and co-workers realized the first asymmetric hydrogenation of 3,4-disubstituted IQs via dynamic kinetic resolution (DKR), enabling hydrogenation under a hydrogen pressure of 40 psi without further substrate activation (Scheme 19) [66]. The asymmetric hydrogenation experienced a tandem process of Ir-H addition to the C=N bond of substrate 34a and tautomerization of enamine 24 to imine (R)-25b and (S)-25b, followed by asymmetric reduction of imine 25. The in situ generated Brønsted acid promoted enamine–imine tautomerization, which was crucial for the rapid conversion of imine (R)-25b and (S)-25b back to enamine 24 (k₋₁ >> k₂). Meanwhile, the rate of asymmetric hydrogenation of (R)-enantiomer 25 was faster than that of (S)-enantiomer 25 (k₋₁ >> k₂ >> k₃), resulting in high diastereoselectivity and enantioselectivity of hydrogenation to form the hydrogenated product cis-14.

3. Asymmetric Synthesis of Chiral THIQs via Catalytic Transfer Hydrogenation

The asymmetric transfer hydrogenation of ketones, imines, and N-heteroaromatics has been extensively developed in the past decades [67]. From the perspective of catalytic systems, asymmetric transfer hydrogenation includes both transition-metal-catalyzed [23] and organocatalytic [22,26,27,28,29] transfer hydrogenation. In transition-metal-catalyzed asymmetric transfer hydrogenation, HCOOH, HCOONa, and isopropanol are the most widely used hydrogen sources, facilitating the catalytically active chiral metal-hydride species for further reduction of C=N bonds. In organocatalytic asymmetric transfer hydrogenation, HEH and DHPD are generally utilized to obtain the desired chiral products via hydrogen bonding.

3.1. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Ruthenium-Catalyzed Transfer Hydrogenation

3.1.1. Asymmetric Transfer Hydrogenation of DHIQs

Since the first example of Ru/TsDPEN-catalyzed asymmetric transfer hydrogenation of imines, presented by the Noyori group in 1996, Ru/chiral diamines catalytic systems have been extensively applied to the asymmetric transfer hydrogenation of DHIQs 13 (Table 1) [68]. With Ru/TsDPEN as a catalyst and HCOOH/Et₃N (the ratio of volume 5/2) as a hydrogen source, the asymmetric transfer hydrogenation of DHIQs 13 proceeded smoothly in an aprotic solvent, yielding chiral THIQs with elegant reactivity and stereoselectivity (entry 1).

Subsequently, Vedejs and co-workers applied the Noyori catalyst to the asymmetric transfer hydrogenation of DHIQs 13, furnishing chiral diamines (for X = NH₂, NBnTs, N(MOM)Ts) with a THIQ core with moderate to good yields and excellent enantioselectivity (entry 2) [69]. By introducing ortho amino group to the phenyl ring at C-1 position, a series of chiral diamines was obtained. Additionally, an imine substrate with a o-bromo group on the phenyl ring at C-1 position underwent Ru-catalyzed asymmetric transfer hydrogenation followed by Ullmann reaction, also resulting in chiral diamines. As privileged structural motifs, chiral diamines serve as elegant chiral organocatalysts or ligands and have a wide range of applications in the field of catalytic asymmetric reactions, including Michael addition, Mukaiyama aldol condensation, hydrogenation, dynamic kinetic resolution, etc. [22,70,71,72].

Table 1. Asymmetric transfer hydrogenation of DHIQs via ruthenium catalysis.

