Ruthenium-Catalyzed Reductive Cleavage of Unstrained Aryl–Aryl Bonds: Reaction Development and Mechanistic Study

Abstract

Cleavage of carbon–carbon bonds has been found in some important industrial processes, for example, petroleum cracking, and has inspired development of numerous synthetic methods. However, nonpolar unstrained C(aryl)–C(aryl) bonds remain one of the toughest bonds to be activated. As a detailed study of a fundamental reaction mode, here a full story is described about our development of a Ru-catalyzed reductive cleavage of unstrained C(aryl)–C(aryl) bonds. A wide range of biaryl compounds that contain directing groups (DGs) at 2,2′ positions can serve as effective substrates. Various heterocycles, such as pyridine, quinoline, pyrimidine, and pyrazole, can be employed as DGs. Besides hydrogen gas, other reagents, such as Hantzsch ester, silanes, and alcohols, can be employed as terminal reductants. The reaction is pH neutral and free of oxidants; thus a number of functional groups are tolerated. Notably, a one-pot C–C activation/C–C coupling has been realized. Computational and experimental mechanistic studies indicate that the reaction involves a ruthenium(II) monohydride-mediated C(aryl)–C(aryl) activation and the resting state of the catalyst is a η⁴-coordinated ruthenium(II) dichloride complex, which could inspire development of other transformations based on this reaction mode.

Graphical Abstract

INTRODUCTION

Oxidative addition of a transition metal (TM) into a carbon–carbon (C–C) bond represents an important means of activating C–C bonds, which has led to the development of numerous synthetically valuable methods.¹ This process converts one relatively inert C–C bond into two more reactive C–TM bonds that can undergo further transformations, affording dual functionalization of both carbon terminuses (Scheme 1A).² To date, a number of catalytic C–C cleavage/functionalization methods have been developed on the basis of such a mode of activation. However, the scope of C–C bonds that can undergo oxidative addition with TMs is still narrow (Scheme 1B). The major class of suitable substrates contains a three- or four-membered ring, in which strain release becomes the main driving force for the C–C cleavage.³ On the other hand, more polar C–C bonds, such as C–CN,⁴ C–carbonyl, and C–iminyl bonds in less strained substrates,⁵ can also be activated by low valent TMs due to favorable interactions between the low-lying σ* orbital in these moieties and TM filled d orbitals, which promotes forming the requisite C–C/TM σ complex.

Scheme 1. C–C Bond Activation via Oxidative Addition.

In contrast, TM insertion into nonpolar and unstrained C–C bonds has been extremely rare. In 1993, Milstein and co-workers reported a phosphine-directed activation of an aryl–alkyl bond in a pincer-type substrate, which was driven by forming a two–five-membered-fused rhodacycle (Scheme 2A).⁶ The catalytic transformation was also developed a few years later by the same group.⁷ Recently, Kakiuchi and co-workers developed a novel Rh-catalyzed cleavage of unstained aryl–allyl bonds, albeit through a β-carbon elimination mechanism.⁸ Activation of unstrained aryl–aryl bonds has been elusive⁹ until our recent work (Scheme 2B).¹⁰ The C(aryl)–C(aryl) bonds in 2,2′-biphenols were catalytically cleaved with hydrogen gas using a rhodium catalyst and phosphinite directing groups (DGs). Despite this promising initial result, ortho substituents in the 2,2′-biphenol substrates were required for this transformation, and our general understanding of activating unstrained aryl–aryl bonds is still limited. A number of questions remain to be addressed. For example, can other types of DGs, besides those strongly coordinative phosphorus-based ones, be used in the C(aryl)–C(aryl) bond activation? Can other TMs besides expensive rhodium be employed as the catalyst? Can other reagents besides hydrogen gas react with the C–C cleavage intermediate? How does a metal catalyst approach the C(aryl)–C(aryl) bond to be cleaved? Answers to these questions could be important for expanding the substrate scope and reaction varieties of this transformation. In this Article, we describe a detailed development of a ruthenium-catalyzed reductive cleavage of unstrained C(aryl)–C(aryl) bonds and the mechanistic study of this reaction (Scheme 2C). Nitrogen-based heterocycles were found to be excellent DGs, and, besides hydrogen gas, secondary alcohols and silanes could also be employed as the reductant for this transformation.

Scheme 2. Activation of Nonpolar Unstrained C–C Bonds.

RESULTS AND DISCUSSION

Pyridine and related heterocycles have been frequently employed as DGs in catalytic C–H activation reactions.¹¹ They have also been used in C–C activation of ketones.¹² Thus, the 2,2′-(3-methylpyrdinyl)-substituted biphenyl (1a) was chosen as the initial substrate. Rhodium-based catalysts were naturally examined first. Using [Rh(C₂H₄)Cl]₂, [Rh-(COD)Cl]₂, or Rh(COD)₂NTf₂ as the catalyst, trace or no desired product was observed (Table 1, entries 1, 2, and 4). However, adding NaI as the additive to the [Rh(COD)Cl]₂-catalyzed reaction, 39% yield of the desired C–C cleavage product was obtained (Table 1, entry 3). This result showed the feasibility of using pyridine as DGs for C(aryl)–C(aryl) bond activation, although the exact role of NaI is still unclear. Note that using the pyridine DG, ortho substitution at the 3,3′ positions was not required, which is distinct from the prior 2,2′-biphenol activation.¹⁰ This motivated us to test other readily available TM complexes as precatalysts for this transformation. While the Ni(0), Co(0), and Ir(I) complexes gave no desired cleavage product (entries 5–7), Ru(II) dichloride complexes nevertheless exhibited remarkable reactivity (entries 11–16). RuCl₃· xH₂O showed moderate reactivity (entry 10), but Ru₃(CO)₁₂^12b and Cp*Ru(COD)Cl were not reactive. Among the various Ru complexes examined, Ru(COD)Cl₂ was found to be most efficient (entry 14). Besides 1,4-dioxane, other solvents, such as toluene and THF, were also suitable for this transformation (entries 15 and 16).

