C–H Activation via Group 8–10 Pincer Complexes: A Mechanistic Approach

Abstract

C–H bond activation is a crucial synthetic strategy widely utilized in both academic and industrial settings. Due to the strong and kinetically inert nature of the C–H bond, its functionalization typically requires metal-based catalysts. This review highlights the most significant advancements in homogeneously catalyzed reactions using pincer complexes with metals from groups 8–10, capable of promoting challenging C–H activation, published since 2010. In particular, it focuses on C–H bond activation for borylation, isomerization, and dehydrogenation, among other processes, discussing their scope and mechanistic insights.

1. Introduction

Functionalizing C–H bonds is one of the most crucial steps in chemical processes for synthesizing chemicals and intermediates essential for manufacturing functional materials [1,2,3]. Additionally, this strategy can reduce waste and energy consumption in synthetic processes. However, this process is challenging because C–H bonds are much less reactive than other bonds, such as C–O or C–Hal. Furthermore, due to the nature of C–H bonds, controlling selectivity is complicated, necessitating the use of directing groups [4,5].

An alternative for activating C–H bond involves the use of transition metals, which represents an attractive strategy both environmentally and economically due to savings in synthesis steps [2,6,7,8,9,10]. To date, C–H activation has been explored using various complexes, including metals such as palladium, rhodium, ruthenium, iridium, and others. The general mechanism typically includes an organometallic step where C–H bond cleavage and the formation of a C–M bond occur; these processes can proceed through mechanisms such as oxidative addition, electrophilic activation, or σ-bond metathesis [11]. Many instances of C–H bond activation exemplify stoichiometric activation at transition metal centers, which is crucial given the challenge of controlling chemoselectivity in reactions mediated by C–H activations [12,13]. Additionally, C–H activation plays a pivotal role in various catalytic processes (e.g., C–H borylation, dehydrogenation, and alkene isomerization), influencing both chemical and steric selectivity in substrates entering the catalytic cycle [14].

Among the different catalysts for C–H activation, those containing cyclometallated species have shown a remarkable role in this process, particularly highlighting the use of metallacycles employing pincer-type ligands [10,12].

Pincer systems are characterized by bearing a tridentate ligand in a meridional conformation [15,16,17,18,19,20]. These pincer ligands can be either neutral or ionic and form two metallacycles, stabilizing the metal through the chelate effect. A general representation of pincer compounds is shown in Figure 1. Pincers are typically named based on their donor atoms (D, D’: C, Si, N, P, O, S, etc.), and sometimes the connectors (Y: CH₂, O, N, S, etc.) are also specified [21,22,23]. For example, an NCN pincer contains donor atoms N, C, and N, while a PSCOP pincer contains donor atoms P, C, and P with connectors O and S. Since their first report by Moulton and Shaw in 1976 [24,25,26], pincers have been extensively studied due to their high thermal stability and remarkable catalytic activity in a wide variety of reactions, including C–H activation [18,27,28,29,30,31]. According to the literature, C–H activation is favored using late transition metals (group 8–10), owing to their high catalytic activity, stability, and controlled stereochemistry [4,32,33].

In this review, we summarize the advances in C–H activation catalyzed by group 8–10 metal pincer complexes published since 2010. The scope of this review is limited to intermolecular C–H activation catalyzed with metal complexes. Additionally, we focus on mechanistic studies of C–H activation, which are crucial for clarifying the catalytic mechanisms involved.

2. Group 8 Pincer Complexes

2.1. Iron

2.1.1. C–H Borylation

C–H borylation is a crucial synthetic tool for introducing complex functional groups into simple and abundant starting materials. Borylated compounds can further undergo C–C couplings, oxidation, and other reactions. Catalysis for C–H borylation has been achieved with high yields using metal complexes [34,35,36]. In recent years, efforts have focused on employing more abundant metals, such as iron, for this catalytic process [2,37].

In 2019, Kato et al. [38] reported the borylation of C–H bonds catalyzed with PNP–iron complexes based on 4,5,6,7-tetrahydroisoindole-2-ide. The pincer complexes achieved C–H borylation with various substrates, including arenes and heteroarenes such as pyrrole, furan, and thiophene. Mechanistic studies involved stoichiometric reactions and reactions with substrates in excess to obtain the corresponding PNP–Fe complexes (Scheme 1). Stoichiometric reaction of 1-Ph with B₂Pin₂ (bis(pinacolato)diboron) yielded the corresponding BPin (pinacolborane) derivative (1-BPin), which reacted with excess benzene to produce 1-H₂ and the borylated product (PhBPin). This complex is a dimer from the hydride iron pincer (1-H), which could be the active catalyst during C–H borylation catalysis. Moreover, 1-H₂ was able to activate the C–H bond of 2-methylfuran, generating the corresponding furanyl derivative (1-Furanyl).

Similar mechanistic studies were reported by Kamitani et al. in 2019 [39], who investigated the non-targeted borylation of various aromatic compounds using PNN pincer-type iron catalysts. In this study, stoichiometric reactions of the chloride (2-Cl) or hydride (2-H) iron pincer complexes with B₂Pin₂ produced the BPin derivative (2-BPin), which reacted with an arene to yield the borylated product. A mechanism for this reaction was proposed and is outlined in Scheme 2. The complex enters the catalytic cycle by reacting with B₂Pin₂ (step 1). Then, C–H activation could occur via a σ-metathesis or oxidative addition pathway (step 2), releasing the borylated product (step 3). Complex 10 would then react with B₂Pin₂ to continue the catalytic cycle. Kinetic isotope effect (KIE) determination led to a value of 2.0, indicating that the C–H activation process would be the rate-determining step in this catalytic process.

