Recent Development in the Synthesis and Catalytic Application of Iridacycles

Cyclometallated complexes are well‐known and have found many applications. This article provides a short review on the progress made in the synthesis and application to catalysis of cyclometallated half‐sandwich Cp*Ir(III) complexes (Cp*: pentamethylcyclopentadienyl) since 2017. Covered in the review are iridacycles featuring conventional C,N chelates and less common metallocene and carbene‐derived C,N and C,C ligands. This is followed by an overview of the studies of their applications in catalysis ranging from asymmetric hydrogenation, transfer hydrogenation, hydrosilylation to dehydrogenation.


Introduction
The term "cyclometallation" was defined by Trofimenko [1] and this type of reactions has been known since the early 1960s. A lot of transition metals, such as Pd, Ru, Rh and Ir, have been studied for this reaction to provide metal complexes having chelating rings embedded with a metalÀ carbon bond. [2] The use of iridium in cyclometallation leads to iridacycles. Iridacycles have found applications in a diverse range of areas, such as medicine, [3][4][5] organic light emitting diodes [6] and bioimaging/sensing applications. [7] In particular, they are active as catalysts for various reactions, such as hydrogenation, transfer hydrogenation (TH), reductive amination, dehydrogenation, hydrosilylation, oxidation, racemisation, etc.. [8] Cyclometallation reactions mostly occur through CÀ H activation. The use of acetate in CÀ H activation provides facile access to cyclometallated complexes. Shaw and co-workers made the pioneering studies, showing that sodium acetate promotes cyclometallation reactions. [9] Later in 1998, Beck et al. performed an initial study of cycloiridation in the presence of sodium acetate, reacting 2-phenyl-4-R-5(4H)oxazolones (R = Me, CH 2 Ph, CHMeEt) with the iridium dimer [IrCl 2 Cp*] 2 . [10] In 2003, Davies 2 , and [RuCl 2 (p-cymene)] 2 in the presence of NaOAc at room temperature. [11] More recently, Muller et al. isolated and crystallographically characterised phosphinine-Ir(III)/Rh(III) complexes via an unprecedented CÀ H activation of 2,4,6-triphenylphosphinine with [MCl 2 Cp*] 2 (M = Ir or Rh). [12] The role of acetate in CÀ H activation was established in subsequent studies, both computationally and experimentally, by Davies, Jones, Zheng and Muller. [13] Although there are some early reports of cyclometallation with iridium, [14] the use of iridacycles in catalysis is later. Scheme 1 shows selected half-sandwich Cp*-iridacycles, with their applications in catalysis highlighted. For example, in 2007, Peris and Fernandez et al. first reported catalytic diboration of olefins with iridium complex 1, giving organodiboranates with high conversions and chemoselectivities. [15] In 2008, Ikariya et al. reported asymmetric transfer hydrogenation (ATH) of acetophenone with chiral iridium complex 2, giving (S)-1-phenylethanol in high yield (95 %) and moderate ee (66 %). [16] In 2009, de Vries, Feringa et al. reported the use of cationic half-sandwich iridacycles, such as complex 3, as racemisation catalysts for chiral alcohols and amines. [17] In 2009, Eisenstein, Brudvig and Crabtree et al. reported Cp*Ir complexes 4 for catalytic water oxidation. [18] In 2011, Fujita and Yamaguchi et al. developed the α-hydroxypyridine-ligated Cp*Ir complex 5 for dehydrogenative oxidation of primary and secondary alcohols, yielding aldehydes and ketones, respectively. [19] In 2012, Morris et al. reported that complex 6 could catalyse the hydrogenation of acetophenone and benzophenone. [20] In the same year, Djukic et al. reported the catalytic activity of Cr(CO) 3 -bound half-sandwich iridacycle 7 in tandem transformation of terminal alkynes into N-phenylamines, including hydroamination and hydrosilylation/protodesilylation reactions under mild conditions. [21] In 2014, Sarkar et al. reported that complex 8 catalyses TH of benzaldehyde to form benzylalcohol in isopropanol/KOH; but the TH of acetophenone was slower under the same conditions. [22] In 2010, our group reported a highly effective catalyst 9, which enables transfer hydrogenative reductive amination of various carbonyl compounds. [23] In the following several years, we and others have found that iridacycles comprised of such simple imino ligands catalyse a number of interesting reactions. The progress made prior to 2017 has been summarised. [8] The aim of this minireview article is to show new iridacycles that have appeared and their applications in catalysis since 2017. Earlier publications are mentioned where appropriate. The coverage is restricted to half-sandwich iridacycles, i. e. those bearing Cp* and related ligands.