Entry	Substrate	Catalyst	Reaction Conditions	Yield (%)	Ee (%)	Group (Year)	Ref.
1		[Ru/(S,S)-N-N* a] or [Ru/(R,R)-N-N* b] or [Ru/(S,S)-N-N* c] (0.5 mol%)	HCOOH/Et3N (2.5 mL, 5:2) CH3CN or CH2Cl2, 28 °C	90–97	84–95	Noyori (1996)	[68]
2		[Ru/(R,R)-N-N* e] (1–2 mol%)	HCOOH/Et3N (0.25 mL, 5:2) CH2Cl2, RT	63–76	85–99	Vedejs (1999)	[69]
3		RuCl2(p-cymene)/ (R,R)-N-N* f (1 mol%)/(2.2 mol%)	HCOONa (5.0 eq.) CTAB (50 mol%) H2O, 28 °C	68–97	90–95	Zhu and Deng (2006)	[73]
4		[Ru/(S,S)-N-N* a] (0.6 mol%)	HCOONa (5.0 eq.) CTAB (50 mol%) H2O, 40 °C	87–90	99–99.5	Pihko (2009)	[74]
5		[Ru/(R,R)-N-N* a] or [Ru/(R,R)-N-N* c] or [Ru/(R,R)-N-N* d] (1 mol%)	HCOOH/Et3N (6.3 eq., 5:2) CH3CN or CH2Cl2, 40 °C	171–261 3.8–45 <1–265 (TOF/h)	82–94 39–80 63–87	Kacer (2013)	[75]
6		[Ru/(R,R)-N-N* e] (1 mol%)	HCOOH/Et3N (2 eq., 5:2) iPrOH, 30 °C	72–97	82–98	Ratovelomanana-Vidal (2013)	[76]
7		[Ru/(R,R)-N-N* e] (1 mol%)	HCOOH/Et3N (2 eq., 5:2) iPrOH, 30 °C	71–97	15–99	Ratovelomanana-Vidal (2015)	[77]
8		[Ru/(R,R)-N-N* g] (1 mol%)	HCOOH/Et3N (0.25 mL, 5:2) iPrOH, RT	25–90	88–93	Wills (2020)	[78]

Table 1. Asymmetric transfer hydrogenation of DHIQs via ruthenium catalysis.

Entry	Substrate	Catalyst	Reaction Conditions	Yield (%)	Ee (%)	Group (Year)	Ref.
1		[Ru/(S,S)-N-N* a] or [Ru/(R,R)-N-N* b] or [Ru/(S,S)-N-N* c] (0.5 mol%)	HCOOH/Et3N (2.5 mL, 5:2) CH3CN or CH2Cl2, 28 °C	90–97	84–95	Noyori (1996)	[68]
2		[Ru/(R,R)-N-N* e] (1–2 mol%)	HCOOH/Et3N (0.25 mL, 5:2) CH2Cl2, RT	63–76	85–99	Vedejs (1999)	[69]
3		RuCl2(p-cymene)/ (R,R)-N-N* f (1 mol%)/(2.2 mol%)	HCOONa (5.0 eq.) CTAB (50 mol%) H2O, 28 °C	68–97	90–95	Zhu and Deng (2006)	[73]
4		[Ru/(S,S)-N-N* a] (0.6 mol%)	HCOONa (5.0 eq.) CTAB (50 mol%) H2O, 40 °C	87–90	99–99.5	Pihko (2009)	[74]
5		[Ru/(R,R)-N-N* a] or [Ru/(R,R)-N-N* c] or [Ru/(R,R)-N-N* d] (1 mol%)	HCOOH/Et3N (6.3 eq., 5:2) CH3CN or CH2Cl2, 40 °C	171–261 3.8–45 <1–265 (TOF/h)	82–94 39–80 63–87	Kacer (2013)	[75]
6		[Ru/(R,R)-N-N* e] (1 mol%)	HCOOH/Et3N (2 eq., 5:2) iPrOH, 30 °C	72–97	82–98	Ratovelomanana-Vidal (2013)	[76]
7		[Ru/(R,R)-N-N* e] (1 mol%)	HCOOH/Et3N (2 eq., 5:2) iPrOH, 30 °C	71–97	15–99	Ratovelomanana-Vidal (2015)	[77]
8		[Ru/(R,R)-N-N* g] (1 mol%)	HCOOH/Et3N (0.25 mL, 5:2) iPrOH, RT	25–90	88–93	Wills (2020)	[78]

In 2013, a novel ligand, designated as (1R,2R)-N-((1S,2S)-borneol-10-sulfonyl)-1,2-diphenylethylenediamine ((R, R)-N,N* d), was developed for the asymmetric transfer hydrogenation of DHIQs 13, providing the hydrogenated products 14 with a turnover frequency (TOF) of up to 265 h⁻¹ and 87% ee (entry 5) [75]. Comparable catalytic activity and stereoselectivity were observed with the traditional Noyori catalyst. Meanwhile, Ratovelomanana-Vidal expanded the substrate scope to various DHIQs 13 (more than 90 examples) in a protic solvent, including substrates with steric hindrance groups at C-1 position (entry 6, 7) [76,77]. In 2020, the Will group developed a TsDPEN ligand containing a heterocyclic group on nitrogen atom for Ru-catalyzed asymmetric transfer hydrogenation of non-electron-rich and unhindered DHIQs 13 with high enantioselectivity (entry 8) [78].