2.1.2. Hydrogen Isotope Exchange (HIE)

Tritium- and deuterium-labeling systems are very important in medicinal chemistry, with several applications such as the identification of receptor dynamics and the localization of potential drugs in vivo [40,41,42,43]. The use of direct hydrogen isotope exchange (HIE) with metal complexes is a strategic method to study mechanisms of action, avoiding conventional multistep synthesis [44,45]. This information allows the tuning of metal complexes to induce selectivity [46].

In this context, Pony Yu et al., in 2016 [47], reported an HIE catalysis with a CNC–Fe(0) pincer (3) for the deuteration and tritiation of pharmaceutical products such as Varenicline (Chantix and Pfizer), Papaverine, Paroxetine (GSK), Loratadine (Claritin and Schering/Merck), MK-6096 (Merck), Cinacalcet (Sensipar and Amgen), Flumazenil (Roche), and Suvorexant (Belsomra and Merck) (Figure 2), employing D2 and T2 as isotope sources. C–H activation selectivity was identified for electron-poor substrates with sterically accessible C–H bonds.

Mechanistic studies on the catalytic activity of complex 3 were carried out by the Chirik group. A stoichiometric reaction with H₂ (1 atm) showed a stable dihydride with a σ-H₂ ligand (4-H₂), which partially incorporated deuterium from a deuterated solvent (benzene-d₆) to yield the isotopomer with a σ-H-D ligand (4-HD) [48]. In 2020, Corpas et al. [49] carried out mechanistic studies with 3 in C–H activation and identified that H₂ acted as an activator of the catalyst. Given this activation, benzene-d₆ could be used as a deuterium source since H₂ could facilitate C₆H_nD_6-n (n = 1–6) elimination, yielding the active catalyst 5 (Scheme 3). Although dinitrogen coordination in 5 would be plausible, it is readily displaced under catalytic conditions.

Based on these results, the authors proposed the mechanism shown in Scheme 4. The CNC–Fe(0) is activated with H₂ and C₆D₆ producing the corresponding dideuteride Fe(II) pincer (step 1). The iron dideuteride compound causes C–H activation via σ-C–H bond coordination (step 2), which incorporates a deuterium atom into the aryl ligand via either a sigma metathesis or a Fe(II/IV) cycle (step 3). Then, the deuterated product is eliminated, and the catalyst is again activated with H₂ and C₆D₆ (step 4).

This mechanism is similar to that proposed by the de Rutier group using a PCP–Fe(II) pincer (6) as catalyst and benzene as substrate [50]. In addition, PCP–Fe(0) pincers exhibited the expected selectivity for the most sterically accessible position in reactions with different arene substrates. Computational studies supported a σ-complex-assisted metathesis (σ-CAM), as the critical step in C–H activation (Scheme 5). This leads to the nonclassical hydride isotopomer, which rotates to position the deuterium close to the aryl carbon. Subsequently, the deuterium is incorporated into the aryl ring, forming the σ-C-D adduct of the deuterated product, which is then released.

2.2. Ruthenium

2.2.1. Hydrogen Isotope Exchange (HIE)

Deuterium oxide is an inexpensive source of deuterium for HIE catalysis, but its use is limited by solubility problems that restrict the number of potential catalysts. Therefore, new strategies are required to overcome this challenge. The Periana group implemented strongly basic conditions as a strategy to analyze an NNN–Ru(III) pincer (7) [51]. This complex was reduced in situ with Zn to generate the Ru(II) active catalyst. In the catalytic cycle, a nucleophilic C–H activation pathway is proposed, facilitated by the OD ligand (Scheme 6). Subsequently, HOD is substituted with deuterium oxide (D₂O), followed by the deuterium transfer to the arene.

2.2.2. Alkane Dehydrogenation

Alkenes are very important organic compounds for both industrial and academic purposes. They can be selectively obtained from more abundant sources such as alkanes [52,53]. The main disadvantages of industrial dehydrogenation are the high temperatures required (approx. 400–600 °C) and the use of heterogeneous catalysts, which often do not provide optimal regioselectivity. Therefore, alternative options, such as homogeneous catalysts with metal complexes, have emerged in recent years, allowing these processes to be carried out with higher regioselectivity and at lower reaction temperatures [54,55,56]. For instance, Zhou et al. [57] performed the catalytic alkane dehydrogenation of cyclooctane (COA) using a PSP–Ru pincer (8, 1 mM), with t-butylethylene (TBE) (300 mM) as a hydrogen acceptor (Scheme 7). Turnover frequency (TOF) values of 0.1–0.8 s⁻¹ were obtained at 120–180 °C, with the reaction nearly complete at 180 °C after 5 min. DFT calculations highlighted the C–H activation via oxidative addition, followed by β-H elimination, as determining steps in the catalytic cycle. Furthermore, acceptorless dehydrogenation assays were conducted with complex 8 (1 mM) in a COA solution at 151 °C, showing a low yield of cyclooctene (COE, 14 mM) after 48 h.