Conventional C,N Ligands
Aiming for asymmetric catalysis, a number of chiral iridacycles have been synthesised in the past a few years. In 2018, Leung et al. synthesised 1-naphthylalkylamine-based chiral iridacycles 10 and 11 by cycloiridation of the optically-active secondary amine ligand (S)-N-methyl-1-(naphthalen-1-yl)ethan-1-amine with [IrCp*Cl 2 ] 2 , with each iridacycle observed as only one stereoisomer (Scheme 2). [24] This observation is notable, because many Cp*-based iridacycles are mixtures of diastereomers upon cyclometallation. For example, whilst complex 10 has three stereogenic centres, only the isomer indicated was observed. This is believed to be due to a structurally locking mechanism. The demethylation product 11 was isolated as a by-product with a low yield; but it could be obtained from direct cyclometallation of the primary amine with a high yield of 90 %. In comparison, the phenyl derivative gave the racemic imino-iridacycle 12 as the major product, with a moderate yield of 42 % (Scheme 2).
In the same year, the same group carried out the reaction of (R)-N,N-dimethyl-1-(1-naphthyl)ethylamine with [IrCp*Cl 2 ] 2 , employing sodium acetate in 1,2-dichloroethane (DCE) at reflux temperature, in order to gain more insight into the selectivity of cyclometallation. [25] This reaction generated six isolable products: cycloiridated complexes 13, 15, and 16, dimethylamine coordinated complex 14 and organic molecules 1-acetonaphthone and 1-ethylnaphthalene, as shown in Scheme 3. It was suggested that the formation of 13, 14 and 1-acetonaphthone could start from a β-hydride elimination-like process, 15 could be derived from 13 via a dehydrogenation procedure, and 16 could be formed from the dehydroamination of a complexed cyclic intermediate, hydrogenation of which could produce 1-ethylnaphthalene. To investigate the steric effect on the reaction pathway, the authors carried out cycloiridation of a phenyl derivative and an achiral naphthalene-based ligand, as shown in Scheme 4. The reaction of the phenyl ligand under the similar conditions gave complex 14, acetophenone and racemic compound 17, while the achiral naphthalene ligand provided the racemic com-pound 18. These results indicate that steric effects in the coordination sphere favour a β-hydride elimination-like pathway, competing with ortho-metalation.
In 2018, we demonstrated the importance of reaction conditions on ligand coordination mode in attempted cyclometallation. [26] The reaction was carried out with [IrCp*Cl 2 ] 2 and methyl (S)-2-phenyl-4,5-dihydrooxazole-4carboxylate in the presence of anhydrous sodium acetate at room temperature, affording a mixture of two half-sandwich Cp*Ir(III) complexes, the expected N,C-chelated complex 19 and an "abnormal" N,O-chelated complex 20 (Scheme 5). Interestingly, a simple change in reaction conditions, such as water content (in the solvent) and additive, affected the ratio of these complexes in a dramatic way (up to > 99 : 1 ratio for each complex). Each complex is air stable and appears as a mixture of two diastereomers.
We have recently reported a range of chiral iridacycles, obtained by cyclometalation of [Cp*IrCl 2 ] 2 with chiral oxazoline and imidazoline ligands via CÀ H bond activation under generally mild reaction conditions. [27] These iridacycles appeared mostly as a mixture of two diastereoisomers, presumably due to the small difference between their thermodynamic stability. Some examples (21)(22)(23)(24)(25)(26) are shown in Scheme 6.
In 2020, Cramer et al. reported air and moisture-stable C,N-chelated Ir(III) complexes bearing a chiral cyclopentadienyl ligand. [28] The complexes were synthesised via a twostep sequence (Scheme 7). Some complexes are presented in Scheme 7c.