H₂O, recognized as a green and clean solvent, has received widespread attention among chemists in terms of environmental problems in past decades. In 2006, the Zhu and Deng group successfully developed a water soluble and recyclable ruthenium catalyst for the asymmetric transfer hydrogenation of imines and iminium salts under mild conditions in aqueous media (entry 3) [73]. By employing water-soluble o,o’-9-disulfonated N-tosyl-1,2-diphenylethylene diamine ((R,R)-N,N* f) as a chiral ligand and HCOONa as a hydrogen source, a variety of IQ-type secondary and tertiary chiral amines were obtained. The catalytic activity and enantioselectivity could be enhanced significantly in the presence of surfactants, such as cetyltrimethylammonium bromide (CTAB). However, using HCOOH as a hydrogen source, the reaction did not proceed. Researchers showed that the pH values of aqueous media influenced the reactivity and enantioselectivity during the transfer hydrogenation, and the reaction was more efficient in a higher pH value media (pH = 9–10). This strategy was only applicable to substrates bearing an alkyl group at the C-1 position. The catalyst could be reused for three cycles with the remaining of reactivity and enantioselectivity in the chiral synthesis of γ-sultam (3-^tBu-2,3-dihydrobenzo[d]isothiazole 1,1-dioxide).

In 2009, utilizing sodium formate as a hydrogen source, the Pihko group documented the ruthenium-catalyzed asymmetric transfer hydrogenation of IQ-type imines 13 in aqueous media. This process employed a commercially available chiral diamine as a ligand ⁿC₁₆H₃₃N⁺(CH₃)₃Br¯ (CTAB) as a phase transfer catalyst (Scheme 20) [74]. Excellent reactivity and enantioselectivity were achieved. However, for benzyl and phenyl substrates, the hydrogenation encountered significant challenges. A series of Lewis acids were tested for the ruthenium-catalyzed asymmetric transfer hydrogenation. Fortunately, the catalytic activity and enantioselectivity of the catalyst were improved significantly in the presence of Lewis acid. With AgSbF₆ as an additive, 1-benzyl-6,7-dimethoxy-1,2,3,4-THIQ 14i was obtained with 90% conversion and 98% ee. With AgSbF₆/Bi(OTf)₃ as additives, substrate 14a bearing a phenyl group at C-1 position was obtained, providing the hydrogenated product with 87% conversion and 94% ee.

3.1.2. Asymmetric Transfer Hydrogenation of DHIQ-Type Iminium Salts

THIQs bearing a polycyclic skeletal structure are among the most privileged structures in natural alkaloids, such as (−)-crispine A, which has promising pharmacological antitumor activity. The asymmetric transfer hydrogenation of DHIQ-type iminiums 42 is the most direct and atom-economical process for the enantioselective synthesis of polycyclic skeletal THIQs 43 (

Table 2. Asymmetric transfer hydrogenation of DHIQ-type iminium salts via ruthenium catalysis.

Entry	Substrate	Catalyst	Reaction Conditions	Yield (%)	Ee (%)	Group (Year)	Ref.
1		[RuCl(C6H6)(S,S)-N-N* e] (10 mol%)	HCOOH/Et3N (2.5 mL, 5:2) CH3CN, 0 °C	96	92	Czarnocki (2005)	[79]
2		[RuCl2(p-cymene)]2 /(R,R)-N-N* f (1 mol%/2.2 mol%)	HCOONa (5.0 eq.) CTAB (50 mol%) H2O, 28 °C	86 (R = Me) 94 (R = Ph)	90 (R = Me) 95 (R = Ph)	Zhu and Deng (2006)	[73]
3		[RuCl(C6H6)(S,S)-N-N* e] (10 mol%)	HCOOH/Et3N (2.5 mL, 5:2) CH3CN, 0 °C	91 (n = 1) 92 (n = 2)	92 (n = 1) 87 (n = 2)	Czarnocki (2007)	[80]
4		[RuCl(p-cymene)(S,S)-N-N* a] (1.3 mol%), AgSbF6 (7 mol%)	HCOONa (15.0 eq.) CTAB (1.0 eq.), H2O, 40 °C	45 (n = 1) 65 (n = 2)	94 (n = 1) 96 (n = 2)	Pihko (2009)	[74]
5		[RuCl(benzene)((R,R)-N-N* h)] [RuCl(benzene)((R,R)-N-N* i)] (0.2 mol%)	HCOOH/Et3N (2.5 mL, 5:2) CH3CN, 0 °C	72 (n = 1) 89 (n = 2)	84 (n = 1) 56 (n = 2)	Czarnocki (2013)	[81]