2.2.3. Direct Alkenylation

C–C cross coupling catalysis is a powerful synthetic tool in organic chemistry for designing complex molecules or inducing specific stereochemistry, often requiring pre-functionalized substrates [58]. However, direct C–C cross-coupling can be achieved from transition-metal-catalyzed C–H activation products [59,60]. For instance, Cai et al. [61] reported the direct alkenylation of several heteroarenes and 2-olefinpyridines using a CCC–Ru(IV) pincer (9). A kinetic isotope effect (KIE) value of 2.3 implied that the C–H activation step was rate-determining, with 2-(o-tolyl)pyridine as the substrate. Additionally, only trace amounts of the alkenylation product were obtained when the radical scavenger TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl) was added. Based on these results, a mechanism is proposed and is shown in Scheme 8. Complex 9 is reduced in situ to produce the active Ru(III) catalyst (step 1), which undergoes a single electron transfer (SET) (step 2). This intermediate was detected via real-time high-resolution mass spectrometry (HRMS). Then, coordination of the directing group pyridine favors the C–H activation (step 3), followed by the release of the alkenylated product (step 4) [62].

2.2.4. Dehydrogenative Silylation

Arylsilanes are compounds with prominent applications in organic synthesis, cross-coupling, optoelectronics, and more. These compounds can be tuned via silylation catalysis to yield chiral compounds like spirocyclic scaffolds. For example, the Huang group carried out dehydrogenative silylation catalysis [63] using a PCP–Ru(II) pincer (10). Silicon-centered spirocycles with different backbones were obtained in moderate to high yields (Scheme 9). Although mechanistic studies have not yet been reported, silicon likely acts as a directing group during the C(sp³)–H activation step, according to computational studies with similar Ru complexes [64].

2.3. Osmium

Alkane Dehydrogenation

The Roddick group [65] evaluated the catalytic activity of the PCP–Os(II) pincer (11) in alkane dehydrogenation. An initial turnover frequency (TOF) value of 1520 h⁻¹ was obtained, utilizing COA and TBE (1:1, 0.033 mol%, 200 °C, 10 min). Furthermore, activity inhibition of complex 11 was detected when the concentration of COE or H₂ increases. H₂ inhibition produced a dimeric compound (12) (Scheme 10). Acceptorless dehydrogenation assays of COA were made and produced trace amount of the alkynated product.

3. Group 9 Pincer Complexes

3.1. Cobalt

3.1.1. C–H Borylation

PNP–Co pincer compounds have been evaluated as catalysts in C(sp²)–H borylation catalysis by the Chirik group, contributing to several mechanistic studies. For instance, Obligacion and colleagues assessed the catalytic activity of complex 13 using substituted arenes and pyridines (Scheme 11) [66]. For pyridines, catalysis was carried out under two different conditions (conditions A and B) with 0.5 equiv and 1 equiv of B₂Pin₂. A higher yield was observed under condition B compared to condition A. Additionally, the byproduct obtained under condition A was H₂, whereas under condition B it was HBPin. This study demonstrated that several meta-borylation products were favored by the Co(I) pincer catalyst.

Kinetics studies, including Kinetic Isotope Effect (KIE) measurements and rate law determinations, were conducted for complexes 13, 14-pyrr, and 14-Me using different substrates and borane sources (Scheme 12,

Table 1. Kinetic Isotope Effect (KIE) and rate law determination for C(sp²)–H borylation with Co pincer compounds.

Cat.	Substrate	Experimental Conditions	KIE	Rate Law	Ref.
13		3 mol%, B2Pin2, 80 °C	2.9(1)	rate=kobs[13][substrate]1[B2Pin2]0	[67]
14-pyrr		3 mol%, B2Pin2, 80 °C	1.6(1)	rate=kobs[14-pyrr][substrate]1[B2Pin2]0	[67]
13		1 mol%, HBPin, 80 °C	1.9(1)	rate=kobs[13][substrate]0[HBPin]0	[68]
14-Me		10 mol%, B2Pin2, 50 °C	1.1(1), 0.9(1) *	rate=kobs[14-Me][substrate]0[B2Pin2]0 [HBPin]0.5	[69]

* ortho and meta borylation, respectively.

) [67,68,69]. The rate laws showed no apparent dependency on the borane source. However, Pabst and colleagues identified that the byproduct HBPin exhibited a half-order behavior for complex 14-Me. In addition, complexes 14-Me and 14-pyrr showed a KIE close to one, indicating a different turnover step from C–H activation. [69]. Complex 13, on the other hand, exhibited higher KIEs with 2,6-lutidine and benzofuran, suggesting that C–H activation is likely the turnover-determining step in the catalysis. For benzofuran, a zero-order behavior of the benzofuran concentration eliminates this proposal [67,68].

To determine the catalyst resting state, C–H borylation was examined via ¹H and ³¹P NMR spectroscopies in THF-d₈ at 23 °C. The catalyses were performed with 2,6-lutidine, B₂Pin₂, or HBPin, employing a high loading of 13 (10 mol%) at 80 °C. Higher catalyst loading facilitated compound identification in solution. C–H borylation yields were 85% after 10 h and 19% after 32 h with B₂Pin₂ and HBPin, respectively. Complexes 15 and 14-BPin were identified as resting states when B₂Pin₂ was employed as the borane source. Stoichiometric reactions with B₂Pin₂ and HBPin confirmed these compounds, with no inhibitory effect detected for dinitrogen (N₂) [67]. In contrast, 14-H and 14-BPin were obtained when HBPin was utilized (Scheme 13). Interestingly, a catalysis with 2-methylfuran (instead of 2,6-lutidine) only presented 14-H as the resting state. This catalysis showed a C–H borylation yield of 62% after 24 h at 23 °C in C₆D₆ [68]. Thus, the catalyst resting state appears to depend on the concentration of the BPin unit during the catalysis.