Metallocene-based Ligands
Ferrocene-featuring iridium complexes have long been known. [29] Among them, ferrocene-based planar chiral iridacycles was reported by Richards and co-workers in 2016. [30] The work was extended to planar chiral iridacycles featuring oxazoline-appended cobalt sandwich complexes in 2018. [31] Thus, diastereoselective cycloiridation of cobalt oxazoline and ferrocenyloxazoline gave iridacycles 33-36, as shown in Scheme 8. The influence of base, water, time, and auxiliary at the oxazoline on the cycloiridation was studied. For example, when using iso-propyl (i-Pr) oxazoline, the selectivity of cycloiridation in the absence of water resulted in a higher selectivity (33 : 34 = 4.8 : 1) with a lower yield (39 %) than in the presence of water (33 : 34 = 2.9 : 1 with 78 % yield) (Scheme 8a). But with the tert-butyl (t-Bu) oxazoline, the diastereoselectivity was not affected by the addition of water, and only the yield increased from 29 % to 45 % (Scheme 8c). Notably, the iridacycles 33 and 34 could be converted into an iridocenium cation in deuterated acetonitrile on heating at 45°C. This iridocenium cation was obtained in their previous work in 2016 [32] from a ferrocene-derived iridacycle, as shown in Scheme 9a. Also, stereospecific ligand exchange with chloride and triphenylphosphine at iridium was performed, providing neutral chloride-ligated diastereoisomers and cationic isomers (37, 38 and 39, 40), respectively, as shown in Scheme 10.  In 2017, Richards et al. reported iridacycles via cycloiridation of ferrocenyl and phenyl-containing imines with [Cp*IrCl 2 ] 2 in the presence of NaOAc in CH 2 Cl 2 (Scheme 11). [33] They studied the exo vs. endo and phenyl vs. ferrocenyl selectivity arising from α or ortho CÀ H bond activation, observing, interestingly, that the iridacycles resulted from ortho-phenyl instead of α-ferrocenyl CÀ H activation. The complexes derived from FcCH=NCH(R)Ph (R = H, Me) were a mixture of E and Z imine diastereoisomers, and while the ligands FcCH(R)N=CHPh were endo selective, the FcCH=NCH(R)Ph were exo selective (Fc: ferrocenyl; Ph: phenyl).
More recently, Richards et al. reported the synthesis of ferrocene-based planar chiral iridacycles by using transmetallation as an alternative approach to cycloiridation (Scheme 12). [34] This has made possible previously unavailable iridacycles by direct introduction of iridium via CÀ H activation. In the presence of tetramethylethylenediamine (TMEDA) and sec-butyllithium (s-BuLi), diastereoselective lithiation of ferrocenyloxazoline followed by the addition of HgCl 2 resulted in mercury-substituted ferrocenyloxazoline 45 with d.r. > 99 : 1 (Scheme 12). In comparison, direct cyclomercuration of ferrocenyloxazoline by CÀ H activation resulted in a mixture of two mercury-substituted complexes (45 and its S,R p isomer) with a very low diastereoselectivity (d.r. = 1.1-1.3) (not shown in scheme). Then, transmetallation of 45 with the iridium dimer in the presence of tetrabutylammonium chloride (TBAC) in acetone resulted in iridacycle 46 [30] as only a single diastereoisomer (d.r. > 99 : 1). Transmetallation of 45 in the presence of KPF 6 in acetonitrile produced cationic complex 47, affording again a single diastereoisomer (d.r. > 99 : 1), as shown in Scheme 12. Complex 47 was also produced via ligand exchange from 46 by using KPF 6 in acetonitrile with d.r. > 99 : 1. Mercuration of a cobalt sandwich analogue of ferrocenyloxazoline by the use of Hg(OAc) 2 led to a mixture of isomeric mercury complexes with d.r. = 2 : 1 (S,S p and S,R p ). Transmetallation of this mixture with [Cp*IrCl 2 ] 2 in the presence of KPF 6 in acetonitrile gave a 2 : 1 ratio of cationic iridacycles, but with a low yield (not shown in scheme). The group also synthesised deuterated ferrocenyloxazoline-derived iridacycles 49 and 48 with d.r. > 99 : 1 for each. Lithiation of ferrocenyloxazoline, followed by a sequence of addition of MeOD-d 4 , second lithiation, mercuration and finally transmetallation with [Cp*IrCl 2 ] 2 , afforded 48 in the presence of KPF 6 and 49 in the presence of TBAC (Scheme 13). A similar approach was used to form the chiral iridacycle 50 (Scheme 14a) derived from Ugi's amine. The application of lithiation to 51, a bulky cobalt sandwich complex analogous to Ugi's amine, followed by the addition of excess HgCl 2 , led to 52 (Scheme 14b). However, transmetallation of 52 with [Cp*IrCl 2 ] 2 was found unsuccessful.