). In 2005, Czarnocki and co-workers employed the Noyori catalyst ([RuCl₂(C₆H₆)]₂/(S,S)-N,N* e) for the asymmetric transfer hydrogenation of iminium salts 42a containing an IQ core in the presence of a hydrogen source (HCOOH/Et₃N), achieving the enantioselective synthesis of (−)-crispine A with 92% ee (entry 1) [79]. They applied this catalytic system to the enantioselective synthesis of other polycyclic skeletal THIQs 43 and tetrahydro-β-carboline (R)-(+)-harmicine (entry 3) [80]. Subsequently, a series of novel chiral diamine ligands bearing three stereogenic centers deriving from (R)-(+)-limonene was prepared, proving to be effective for the asymmetric transfer hydrogenation of DHIQ-type iminiums 42 (entry 5) [81].

Water soluble ruthenium/chiral diamine catalysts have emerged as some of the most attractive catalysts for the asymmetric transfer hydrogenation of iminiums. In 2006, the Zhu and Deng group realized a highly enantioselective synthesis of tertiary amines bearing a THIQ moiety via Ru-catalyzed asymmetric transfer hydrogenation (entry 2) [73]. This transformation employed RuCl₂(cymene) as a ruthenium precursor, water soluble (R,R)-N,N* f as a chiral ligand, HCOONa as the hydrogen source, CTAB as the surfactant, and H₂O as the solvent. The methodology proved suitable for both alkyl and aryl substituted substrates. Lewis acids, such as AgSbF₆, have been utilized to improve the catalytic activity and enantioselectivity of the catalyst in Pinko’s work (entry 4) [74].

3.2. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Rhodium-Catalyzed Transfer Hydrogenation

In 1999, the Baker group reported a rhodium/diamine-catalyzed asymmetric transfer hydrogenation of DHIQs 13 with HCOOH/Et₃N as a hydrogen source in organic solvents, providing THIQs with up to 96% yield and 99% ee (Scheme 21) [82]. This methodology was subsequently extended to the asymmetric transfer hydrogenation of cyclic sulfonimides 44 and acyclic N-benzyl-substituted imines with respectable reactivity and enantioselectivity.

Recently, Rimoldi and co-workers reported the rhodium-catalyzed asymmetric transfer hydrogenation of DHIQ 13 in aqueous media with chiral diamine ligand bearing an 8-amino-5,6,7,8-tetrahydoquinoline skeleton (named CAMPY) and HCOOH/Et₃N as a hydrogen donor, providing various THIQs with excellent yields and moderate ee values [83]. Additionally, Lewis acids, such as La(OTf)₃, effectively promoted the hydrogenation and the reaction reactivity was improved, apparently with a retention of stereoselectivity.

Kinetic studies on the rhodium-catalyzed asymmetric transfer hydrogenation of DHIQ were reported by Blackmond and co-workers in 2006 (Scheme 22) [84]. The initial reaction rate exhibited a linear relationship with increasing concentrations of the initial imine, which revealed that the reaction followed first-order kinetics with respect to imine. However, as the reaction progressed to 90% conversion, it exhibited zero-order kinetics in imine. The observed anomalous concentration dependences in the kinetic studies were ascribed to the equilibria between formic acid and both Et₃N and imine. A possible reaction mechanism was proposed according to the mechanistic and kinetic study of rhodium-catalyzed asymmetric transfer hydrogenation of ketones [85,86]. The rhodium catalyst 47 was converted into an active catalyst species 48 with Et₃N to remove HCl. Then, the metal-hydride 49 was formed in the presence of formic acid, transferring hydride to the protonated iminium salt 13b‧HCOOH and providing amine product 14b. Notably, Et₃N plays a crucial role not only in the formation of rhodium-hydride, but also in maintaining the acid-base equilibria by adjusting the concentration of free formic acid in the catalytic cycle.