Based on a Co(I)–Co(III) cycle, a general mechanism is proposed for C(sp²)–H borylation with PNP–Co pincers, considering the experimental data as presented in Scheme 14. The cobalt pincer precatalyst initiates the catalytic cycle by reacting with either HBPin or B₂Pin₂, yielding the corresponding Co(I) hydride pincer or its BPin derivative, respectively (steps 1 or 1′). In cases where the Co(I) complex lacked para-substituents, BPin would be added at its para position [68]. The Co(I) hydride complex undergoes oxidative addition with XBPin (X = H, BPin) to form a Co(III) complex (step 2), which then eliminates HX via reductive elimination, yielding the BPin derivative (step 3). Subsequently, C(sp²)-H oxidative addition of the corresponding arene to the BPin derivative is kinetically favored, leading to an isomer where the boryl and aryl ligands are in the trans position (step 4) [70]. Following this, isomerization of the Co(III) pincer to the cis stereoisomer relative to the boryl and aryl ligands occurs through reductive elimination and oxidative addition of HBPin (step 5), ultimately resulting in the formation of the aryl boronate product via reductive elimination and regeneration of the Co(I) hydride pincer complex (step 6). A higher concentration of HBPin during catalysis directs the Co(I) hydride pincer to the resting state (step 7) or, alternatively, via rearrangement when X = H (step 8).

Selectivity for a specific borylation position in arenes is determined in the C(sp²)–H oxidative addition (Scheme 14, step 4) and depends on several factors such as steric accessibility, M–C/C–H bond thermodynamics, and chelation assistance [71]. For example, the Chirik group identified a high ortho-to-fluorine selectivity in C–H borylation with different functional groups on the arenes. In addition, they reported a four-step synthesis of the anti-inflammatory drug flurbiprofen employing the ortho-to-fluorine selective C–H borylation, which reduced the previously reported eight-step route [72]. Likewise, Pabst and coworkers compared the cobalt–carbon bond dissociation free energies (BDFEs) of nine fluoroaryl complexes (Co-ArF_n, n = 0–2) versus their C–H BDFEs [69]. Their computational study showed a higher M–C stabilization as the ortho fluorine substitution increased. For 1-fluoro-3-(trifluoromethyl)benzene, a ortho/meta selectivity of 99:1 was predicted since the ortho isomer (10b) ground energy was calculated to be 2.6 kcal/mol more stable than that of the meta isomer (10a). This calculation was in good agreement with the experimental selectivity of 95:5 (ortho/meta) (Scheme 15).

The connectors between the phosphine and the pyridine ring in PNP–Co pincers were observed to be key points for their catalytic activity. For example, the Hall group demonstrated through computational studies that conformational rearrangements of these methylene connectors in PNP–Co were relevant throughout the catalytic cycle [70]. Moreover, methylation of the methylene connectors led to a decrease in the catalytic activity and selectivity. This modification increases the steric profile of the connector, thereby reducing the flexibility of the pincer [73,74]. Substituting the methylene connector with an oxygen atom has not yet yielded a stable Co pincer complex [75].

Terpyridine (NNN)–cobalt complexes were also evaluated as ligands in C(sp²)-H borylation; however, they exhibited lower yields compared to PNP pincers. To enhance catalytic activity, Léonard and colleagues added LiOMe as a base with an NNN–Co(II) pincer to control the excess HBPin concentration, forming a methoxide–borohydride adduct and avoiding an off-cycle resting state [76]. Interestingly, Pabst and coworkers employed an NNN–Co(I) pincer to favor meta-to-fluorine selective C–H borylation. For complex 18, a competition deuterium kinetic isotope experiment (KIE) was carried out with 3-fluoro-N,N-dimethylbenzenesulfonamide as the substrate (Scheme 16). The result of a KIE of 2.9(2) suggests that the C–H activation step is turnover limiting. Isomerization following C–H oxidative addition to the complex was disregarded, as borylation of fluoroaryl Co(I) pincers exclusively yielded the corresponding borylated product. Given that the ortho position is stabilized by fluorine substituents, meta selectivity would result from the kinetic product of irreversible C–H activation [77].

Bis(carbene)- and bis(silylene)pyridines Co pincer compounds have been applied as electron-rich catalysts in C–H borylation. For example, Co(I) carbenes synthesized by Roque and coworkers exhibited high catalytic activity in C–H borylation, displaying meta-to-fluorine selectivity at room temperature. However, reaction of 19 with an excess of 1,3-difluorobenzene demonstrated isomerization of the fluoroaryl complex to the ortho-isomer after 16 h (Scheme 17a), yielding the corresponding borylated product. They designed two different experimental conditions to favor either the meta-isomer (2 equiv B₂Pin₂, rt) or the ortho-isomer (1 equiv HBPin, 50 °C) (Scheme 17b) [78].

On the other hand, N-Heterocyclic silylene–pyridine Co pincer compounds were first used by the Cui group in C–H borylation. This Co(II) pincer catalyzed selectively meta-to-fluorine C–H borylation with a catalyst loading of 4 mol% and 8 mol% of NaBHEt₃ as a reductant. In addition, 0.5 mmol of cyclohexene was added to control excess concentrations of HBPin [79]. Subsequently, Arevalo and coworkers evaluated the catalytic activity of a series of SiNSi–Co(III) pincers (20, 20-Ph). High catalyst loading was required to identify any resting state during the catalysis. For 2-methylfurane (Scheme 18a), two resting states were observed, involving the BPin–Co(III) derivative (21-Ph) and the complex bearing a σ-H₂ ligand (22). For 2,6-lutidine, the resting states resembled 1-Ph and depended on the aryl substituent of the silylene fragment (Scheme 18b). With a non-substituted aryl fragment, borylation at the meta position of the catalyst was observed (21-BPin) [80]. These resting states are similar to those of the PNP pincer and, therefore, could follow the same mechanism.