Carbene-based Ligands
N-Heterocyclic carbenes (NHCs) provide another easy entry to cyclometallated complexes, those featuring a C,C-chelate ring. In 2017, Sierra and de la Torre et al. reported the application of 1,2,3-triazolylidine mesoionic carbenes (MIC) bearing enantiopure sulfoxide moieties to the synthesis of chiral cyclometallated Ir(III) and Rh(III) half-sandwich complexes (Scheme 15). [35] Treatment of the enantiopure sulfinyl triazolium salt with Ag 2 O in the presence of NMe 4 Cl and 4 Å MS, followed by transmetallation with [Cp*IrCl 2 ] 2 , gave the neutral Ir(III) complex 53. On treatment of 53 in DCE with NaOAc, cyclometallation took place, affording iridacycle 54 (Scheme 15a). The five and six membered sulfone analogues 56 and 57 and sulfoximine analogue 58 were prepared in a similar fashion (Scheme 15b and 15c). The enantiopure cationic Ir(III) complexes, such as 55, are easily accessible (Scheme 15a), and further, insertion of alkynes into the IrÀ C bond was demonstrated, as seen in the conversion of 59 to 60 (Scheme 16). The process of MIC coordination and CÀ H aromatic activation was found effective and diastereoselective for the formation of enantiopure five-membered metallacycles, whilst the formation of six-membered analogues exhibited a considerably lower diastereoselectivity.
More recently, the same group reported planar chiral bimetallic ruthenocene or ferrocene-derived iridacycles incorporating sulfoxide-substituted mesoionic carbenes (Scheme 17). [36] These iridacycles contain three different types of chirality, sulfur-based point chirality, iridium-centred metal chirality and metallocene-imposed planar chirality. The reactions leading to the formation of complexes 61-66 are shown in Scheme 17. It was observed that whilst the sense of asymmetric induction at Ir(III) remained the same, the planar chirality depended on the nature of the ligands at the iridium centre (carbonate vs chloride). For instance, while 61 was isolated as a single isomer from the Ir-carbonate intermediate in a high yield, the dichloro-Ir yielded a mixture of diastereomeric 61 and 62.

Asymmetric Reduction
Asymmetric hydrogenation (AH) and asymmetric transfer hydrogenation (ATH) have been extensively used to introduce chirality into prochiral molecules, such as enamines, imines, ketones, alkenes and unsaturated carboxylic acids. AH uses hydrogen gas as hydrogen donor while ATH commonly uses isopropanol and formic acid as hydrogen source. In comparison with AH where high pressure H 2 is often required, ATH is generally easier to operate, with no need for special equipment or measures. Both methods of reduction have found wide applications in the synthesis of fine chemicals, pharmaceuticals and agrochemicals. Iridium-catalysed AH or ATH reactions are well-known. The interested reader is referred to selected reviews [37] and articles. [38] However, the use of half-sandwich Cp* type iridacycle catalysts in these reactions is less explored. [39] Recent examples are presented below.
In 2020, Cramer et al. reported an efficient AH of oximes with the iridacycle 32, producing chiral hydroxylamines. [28] The reduction of oximes to chiral hydroxylamines is challenging due to the possible cleavage of the NÀ O bond leading to undesirable products. The iridium complex 32 in the presence of a strong acid serves as an efficient catalyst for this reduction without disturbing the NÀ O bond. The hydrogenation was carried out at room temperature, 1 mol% catalyst, 50 bar pressure in the presence of methanesulfonic acid (MsOH) giving high yields and high enantiomeric ratio (er) for various hydroxylamines, as shown in Scheme 18. A comparison of the analogous iridacycles 27-31 in the AH of an oxime is shown in Scheme 19.