3.3. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Iridium-Catalyzed Transfer Hydrogenation

Although significant progress has been made in the iridium-catalyzed asymmetric hydrogenation of IQ-type imines, the application of iridium catalysts in the asymmetric transfer hydrogenation is much less widely used [23]. Owing to their steric and electronic properties, TsDPENs serve as chiral diamine ligands that present excellent stereoselectivity and have been widely used in asymmetric synthesis. Unfortunately, C-1 substituted 6,7-MeO-1-phenyl-THIQ 14a was obtained in a racemic form in the presence of [Cp*IrCl(S,S)-TsDPEN] and HCOOH/Et₃N (Scheme 23) [77].

The tuning of electronic and steric effects of iridium catalysts has remarkably enhanced the stereoselectivity in the iridium-catalyzed asymmetric transfer hydrogenation of DHIQs 13. In 2017, employing [Cp*Ir(R,R)-TsDPEN] as catalyst and HCOOH as a hydrogen source, Václavíková Vilhanová and co-workers first described the iridium-catalyzed asymmetric transfer hydrogenation of DHIQs 13 [87]. With anhydrous phosphoric acid (APA, prepared from 85% H₃PO₄ and phosphorus pentoxide) as an additive, the conversion and enantioselectivity of hydrogenation were significantly improved, providing 1-aryl THIQs 14 with up to 92% yield and 86% ee.

Recently, Rimoldi reported a chiral diamine ligand (R)-CAMPY with an 8-amino-5,6,7,8-tetrahydroquinoline skeleton, which was employed in the iridium-catalyzed asymmetric transfer hydrogenation of 1-phenyl DHIQ 13a to provide the corresponding THIQ, albeit with only 45% ee [83]. Luckily, the chiral ligand was effective in the rhodium catalysis of transfer hydrogenation.

3.4. Asymmetric Synthesis of Chiral THIQs via Organocatalytic Transfer Hydrogenation

As challenging substrates, the aromaticity and lower reactivity of IQs prohibits chemists from exploring organocatalytic asymmetric transfer hydrogenation strategies. Chiral phosphoric acid (CPA), recognized as an efficient organocatalyst, has been successfully applied in the asymmetric transfer hydrogenation of unsaturated C=C, C=O, C=N and N-heteroaromatics.

Inspired by the previous work, Zhou and co-workers achieved the first asymmetric transfer hydrogenation of IQs via a substrate activation and chiral anion metathesis strategy (Scheme 24) [88]. By employing chloroformate as an activator, IQs were converted into activated N-acyl isoquinolinium chloride. Meanwhile, in the presence of an inorganic base, a chiral phosphoric acid catalyst was transformed into chiral anions, promoting the formation of contact chiral ion pair 51 through anion metathesis. With Hantzsch esters as organic hydrides, the asymmetric transfer hydrogenation of IQs 34 proceeded smoothly though the transition state 52, giving chiral 1,2-DHIQs 50 in high yields and good enantioselectivities. The products could be further transformed into chiral THIQs 14 via the reduction of C=C bonds.

4. Asymmetric Synthesis of Chiral THIQs via Transition-Metal-Catalyzed Reductive Amination

Direct asymmetric reductive amination represents a practical and effective methodology to enantiomerically pure amines [67,89,90,91,92]. Extensive research has significantly expanded this protocol to the synthesis of various chiral compounds, such as chiral THIQs, γ-sultams [93,94], hydrazines [95], linear amines [96], etc.

Intramolecular asymmetric reductive amination has been successfully applied for the construction of chiral amines bearing a THIQ framework 14 via the asymmetric hydrogenation/transfer hydrogenation of the in situ generated imines/iminiums 13 derived from the condensation of amino and carbonyl group of substrates 54 (Scheme 25). However, substrate protection is necessary, owing to the high activity of amino groups toward carbonyl groups. The Boc group serves as an excellent protective group (PG) for amino functionalities and is widely utilized in organic synthesis. The protecting group of substrates 53 could be removed under Brønsted acid conditions, providing a free amino group. The reaction then experienced the in-situ formation of imine, followed by the hydrogenation/transfer hydrogenation of imine intermediate. The stereoselective control of reductive amination depended on the asymmetric reduction of the imine intermediate. Therefore, reductive amination involves a cascade process of deprotection/cyclization/asymmetric reduction.