3.1.2. Hydrogen Isotope Exchange (HIE)

Roque and coworkers reported a catalytic hydrogen isotope exchange (HIE) via C(sp²)–H activation employing a bis(silylene)pyridine–cobalt(III) pincer complex (21) [81]. Deuterium incorporation was achieved with benzene-d₆ and a low loading of cobalt precursor 21 (0.5–5 mol %). Additionally, sterically hindered positions in arenes with methyl substituents (23a-f) were highly deuterated (>97%) (Scheme 19).

A slower rate of deuterium incorporation was observed when HBPin was added to the reaction, which suggested a resting state off-cycle. Therefore, loss of HBPin from the complex 21 was proposed to generate the cobalt(I) hydride complex, which would react via oxidative addition with benzene-d₆ in the HIE catalysis [82]. In addition, different Co(III) pincer precursors were evaluated in HIE to identify the active catalyst; 3-fluoro-N,N-dimethylbenzenesulfonamide was employed as substrate (Scheme 20). Complete deuteration (>98%) was observed with complex 21 after 60 min and with complex 20 after 12 min. When benzofuranyl derivatives were utilized, a higher deuterium incorporation was detected with the monobenzofuranyl complex (24a, <5%) than that of the bisbenzofuranyl complex (24b, 78%). A lower yield with complex 24b could be because this complex would not generate the cobalt(I) hydride.

3.2. Rhodium

3.2.1. C–H Borylation

The Esteruelas group evaluated C–H borylation catalysis using a POP–Rh(I) complex (25) [83]. Stoichiometric reactions with arenes and HBPin produced the corresponding aryl and boryl derivatives, respectively. Borylation of arenes exhibited a preference for the meta position (Scheme 21). An exception was observed with fluorobenzene, where ortho position borylation was favored. This ortho-to-fluorine selectivity in Rh complexes is attributed to thermodynamic and electronic factors [84,85,86]. Catalysis was efficient with either B₂Pin₂ or HBPin, suggesting a catalytic cycle similar to that of Co pincers (vide supra).

3.2.2. Direct Arylation

Lewis et al. reported a catalytic direct arylation with a PNP–Rh(I) pincer (26) [87]. C–H activation in benzene readily yielded the aryl derivative at room temperature using ^tBuOK as the base after 16 h, which reacted with iodoarene to form the corresponding C–C coupling product. Arylation of several arenes showed functionalization at both C(sp²) and C(sp³) positions (Scheme 22). Selectivity was identified for C(sp²) and at ortho positions. This selectivity was similar to that observed in the reaction without a catalyst, which could indicate a radical pathway. Therefore, the Rh(I) would accelerate the reaction via radical intermediates.

3.3. Iridium

3.3.1. C–H Borylation

8-Hydroxyquinoline-2-carboxylic acid-derived complexes 27, 27-py, 28, and 28-py were found to be active catalysts in the borylation of arenes [88]. The catalytic activity of these complexes was evaluated using HBPin as the boron source under an argon atmosphere in neat benzene, with a catalyst loading of 5.0 mol% (Scheme 23). Conversion percentages ranged from 20% to 72%, with the highest percentages observed for the pyridine derivatives (27-py and 28-py). Complex 28-py, for instance, achieved a 71% yield of the aryl-borane product from the C–H borylation of anisole, with selectivity favoring the meta-isomer. Although mechanistic studies have not yet been conducted, the preference for the meta-position is likely associated with steric factors.

The Ozerov group [89,90] has developed catalytic C–H borylation of arenes with POCOP–Ir(III) pincers (29a-c), with the addition of hydrogen acceptors (HAs). Hydrogen acceptors were required to reduce the metal complex, increasing the borylation percentage on the arene from 4% to ca. 90%. For complex 29c, catalyst loading was reduced to 0.004 mol%, yielding a conversion percentage of 83% and a turnover frequency (TOF) of 20,750 h⁻¹ with benzene-d₆ as the substrate and ethylene as the H₂ acceptor. Conversely, side reactions of HBPin with an HA mainly produced the corresponding alkane. On the other hand, side reactions of B₂Pin₂ and HAs led mainly to the corresponding alkane. Stoichiometric reaction of 29b with fluorobenzene resulted in the fluoroaryl derivative (31), which next reacted with 1 equiv HBPin to yield the borylated product and the dihydride iridium complex (32). A proposed mechanism (Scheme 24) begins with the formation of the 14-electron pincer intermediate via a dehydrochlorination reaction of complex 29 using a base (step 1). Subsequently, oxidative addition of the arene occurs (step 2), followed by formation of the borylated product (step 3). And the 14-electrons pincer intermediate is regenerated via alkene reduction (step 4). Reaction of 32 with H₂ formed complex 33, which could act as an off-cycle step. Hence, HAs are necessary to prevent this by consuming H₂ (step 5). Additionally, auxiliary off-cycle hydride/boryl redistribution equilibria from 32 were observed when reacting with a stoichiometric amount of HBPin.