Earlier, Leung et al. reported the use of iridacycles 10 and 11 for ATH of acetophenone (Scheme 20). [24] Both complexes provided high conversions (ca. 90 %) and modest ee (60 %) at À 15°C in 30 min. When the ATH reaction was performed with 10 at a lower temperature (-30°C), the enantioselectivity was improved to 69 % ee.
We compared the N,C-and N,O-chelated Cp*Ir(III) complexes 19 and 20 for ATH of substituted acetophenones. [26] The N,O-chelated 20 exhibited much higher catalytic activity and enantioselectivity than the N,Canalogue 19, with differing sense of asymmetric induction, as exemplified in Scheme 21. Thus, complex 20 was applied to ATH of various aromatic ketones, providing high enantioselectivities (90-99 % ee) as shown in Scheme 22. These results reveal that the chelation mode of a ligand can dramatically affect the outcome of ATH reactions. The suggested mechanism for ATH with 20 includes the formation of iridiumhydride and participation of isopropylammonium cation in the transition state of enantioselectivity-determining step. A hydrogen bonding network was suggested, which would be expected to lower the barrier of the diastereomeric transition state, enhancing the enantioselectivity of the hydride transfer.
In a related work, we examined the catalytic activity of a series of oxazoline and imidazoline-based iridacycles in direct asymmetric reductive amination (DARA) of acetophenone with p-methoxy aniline and in ATH of N-benzyl pyridinium bromide salts. [27] Among them, the 3,4,5-trimethoxy imidazoline-bearing iridacycle, complex 23, was found the most effective catalyst for DARA. Applying 23 to DARA of a range of ketones with amines afforded chiral amines with high yields and moderate enantioselectivities in most cases. Some chiral amine products are presented in Scheme 23. The dioxole iridacycle, complex 24, provided the best enantioselectivity (up to 77 % ee) for ATH of pyridiniums by using HCO 2 H/Et 3 N in isopropanol (Scheme 24).
Iridacycle 49 was investigated as a catalyst for reductive amination of acetophenone with benzylamine. However, the catalyst showed no asymmetric induction (Scheme 25). [34] More recently, Leung et al. [40] explored the iridacycle below, an isomer of 10, for the catalytic asymmetric 1,2dihydrophosphination of acrylonitrile, obtaining a remarkably high ee of 99 %. However, the conversion was low under the conditions employed even after three days of reaction (30 %), as shown in Scheme 26.

Achiral Hydrogenation
A great number of chiral but racemic iridacycles have been synthesized in the past, some of which have been explored for achiral hydrogenation with H 2 . The recent results are summarized below. The Choudhury group reported an "abnormal NHC"-derived iridacycle [41] for atmospheric pres-Scheme 17. Synthesis of chiral ferrocene and ruthenocene-derived iridacycles.
sure hydrogenation of various aldehydes. [42] Based on their previous studies, [43,44] the free, basic imino-nitrogen from the benzimidazoleto moiety is prone to protonation and thus has the ability to capture the proton generated from heterolytic H 2 splitting. Additionally, the imidazole-5-ylidene moiety, an abnormal NHC with strong σ donor capacity, facilitates hydride transfer. The bifunctional iridium catalyst showed efficient, chemoselective hydrogenation of aldehydes in aque- Later, Choudhury et al. reported a bifunctional iridium complex bearing an uracil-abnormal NHC ligand, which undergoes acid/base switchable hydrogenation catalysis. [45] Inspired by nature, chemists have long been interested in the development of catalysts whose activity could be influenced by external chemical or physical stimuli, such as light, pH, metal ion coordination and mechanical forces. [46] In this work, the   [47,48] In 2018, Deng et al. reported benzothiazole-derived iridacycles for hydrogenation of imines and quinolines. [49] The airstable C,N-chelated iridium complexes were prepared via CÀ H activation of benzothiazole ligands with [Cp*IrCl 2 ] 2 ; no C,Schelated product obtained. Various imines and quinolines were hydrogenated in high yields (Scheme 29). In the presence of AgOTf, the yields were significantly improved.