4.1. Ruthenium-Catalyzed Asymmetric Reductive Amination

In 2003, Wills and co-workers reported the first intramolecular asymmetric reductive amination for the synthesis of THIQs via the Ru-catalyzed asymmetric transfer hydrogenation of imine intermediates (Scheme 26) [97]. With N-Boc protected keto amines 53b as substrates, (S)-6,7-dimethoxy-1-methyl-1,2,3,4-tetrahydroisoquinoline 14b was obtained with 85% yield and 88% ee. Formic acid acts as a Brønsted acid to facilitate the efficient removal of Boc groups in the deprotection process, leading to the formation of imines. Meanwhile, along with triethylamine (a volume ratio of HCOOH/Et₃N in 5/2), formic acid also functions as a hydrogen source for a compatible asymmetric transfer hydrogenation process.

4.2. Iridium-Catalyzed Asymmetric Reductive Amination

With rapid progress in the field of iridium-catalyzed asymmetric hydrogenation of imines containing a THIQ core, asymmetric catalytic reductive amination for the synthesis of THIQs has attracted much attention from chemists (Scheme 27).

In 2017, Chang and co-workers described a one-pot synthesis of enantiopure THIQs via a cascade process of TFA-promoted N-Boc deprotection of substrate 53 and iodine-bridged dimeric iridium-catalyzed asymmetric reductive amination [98]. With Ti(OⁱPr)₄ and iodine accelerating the hydrogenation of imine intermediate and TsOH improving the enantioselectivity, a range of enantiomerically pure THIQs was obtained with excellent yields and ee values. Subsequently, Zhang et al. employed a novel ferrocenyl bisphosphine scaffold ZhaoPhos as a chiral ligand, achieving iridium-catalyzed one-pot enantioselective synthesis of THIQs and THQs via N-Boc deprotection/intramolecular asymmetric reductive amination of substrate 53 [99]. However, the substrate scope was limited to the preparation of 1-aryl substituted THIQs. Recently, with a chiral ferrocenyl bisphosphine ligand JosiPhos, the Jiang and Nie group further expended this protocol for the chiral synthesis of THIQs with an alkyl group at C-1 position via iridium-catalyzed asymmetric reductive amination [100].

5. Asymmetric Synthesis of Chiral Tetrahydroisoquinolines via Deracemization

Classical kinetic resolution generally utilizes chiral resolving agents, such as chiral acids, chiral bases, or chiral amino acids, to facilitate the efficient separation of enantiomers via the formation of diastereomers. Chiral acids and amino acids have been designed and utilized for the kinetic resolution of amines (

Table 3. Kinetic resolution of THIQs.

Entry	Compound	Chiral Acid	ee	Yield a	Entry	Compound	Chiral Acid	ee	Yield a
1			>98% (S)	33%	2			>98% (S)	20%
3			>98% (S)	45%	4			>98% (R)	41%
5			>98% (S)	40%	6			>98% (S)	38%

^a The yield is based on sum of enantiomers of a substrate.

). In 1929, the Leithe group employed (−)-(D)-tartaric acid as a resolving agent for the kinetic resolution of racemic 1-phenyl-1,2,3,4-THIQ 14f, giving (S)-enantiomer with a yield of 33% and an enantiomeric excess exceeding 98% [101]. Besides, (−)-(D)-mandelic acid [102], N-Ts-(L)-phenylalanine [103], and N-Ph-(L)-phenylalanine have also been used for the kinetic resolution of alkaloids with a THIQ core, providing enantiomerically pure (S)-(−)-norcryptostyline I, (R)-(+)-salsolidine, and (S)-(−)-norlaudanosine with excellent ee values. Approximately half equivalent amount of resolving agents were used in the kinetic resolution, and only 50% theoretical yield was obtained for the desired enantiomer, resulting in the waste of chiral resources.

Despite the limitations of kinetic resolution in terms of theoretical yield, synthetic methodologies, including dynamic kinetic resolution (DKR), dynamic kinetic asymmetric transformation (DYKAT), and deracemization, were developed for the enantioselective synthesis of amines deriving from racemates with 100% theoretical yield in recent years (Scheme 28). Deracemization refers to a chemical process that completely converts a racemic mixture into a single enantiomer of the same compound without the separation of intermediates [104]. As a highly efficient technology, deracemization has significant advantages over classical kinetic resolution in the following respects: (1) 100% theoretical yield and atom economy; and (2) one-pot, single-operation process without separation of intermediates or removal of chiral resolving agents.