Ozerov and coworkers carried out a dehydrogenative borylation of terminal alkynes (DHBTA) catalyzed with a SiNN–Ir complex (34) [91]. This reaction typically yields borylated alkanes as major products, with borylated olefins as byproducts. However, they discovered a method to selectively obtain borylated olefins by degassing and introducing 1 atm of CO into the reaction mixture after 10 min (using 4-ethynyltoluene and five equivalents of HBPin at 55 °C for 18 h). Under these conditions, a triborylated olefin was obtained with a higher yield (74%), allowing for fine-tuning of olefins through subsequent cross-coupling reactions. Additionally, substrate scope demonstrated triborylation yields up to 80% with substituted aryl and alkyl alkynes (Scheme 25). An exception was observed for Me₃SiC≡CH, which stopped at the alkynylboronate stage, likely due to steric hindrance. Stoichiometric reaction of CO with 34 generated the mono- and dicarbonyl–iridium pincer, but individual catalysis with the monocarbonyl complex resulted in a lower conversion percentage (48%).

3.3.2. Alkane Dehydrogenation

Alkane dehydrogenation catalyzed with iridium complexes has been extensively studied since the first report of alkane dehydrogenation with an iridium complex by Crabtree in 1979 [92]. Since then, efforts have focused on designing new ligands to either tune or enhance their catalytic activity [93,94,95]. For instance, the Huang group synthesized various PCN–Ir pincer complexes (35–46) [96] for use as homogeneous catalysts in alkane dehydrogenation (AD) (Figure 3). These complexes feature a bulky substituent (e.g., ^tBu) on the P arm to prevent dimer formation, and an N-type arm that could provide favorable steric effects, aiming for asymmetric reactions [97]. Moreover, the complexes are stabilized by electronic effects arising from the push–pull interaction between N and P donor atoms.

Complexes 35–48 were evaluated as precatalysts in the transfer dehydrogenation (TD) reactions of simple alkanes (Scheme 26,

Table 2. Transfer dehydrogenation of COA/TBE catalyzed with PCN–Ir.

Entry	Cat.	Cat. Loading (mol%)	T (°C)	TOF (h−1) 1
1	35	0.2	150	26
2	36	0.2	150	3
3	37	0.2	150	2
4	38	0.2	150	333
5	39	0.2	150	6
6	40	0.2	150	6
7	41	0.025	200	191
8	47 2	0.033	200	448
9	48 2	0.017	200	328

¹ t = 1 h (35–40), 18 h (41), 4 h (47), 15 h (48). ² Catalysts used as control.

). The highly effective precursors 47 and 48 (PCP-IrHCl) served as catalytic controls using cyclooctane (COA) and 2-tert-butyl-1-ethylene (TBE) as substrates (entries 8 and 9) [98,99]. The catalytically active 14-electron species PCP–Ir, along with NaCl and tert-butanol as byproducts, were generated through the dehydrochlorination of PCP–IrHCl complexes using NaO^tBu as the base [100,101]. Iridium complexes 38 and 41 showed the highest catalytic activities with TOF values of 333 h⁻¹ and 191 h⁻¹, respectively. These complexes feature a phosphinite donor arm, characterized by its rigid structure similar to that of 47 and 48. Under the given experimental conditions, the productivity of catalyst 41 was comparable to that of the control catalysts in the catalytic assay.

On the other hand, complex 41 was also used in the transfer dehydrogenation (TD) of simple heterocycles, exhibiting outstanding activity. Heteroarenes like 2,3-dihydrobenzofuran and indole were produced with a high conversion percentage (>95%) and a low catalyst loading of 0.1–0.2 mol% (Scheme 27). On the contrary, pyridines production required a higher temperature (200 °C) and higher catalyst loading (5 mol%) to yield a conversion percentage up to 88%. Those percentages were higher than those of the catalyst of control 48 (up to 63%).

The Huang group proposed that the lower conversion percentage observed with pyridine-containing ligands could be attributed to inhibition by pyridine coordination to the metal. To investigate this, they utilized the agostic complex (46) to identify these coordination species, such as the hexacoordinated Ir complex (49). This complex showed an ortho-C(sp²)-H cyclometallation of the pyridine ring of the PCN ligand, caused by N dissociation and a subsequent C(sp²)–C(sp²) rotation, producing intermediate 50 (Scheme 28). Thus, the formation of these Ir(III) species outside the catalytic cycle could deactivate the catalyst. Therefore, the enhanced activity of complex 46 is attributed to its rigid structure, which precludes C–H activation promoted by the metal center and thereby eliminates the potential for cyclometallation.

Goldman and coworkers [102] conducted comparative studies of regioselectivity in AD using iridium pincer complexes. They combined experimental and computational approaches to elucidate the fundamental factors governing regioselectivity (Scheme 29). Experimental competition studies of AD were performed using n-alkanes and cycloalkanes, such as n-pentane and cyclododecane (CDA), with complexes 30 and 51–56. All complexes, except for 53, showed high selectivity for the dehydrogenation of n-pentane over CDA with selectivity ratios reaching up to 25:1. The main product formed was predominantly 1-pentene.

Computational comparative studies were conducted through DFT calculations based on the catalytic cycle proposed for catalysts 51 and 53 (Scheme 30). For complex 51, the calculations obtained for the electronic structure with n-hexane indicated that the rate-determining step and selectivity are governed by the β-H transfer (step 2). In addition, C–H activation (step 1) at the terminal carbon (C1) is favored over addition at the internal carbon (C2), both kinetically (ΔG‡) and thermodynamically (ΔG°) by ca. 3 kcal/mol. For complex 53, replacing the connector CH₂ with O does not change the preference for C–H activation at C1. However, the rate-determining step for this complex is the olefin dissociation (step 3), instead of β-H transfer (step 2). In step 3, the loss of the olefin is favored by 2 kcal/mol to the internal olefin than the terminal olefin, which is consistent with the inversion of regioselectivity for complex 53. The 14-electron iridium species is regenerated via alkene reduction (step 4).