R e c o r d R e v i e w T H E C H E M I C A L R E C O R D
In 2019, Hartwig et al. reported a linear-selective hydroaminomethylation of α-olefines with a dual-catalytic system (Scheme 30). [50] The transformation comprised two steps, hydroformylation catalysed with a rhodiumÀ diphosphine complex followed by reductive amination enabled with an iridacycle (68), which we reported earlier. [51] The latter step was conducted under transfer hydrogenation conditions, however, using sodium formate as reducing agent. Notably, the iridacycle 68 was not poisoned by carbon monoxide. Aromatic, heteroaromatic, and aliphatic amines were obtained with high yields (up to 88 %) and regioselectivities (90 : 10 to > 99 : 1 n:iso ratios).

Achiral Transfer Hydrogenation
TH has attracted enormous attention due to its simple and safe operation compared to direct hydrogenation using molecular H 2 . [52] Previously, our group reported iridacycle-catalysed TH of carbonyl compounds in water at optimal pH values. [51,53] In 2017, the Albrecht group reported synthesis of O-functionalised triazolylidene Ir(III) complexes (Scheme 31). [54] The activity of the iridium complexes 69-70 was examined in TH of a model substrate benzophenone with isopropanol as hydrogen source.   [55] NHC-bearing transition metal complexes have been successfully applied to numerous catalytic TH reactions. [56][57][58][59][60][61][62][63][64][65][66][67][68][69][70] In 2015, Gulcemal et al. reported a new phenoxide chelated Cp*Ir(NHC)complex for TH reductive amination under aqueous conditions. [71] As was shown before, [72] the catalytic activity was found highly affected by the structure of iridium complexes. Based on these studies, the Gulcemal group synthesized a set of new O-aryloxide chelated NHCÀ Ir complexes, which gave high yields for reductive amination of various carbonyls in water. [73] Scheme 32 shows the previously reported IrÀ NHC complex 71 and modified variants 72-78. The complex 73 with an imidazole moiety gave better conversions than the benzimidazole complex 72. Further modifications revealed that complex 78, where the phenyl ring is replaced by a naphthyl moiety, showed the highest catalytic activity. According to 1 H NMR, the higher activity of the Onaphthyl-based catalyst 78 than the O-phenoxide analogue 73 could be attributed to the faster formation of iridium hydride with the former during catalysis. Interestingly, the formation of the cyclometallated product 73-CM, which was detected by 1 H NMR, slowed the metal hydride formation (Scheme 33).
Most of the NHCÀ metal complexes are mononuclear. [74,75] The Graiff group developed binuclear iridiumÀ NHC complexes, which showed good activities in the TH of ketones. [76] In these iridium complexes, a dicarbene ligand coordinates to two iridium centres in a bridging manner. The complexes 79 and 80 catalysed TH of various ketones in the presence of isopropanol and NaOiÀ Pr (Scheme 34). Interestingly, whilst complex 79 showed high activity for alkyl, aryl and cyclic  By introducing highly hydrophilic substituents, the Kuhn group reported novel water-soluble metalÀ NHC catalysts. [77] The air-stable metal complexes were synthesized via transmetallation from sulfonated AgÀ NHC (Scheme 35). The catalytic TH of acetophenone with these water-soluble complexes was tested in HCO 2 H/HCO 2 Na at 80°C. The rhodium complex 83 was shown to hydrogenate the substrate, affording 96 % yield in 10 min. Ruthenium 81 and iridium 84 took 7 h to reach 87 % and 92 % yields, respectively, whilst the osmium complex 82 afforded only 3 % yield in 7 h.