Despite much progress having been achieved [105,106,107,108,109], the deracemization of racemates remains one of the most challenging problems in organic synthesis. Deracemization is generally realized via the destruction and reconstruction of stereogenic centers, which involves two fundamentally opposite processes in the reaction direction and mechanism. Deprotonation–protonation and oxidation-reduction are the most commonly used strategies for deracemization for amines and alcohols (Scheme 29). In a deprotonation–protonation strategy, a Brønsted base serves as an elegant deprotonating agent that extracts protons from C-H bonds to destroy the stereogenic center and form carbanion intermediate 55. Then, the regeneration of stereogenic center is achieved through the protonation of carbanion intermediate 55 with a Brønsted acid as a protonating agent. In oxidation–reduction strategis, imines and iminiums are crucial prochiral intermediates for deracemization. The combination of amine oxidation and prochiral imine/iminium intermediates 56 reduction enables the deracemization of racemic amines. However, the direct quenching between Brønsted bases and Brønsted acids, oxidants, and reductants makes the deracemization more challenging. To overcome this problem, various strategies have been developed, including sequential operation or physical isolation. Sequential operation often requires a full conversion of the substrate in the first step. Physical isolation, including multiphase separation and cell wall separation, often proceeds in different areas and requires excellent stereoselectivity. Physical isolation with the cell wall has been widely applied in the enzymatic deracemization [110,111].

5.1. Deracemization via a Deprotonation-Protonation Strategy

Deprotonation–protonation (also known as acid-base neutralization) strategies are among of the most important and straightforward methodologies for deracemization. The destruction and reconstruction of chirality of the stereogenic center were achieved via a cascade process of deprotonation and protonation. Due to the weak acidity of protons in C-H bonds, the deprotonation process proceeds under strong alkaline conditions, such as ^sBuLi, ^tBuLi, and so on. Then, the protonation process is conducted with various proton sources, such as alcohols, phenols, amines, acids, H₂O etc. Considering a strong alkaline, and the reaction tendency between the base and protonating agents, the deprotonation and protonation should be separated via a sequential operation or physical isolation.

From the perspective of the construction of chirality in a stereogenic center, deracemization is realized through distinct strategies (Scheme 30). In the first case, the enrichment of chirality occurs during the deprotonation process. With a chiral base to extract proton from substrate rac-31, contact chiral ion pair 57 forms gradually. Followed by a protonation step, protonation occurs in the presence of an achiral proton source, resulting in the production of chiral compound 31 with the retention of chirality. In the second case, the enrichment of chirality occurs during the protonation process. The ion pair 58 is quenched with chiral alcohols, phenols, amines, acids, amino acids, etc., and this protonation process is a stereo-control step.

Early in 1998, Levacher and co-workers reported the (−)-sparteine-mediated deracemization of N-methyl-4-phenyl-THIQ 31 through a process involving deprotonation and protonation (Scheme 31) [112]. This deracemization involved s-BuLi/(−)-sparteine-mediated deprotonation and proton source-promoted protonation. The deprotonation was conducted under a low temperature (−45 °C) in the presence of s-BuLi as a strong base and (−)-sparteine as a chiral ligand, ensuring the complete lithiation of THIQ at the C-4 position and high enantioselectivity. Then, chiral organolithium intermediate 59 was restored through protonation using proton sources (alcohols, H₂O etc.), affording the product under low-temperature conditions. An isotopic labeling experiment (protonation with MeOD) revealed that the formation of the stereogenic center occurred during the deprotonation step. However, the poor incorporation of deuterium into THIQ d-(S)-31 at C-4 position resulted in poor enantioselectivity. Luckily, an impressive 95% incorporation of deuterium was observed during deracemization. Further deprotonation at the C-1 position was also observed at −78 °C to provide diastereomers after quenching with MeOD, which could be extended to the deracemization of C-1 substituted THIQs. The deracemization used a pyramidal carbonion intermediate [113]. This methodology could also be applied to the deracemization of diarylmethanes, such as 2-(1-phenylethyl) pyridine 60. The results showed that moderate enantioselectivities were obtained with EtOH (S configuration) and ^tBuOH (R configuration) as proton sources, respectively.