Goldman group [103], have also investigated the mechanism and regioselectivity of catalytic dehydrogenation of alkyl chains attached to molecules other than alkanes. For example, the formation of enamines from tertiary amines, using PCP–Ir pincer complexes. A competitive reaction was realized with COA, N,N-diisopropylethylamine (DIPEA), and PCP–Ir pincer 57 (51 precursor), showing a selectivity of 2:1 DIPEA:COA (Scheme 31a).

Kinetic isotope effect (KIE) determination was performed employing DIPEA (A: ⁱPr₂NCH₂CH₃) and its isotopomers (B: ⁱPr₂NCD₂CD₃, C: ⁱPr₂NCD₂CH₃, D: ⁱPr₂NCH₂CD₃). A k_A/k_B value of 7.0 revealed a strong rate-limiting step originated in the ethyl chain. KIE of the partially deuterated isotopomers showed values of 2.0 and 3.7 for k_A/k_C and k_A/k_D, respectively. The value of 2.0 for k_A/k_C suggests an equilibrium in the C–H activation in which there is a transfer of H from the carbon atom to the metal center, while the value of 3.7 for k_A/k_D would indicate the β-H elimination as the rate-determining step (r.d.s.) (Scheme 31b).

Due to the high catalytic activity presented by the Ir pincer compounds in alkane dehydrogenation, the Goldman group successfully synthesized benzene from hexane via catalysis [104]. Mechanistic studies through DFT calculations of this reaction were reported by the Sunoj group, utilizing a PCP–Ir catalyst [105]. These studies revealed that the most plausible route for benzene formation could occur through the formation of hex-1-ene by activating the C1-H bond via a σ-agostic interaction between iridium and the terminal methylene group, followed by the formation of hexa-1,3-diene and hexa-1,3,5-triene. Once these species are formed, the final step involves electrocyclization of the triene to cyclohexadiene, followed by benzene releasing a dehydroaromatization (Scheme 32).

3.3.3. Direct Alkylation

Inactive C–H bonds are predominant in organic compounds, where the activation of such bonds becomes desirable to carry out the formation of different high-value compounds. Unlike C(sp²)–H bonds, C(sp³)–H bonds are difficult to activate due to their non-acidic nature, resulting in a smaller dissociation energy. To address this issue, the use of transition metal complexes has been an efficient way to activate such bonds [106,107,108] that could be part of the metal complex as a ligand or be further functionalized [109].

The C(sp³)–H activation has also been achieved in enantioselective catalysis through functionalization via carbenoids according to the Song group [110,111] by using chiral NCN–Ir(III) catalysts (Scheme 33a). From a series of complexes, they determined that the compounds carrying out better enantioselective C–H insertion using α-aryl-α-diazoacetates with protected indoles (Scheme 33b) and various 3-diazo-oxindoles substituted with 1,4-cyclohexadiene (Scheme 33c). Catalyst 58 carried out an insertion at the C3 position of the indole through C–H activation, resulting in the formation of optically active indole derivatives with yields up to 98% and enantioselectivities (ee) ranging from moderate to good (up to 86% ee). On the other hand, catalyst 59 was also employed to carry out insertion reactions, but this time for the formation of oxindoles (Scheme 33b). The insertion of 1,4-cyclohexadiene occurs at the 3-position of the oxindole, resulting in the formation of optically active compounds with yields of 42–95% and enantiomeric excesses ranging from moderate to good (51–99% ee).

DFT calculations were employed to elucidate the mechanism of the catalyst, and based on these results, a proposed catalytic cycle is presented in Scheme 34. The reaction initiates with the dissociation of water from the catalyst, forming a 16-electron species. The energy required for water dissociation to form the trans intermediate is minimal, at 1.4 kcal/mol. This trans intermediate then interacts with 3-diazo-oxindole, resulting in the formation of an iridium carbene complex (−39.1 kcal/mol) and the expulsion of nitrogen. Further calculations indicated that the iridium carbene complex with the trans intermediate exhibits an energy gap of 25.9 kcal/mol, whereas formation via the cis intermediate encounters an energy barrier of 17.5 kcal/mol, suggesting a rearrangement favoring the trans intermediate. The intermediate carbene complex undergoes C–H insertion with 1,4-cyclohexadiene, yielding the desired organic product. Based on computational results, the authors proposed that the enantioselectivity would be originated by a hydride transfer and a C–C formation step via the Si-face approach of the cyclohexadiene to the intermediate complex [110].

Blakey and colleagues [112] synthesized a series of Ir(III) pincer complexes from phebox-type ligands, with the aim of conducting enantioselective functionalization reactions via C(sp³)–H activation (Scheme 35). Similar to the studies conducted by Song, reactions were carried out using different diazoesters and 1,4-cyclohexadiene with the Ir(III) catalyst (33). These reactions exhibited yields ranging from 71% to 99% and enantiomeric excess (%ee) values ranging from 83% to 99%. Computational results would indicate that the enantioselectivity is originated by an axial conformation of the intermediate that exposes only the Re-face to the cyclohexadiene.