R e c o r d R e v i e w T H E C H E M I C A L R E C O R D
In 2017, the Albrecht group reported a series of ditriazolylideneÀ iridacycle complexes containing alkyl or ether linkers with different lengths, and studied the effect of the linkers on the catalytic performance in TH. [78] Scheme 36 shows the synthesis of mono-and bi-metallic iridium complexes with the ditriazolylidene ligands. The monocarbene Ir(III) complex 86 was formed when 1 equivalent of Ag 2 O was used, while the bidentateÀ NHC complex 87 could be synthesised with additional Ag 2 O. In the presence of the iridium dimer and Ag 2 O, the bimetallic complexes 88 was formed from 86. These complexes were examined as catalyst precursors in the TH of benzophenone using isopropanol as hydrogen donor. As shown in Scheme 37, there appears to be    [79] PYAs have an electronically flexible nitrogen donor that can act as a π-acidic imine or as a π-basic pyridinium amide when coordinated to a metal (Scheme 38a). The simple synthetic procedures for PYAs allow for easy modification of their electronic and steric properties, which could be harnessed to improve the catalytic activity. Air-stable iridium complexes 89 were isolated from the reaction of [IrCp*Cl 2 ] 2 with the corresponding ligands (Scheme 38b). In the catalytic TH of benzophenone, higher conversions were achieved with the PYA iridacycles compared with [IrCp*Cl 2 ] 2 , and the PYA ligands exerted a considerable effect on the TH activity, with high activity observed when electron-donating  [80] The complex 89 d was used as the optimal catalyst for TH of various ketones, including ketones with pyridyl moieties. [56] Imines were investigated as substrates under similar conditions with 89 b as catalyst (Scheme 39). Electron-withdrawing substituents on the sub- TH is an attractive method to access tetrahydroquinolines, valuable building blocks for pharmaceutical synthesis. [81] The Fujita group [82] and ours [83] reported work on homogeneous TH of quinolines using [IrCp*Cl 2 ] 2 and iridacycles, respectively. In 2017, Hou et al. developed mesoporous silica (SBA-15) supported iridium catalysts for selective TH of quinolines in water. [84] Two immobilized catalysts, 90 and 91, were In recent years, our group have studied a series of simple iridacycles such as 9, revealing their high catalytic efficiency in the TH of carbonyl compounds, with formic acid as the hydrogen source. [8a,53] Notably, the analogous rhodacycles are active in methanol dehydrogenation even at room temperature, allowing for TH reactions with methanol as the hydrogen source. For example, a wide range of aromatic and aliphatic aldehydes were reduced to the corresponding alcohols with methanol as both the hydrogen source and solvent near room temperature (Scheme 41). [85,86]

Achiral Hydrosilylation
Iridacycle-catalysed hydrosilylation reactions of imines, esters and amides or O-dehydrosilylation of alcohols were reported by the groups of Agbossou-Niedercorn, Michon and Djukic. [21,87] In 2017, Agbossou-Niedercorn and Michon et al. [88] described iridacycle 92 as pre-catalyst for the hydrosilylation of tertiary and secondary amides at low catalyst loadings (Scheme 42). They found that the activity of 92 can be improved through ligand modification, by changing the chelating 2-phenylpyridine ligand to a 1-phenylisoquinoline one (93). When 93 was applied to hydrosilylation of challenging amides, the related amines were obtained with higher yields in a shorter time.
The same group also reported an interesting Ir(III) metallacycle 94 as catalyst for hydrosilylation of various ketones and aldehydes in the presence of sodium tetrakis[(3,5trifluoromethyl)phenyl]borate (NaBArF 24 ) under mild conditions (Scheme 43-44). [89] This complex was also able to catalyse selectively the reduction of esters and carboxylic acids (Scheme 45), resulting in, depending on the substrate, either alcohols/ethers or alcohols/aldehydes, respectively. The authors proposed an ionic hydrosilylation pathway, which starts from metathesis of the pre-catalyst 94 with NaBArF 24 to form a cationic complex.
In 2017, Djukic et al. reported cationic iridacycles 95-97, which catalyse hydrosilylation of nitriles and carbonyls (Scheme 46 and Scheme 47). [90] For example, iridacycle 96 was shown to catalyse double hydrosilylation of a wide variety of aryl nitriles. These nitriles were converted into the correspond-  In 2018, Zhu et al. reported phosphine-containing five-, six-and eight-membered cycloiridated complexes, including their insertion reactions with aromatic alkynes. [91] The iridacycles were screened for their catalytic activities in the hydrosilylation of N-benzylideneaniline, the most active being the six-membered iridacycle 98. The complex was applied to the hydrosilylation of various aldimines and ketimines, affording the corresponding amines in 35-98 % yields (Scheme 48).