Further investigation into this methodology focused on the deracemization of diarylmethanes. In 2001, the Levacher group achieved the deracemization of 2-(1-phenylethyl) pyridine 60 with (−)-sparteine or a chiral protonating agent via two completely different modes of chirality generation [114]. (1) With (−)-sparteine as chiral ligand, the stereogenic center was regenerated either in the deprotonation process via a dynamic kinetic resolution process or in the protonation process via an asymmetric protonation process [112]. (2) With chiral diamines 62 as a proton source, the stereogenic center was regenerated in the protonation process.

5.2. Deracemization via a Redox Strategy

Deracemization via a redox strategy includes two opposite processes in terms of the reaction direction and mechanistic pathway. The destruction and reconstruction of stereogenic centers may be realized through the combination of oxidation and reduction processes. The most challenging problem is the direct quenching between the oxidant and reactant.

The deracemization of amines was achieved through two distinct reaction modes, including linear redox deracemization and cyclic redox deracemization (Scheme 32). In linear redox deracemization mode, racemic amines (rac-31) were converted to prochiral imines/iminiums 30 via non-selectively oxidative dehydrogenation, followed by a selectively asymmetric reduction to provide chiral amines (chiral-31). The enrichment of chirality occurred during the reduction process. The thorough oxidation of amines and efficient stereoselective control of reduction contributed to the excellent observed enantioselectivity. In contrast, in cyclic redox deracemization mode, racemic amines (rac-31) experienced selective oxidative dehydrogenation to give the corresponding prochiral imines/iminium salts 30 and single enantiomer of amines (chiral-31). Then, a non-selective reduction of these prochiral intermediates proceeded to yield the racemic amines (rac-31), and a single enantiomer of amines was accumulated in the cyclic redox deracemization. As the number of cycles increased, the enantiomeric excess values of the products improved remarkably, which revealed superior results to linear redox deracemization mode. The enrichment of chirality was accumulated in each cycle. Cyclic redox deracemization has been widely used in the enzymic deracemization of amines.

In 2015, Zhou and co-workers reported the deracemization of secondary and tertiary amines bearing a THIQ core via a compatible redox strategy, yielding chiral THIQs 14 with up to 95% yield and 98% ee (Scheme 33) [115]. Deracemization involves a cascade process, comprising NBS oxidation of THIQs and iridium-catalyzed asymmetric hydrogenation of DHIQ intermediates. The excellent observed enantioselectivity could be attributed to the dual activation role of NBS, i.e., (1) as an oxidizing halogen agent, NBS could oxidize THIQs to DHIQs efficiently and thoroughly; and (2) NBS could also improve the catalytic performance of the iridium catalyst by facilitating its conversion from Ir^I to active Ir^III, thereby improving both the activity and enantioselectivity. Experimental research was conducted for mechanistic studies. With 1-phenyl-DHIQ 13f as a substrate, the reaction occurred smoothly under standard conditions to give (R)-14f with 98% ee. When (S)-14f was subjected to standard conditions, a reversal phenomenon of chirality was observed, yielding ((R)-14f) in 98% ee. An isotopic labeling experiment with D₂ gas (100 psi) also indicated that DHIQs (13) are important intermediates of deracemization, which are in situ generated in the NBS oxidation of THIQs. The high deuterium incorporation indicated the thorough oxidation of THIQs, which was one of the most critical factors for the perfect enantioselectivity.

6. Conclusions

This review has highlighted recent advances in the enantioselective synthesis of THIQs through the asymmetric reduction of N-heteroaromatics, imines, enamines, and iminium salts using either organometallic or organic catalysts. Strategies such as asymmetric hydrogenation, and transfer hydrogenation, reductive amination, deracemization strategies have been developed to synthesize enantiopure THIQs with exceptional reactivity and enantioselectivity. Despite significant progress in the enantioselective synthesis of THIQs via asymmetric reduction, there still remain great opportunities and challenges for progress and developments in this field of research. Future efforts may concentrate on the following aspects: (1) developing novel homogenous catalytic systems for the direct asymmetric reduction of IQs without substrate activation; (2) designing efficient heterogenous chiral catalysts that emphasize recyclability and reusability; (3) enhancing both the efficiency and stereoselectivity of catalysts; (4) expanding the generality of an asymmetric reduction strategy to a range of more challenging heteroaromatics, imines, enamines, and iminium salts. We believe that more breakthroughs will be achieved, along with potential applications in the asymmetric reduction of various unsaturated compounds.