3.3.4. Alkene Isomerization

Over the past couple of years, there has been a quest for the development of highly active and selective catalysts to conduct isomerization reactions, as they are of vital importance in a wide range of areas. For instance, they play a significant role in Shell higher olefin processes, the formation of aliphatic amines through internal olefins via hydroaminomethylation [113,114,115], or the synthesis of terpenoid [116,117] compounds. Due to this, the group of Chianese and colleagues [118] proposed a potential reaction mechanism in the isomerization of alkenes catalyzed with a CCC–Ir(III) pincer complex (61). Through NMR experiments and H-D cross experiments, as well as isolation and characterization via X-ray diffraction of one of these intermediates (36), they observed different reaction intermediates. The isomerization mechanism (Scheme 36) suggests that the reaction occurs in four steps: (1) Complex 61 undergoes dehydrohalogenation using NaO^tBu as the base, generating a 14-electron iridium intermediate which reacts with 1-alkene to form an η²-1-alkene complex. (2) At this point, oxidative addition occurs via an allylic C–H bond, followed by rotation of the η³-allyl ligand to generate the resting state of the catalyst (62). (3) Subsequently, a reductive elimination occurs to generate an η²-2-alkene complex, and finally, (4) the 2-alkene is replaced by 1-alkene, initiating the catalytic cycle anew.

4. Group 10 Pincer Complexes

4.1. Nickel

Direct Alkylation

The use of nickel is attractive due to its comparable catalytic efficiency to other transition metals, coupled with advantages such as economic feasibility, abundance, and lower toxicity. Nickel pincer complexes have proven to be efficient catalysts in C–H alkylation [119]. Arora et al. emphasize the significance of employing pincer-type compounds with nickel in alkylation and arylation reactions (Figure 4), involving the activation of C–H bonds. They analyze various reports of NNN pincer compounds in catalyzing reactions with substrates including indoles, arenes, pyrazolyl, and carbazolyl, among others, featuring diverse functional groups such as halides, ethers, and amines (Complexes 63–66) [120,121,122,123,124,125]. While the authors do not delve into specific reaction mechanisms, they note that direct activation of C–H bonds predominates in most cases, leading to moderate to high yields across the diverse scenarios presented [119].

In 2014, Gartia et al. published their findings on the catalytic capabilities of a nickel(II) pincer NNN complex featuring the ligand N,N-bis(2,6-diisopropylphenyl)-2,6-pyridinedicarboxamido (L). They explored Grignard reagent couplings by activating C–H bonds in heterocyclic compounds such as tetrahydrofuran and furan. The nickel(II) complex demonstrated excellent activity in catalyzing C–H activation and promoting coupling with various Grignard reagents. Notably, C–H activation occurred under ambient reaction conditions with short reaction times (1–2 h) and a high turnover frequency of 4130 h⁻¹, using a catalyst loading as low as 0.01 mol% [124]. Gartia et al. noted that the C–H activation process occurs at the 2-position relative to oxygen, facilitated by oxidative addition where the heterocyclic compound binds to Ni, positioning it for subsequent coupling with an alkyl or aryl group from the Grignard reagent, resulting in substitution at the 2-position of the heterocycle (Scheme 37).

4.2. Palladium

4.2.1. C–H Borylation

Mao et al. describe oxidative borylation catalyzed with palladium species, proposing a concerted metalation–deprotonation-type mechanism. They propose that initial oxidation of the metal species activates the C–H bond of alkenes, facilitating the formation of new C–B bonds [126]. The proposed catalytic cycle for C–H borylation with NCN–Pd is depicted in Scheme 38, outlining the mechanistic steps involved in the activation of allylic C–H bonds of alkenes for subsequent borylation through a single, concerted metalation–deprotonation process.

4.2.2. Direct Alkylation

Arumugan et al. reported the synthesis of two NNO clamp-type palladium compounds, tested as catalysts in the synthesis of carbazole derivatives. The authors proposed a reaction mechanism highlighting the role of these complexes in the different steps of the synthesis. For our focus, we address the latter part of the synthesis, where the complexes activate the N–H and C–H bonds, leading to the formation of a new five-membered ring and a C–N bond. According to the authors, the catalytic cycle continues with the deprotonation of the N–H group, forming a new palladium complex that immediately activates the adjacent phenyl C–H bond, resulting in a six-membered palladacycle. Finally, reductive elimination releases the N-acetylcarbazole and regenerates the active Pd(0) species for the next catalytic cycle (Scheme 39) [127].

5. Conclusions

C–H activation is a crucial tool in synthesis, and numerous studies have focused on optimizing and fine-tuning this process to specific requirements. As highlighted in this review, pincer complexes have been successfully employed to activate and functionalize C–H bonds, effectively targeting substrates such as alkanes, which are abundant organic compounds.

As demonstrated, C–H bond activation is a crucial step in the catalytic cycle of various processes such as C–H borylation, hydrogen isotope exchange (HIE), and alkane dehydrogenation. It determines the reaction rate and influences regio- and stereoselectivity during catalysis. Moreover, C–H activation facilitates direct alkylation or arylation, enabling direct C–C coupling without the need for a functionalized intermediate.

Various mechanistic insights into the C–H activation pathway have been proposed through the identification of resting states, kinetic isotopic effects (KIE), and computational calculations. These insights are influenced by the choice of metal pincer and substrate. However, additional investigation is necessary to validate these mechanisms and, crucially, to predict and optimize catalysis for specific applications. The perspective outlined above clearly reveals a broad range of opportunities for future developments of more efficient and selective catalysts based on pincer complexes.