Dehydrogenation
Catalytic dehydrogenation (CDH) has attracted much attention from the viewpoint of sustainable and green chemistry, avoiding the use of stoichiometric oxidants. [92,93] In 2013, we reported a versatile iridacycle for acceptorless CDH of benzofused N-heterocycles with high activity. [94] Later, this cyclometallated Ir complex was immobilized onto multiwalled carbon nanotubes (MWCNTs) via π-π stacking interactions, aimed at recycling of the catalyst (Scheme 49). [95] The iridacycle was modified with a pyrene substituent and then immobilized onto MWCNTs. The similar conjugated π systems in pyrene and MWCNTs led to effective π-π stacking. Addition of water into the THF solution of iridacycle 99 decreases its solubility, promoting the π-π interaction because of the hydrophobicity of 99. The activity of the immobilized Ir catalyst 100 was investigated in CDH of indolines in aqueous conditions and was shown to be comparable with that of the homogeneous precursor (Scheme 50).
Since TH often involves the dehydrogenation of the hydrogen donor isopropanol, it is not surprising that cyclometallated iridium complexes have been explored for catalytic dehydrogenation of alcohols, although many methodologies are available for the oxidation of alcohols. [96][97][98][99][100] Dieguez et al. reported iridacycle-catalysed acceptorless oxidation of alcohols with liberation of hydrogen gas. [101] In their study, the iridium catalysts bearing benzoxazole (101) and thiazole (102) moieties show a higher activity than catalysts with a pendant ether group (Scheme 51). Generally, the conversion is substrate dependent, higher for secondary alcohols than for primary in 1,2-dichlorobenzene (DCB). It is noted that the common iridium precursor compound, [Cp*IrCl 2 ] 2 , is active in CDH of alcohols, as showed by Yamaguchi, Fujita and co-workers earlier. [102] With the help of noble metal catalysts, aldehydes and ketones are easily obtained by CDH of alcohols. [103][104][105][106][107][108][109] However, the direct dehydrogenation of alcohols to carboxylic acids without using an oxidant has been less reported. Since the first report by the Milstein group in 2013 using a ruthenium PNP pincer complex, [110] several RuÀ phosphine [111][112][113][114][115] and NHC complexes [116][117][118] have been described. However, there remain problems of high catalyst loading, low turnover numbers, and/or long reaction time. Very recently, the Das group reported an iridium(III)À NHC complex which catalyses efficient acceptorless dehydrogenation of alcohols to carboxylic acids. [119] Under a low metal loading of 0.1 mol%, a wide range of alcohols (aromatic and aliphatic) were converted to the corresponding carboxylic acids in high yields (Scheme 52). The acids may be formed via a pathway involving dehydrogenation to aldehydes followed by a baseinduced Cannizzaro reaction. Notably, this iridium complex could be recycled at least three times without losing its activity under the conditions employed.
In 2017, Zhao and co-workers reported the first highly enantioselective synthesis of 2-substituted 1,2,3,4-tetrahydroquinolines through the borrowing hydrogen methodology. [120] The dehydrogenation and hydrogenation were effected by a racemic iridacycle, with the asymmetric induction in the hydrogenation step brought about by a chiral phosphoric acid,  [121][122][123] In 2015, the iridacycle 9 was shown to catalyse alkylation of amines with high yields. [124] Later in 2017, Wang and coworkers reported that the same iridacycle catalyses dehydrogenative coupling of indolines with alcohols. [125] Interestingly, Nand C3-alkylation could be effected selectively by simply varying the sequence of base addition (Scheme 54). Mechanistic studies suggest a borrowing hydrogen-dehydrogenation process and a dehydrogenation-borrowing hydrogen process for the N-alkylation and C3-alkylation reactions, respectively. The iridacycle was believed to play multiple roles in the reaction, catalysing the dehydrogenation of both amines and alcohols as well as the subsequent coupling reactions.

Conclusions
The past several years has witnessed the reporting of a number of iridacycle complexes. Apart from the more conventional C,N-chelated iridacycles, metallocene-based C,N ligands have been introduced, affording novel ferrocenyl and closely related iridacycles. NHCs directed CÀ H activation has also been explored to form C,C-chelated iridacycles. A great deal of effort has been made in generating enantiomerically pure iridacycles, achieving various degrees of control of chirality at iridium. Most of these iridacycles have been examined in reactions such as hydrogenation, transfer hydrogenation, hydrosilylation, or dehydrogenation, with some displaying high versatility and high activity. However, with few exceptions, asymmetric reduction with chiral iridacycles remains challenging in terms of acceptable enantioselectivity. This stems at least partly from the difficulty in controlling the absolute configuration of the iridium centre.