Table 1.

Selected Optimization Study for the Hydrogenation Condition

entrya	conditions	yieldb
1	10 mol % [Rh(C2H4)Cl]2	trace
2	10 mol % [Rh(COD)Cl]2	n.d.
3	10 mol % [Rh(COD)Cl]2 + 100 mol % NaI	39%
4	20 mol % Rh(COD)2NTf2	n.d.
5	20 mol % Ni(COD)2	n.d.
6	10 mol % Co2(CO)8	n.d.
7	10 mol % [Ir(COD)Cl]2	n.d.
8	6.7 mol % Ru3(CO)12	n.d.
9	20 mol % Cp*Ru(COD)Cl	n.d.
10c	20 mol % RuCl3·xH2O	26%
11	10 mol % [Ru(p-cymeme)Cl2]2	50%
12	10 mol % [Ru(p-cymeme)I2]2	66%
13	20 mol % Ru(COD)Cl2	89%
14	10 mol % Ru(COD)Cl2	89% (83%)
15	10 mol % Ru(COD)Cl2, tol as solvent	88%
16	10 mol % Ru(COD)Cl2, THF as solvent	77%

^aReaction conditions: 1a (0.1 mmol), 20 mol % monomer or 10 mol % dimer or 6.7 mol % trimer of metal catalyst, 1,4-dioxane (0.075 M), 130 °C, 18 h, Q-tube filled with 150 psi H₂ gas.

^bUnless otherwise noted, the yields were determined by 1H NMR using 1,1,2,2-tetrachloroethane as internal standard; n.d. = not detected; the yield in parentheses is isolated yield.

^cThe catalyst loading was based on the formula of RuCl₃.

Alternative reductants besides H₂ gas were then sought, which, if successful, could provide a more convenient way to operate this C–C cleavage reaction (Table 2). To our delight, a variety of mild reductants were found reactive under this Ru-catalyzed condition, and afforded the desired product. For example, potassium formate salt and Hantzsch ester gave 9% and 58% yields of product 2a, respectively (entries 2 and 3). Diverse secondary alcohols could also serve as a hydride source through a transfer hydrogenation process (entries 4–8). Among all of the alcohols tested, cyclopentanol proved to be most efficient (entry 8), although an excess amount was needed for a higher conversion (entries 9–13). 84% yield was achieved using 50 equiv of cyclopentanol with toluene as solvent. In addition, a combination of silane and water (1:1) was found to be an excellent reductant (entries 14–18).¹³ An optimal result (85% yield) was obtained when using 5.0 equiv of diphenylmethylsilane with 5.0 equiv of H₂O (entry 18).

Table 2.

Screening for Alternative Reductants^a

entrya	conditions	yieldb
1	10 equiv of HCOONH4	n.d.
2	10 equiv of HCOOK	9%
3	10 equiv of Hantzsch ester	58%
4	10 equiv of diphenylmethanol	20%
5c	2-propanol	11%
6c	2-butanol	22%
7c	3-pentanol	47%
8c	cyclopentanol	70%
9	50 equiv of cyclopentanol	66%
10	30 equiv of cyclopentanol	62%
11	10 equiv of cyclopentanol	51%
12	50 equiv of cyclopentanol, Tol as solvent	84% (81%)
13	50 equiv of cyclopentanol, THF as solvent	46%
14	10 equiv of (TMS)3SiH + 10 equiv of H2O	12%
15	10 equiv of Et3SiH + 10 equiv of H2O	28%
16	10 equiv of PhMe2SiH + 10 equiv of H2O	42%
17	10 equiv of Ph2MeSiH + 10 equiv of H2O	80%
18	5 equiv of Ph2MeSiH + 5 equiv of H2O	85% (78%)

^aReaction condition: 1a (0.1 mmol), 10 mol % Ru(COD)Cl₂, 1,4-dioxane or other solvents (1.0 mL), 130 °C, 24 h, sealed vial.

^bUnless otherwise noted, the yields were determined by ¹H NMR using 1,1,2,2-tetrachloroethane as internal standard; n.d. = not detected; the yields in parentheses are isolated yields.

^cThe indicated alcohols were used as solvent.

With three high yielding conditions in hand, the substrate scope was investigated next (Table 3). First, besides 3-methylpyridine, simple pyridine can also serve as an effective DG (entry 2). Substitutions on the arene at 3, 4, or 5 positions were all tolerated (entries 3–6), and the yield was lower for the 3,3′-disubstituted substrates likely due to the steric hindrance (2c). In addition, phenyl and furyl-substituted substrates (2g and 2h) showed good reactivity. A range of functionalization groups, such as fluoride (2i), chloride (2j and 2p), bromide (2k), trifluoromethyl (2l), OCF₃ (2m), ester (2n), amide (2o), OMe (2q), and silyl ether (2r), were found compatible. Interestingly, when a fluorine substituent is ortho to the DG (entry 20), partial C–F bond activation/cleavage product was obtained;¹⁴ for comparison, fluorine substitutions at other positions (2i and 2w) were intact. When a ketone moiety was present (entry 21), partial reduction to the corresponding alcohol was observed, particularly under the transfer hydrogenation conditions. Unsurprisingly, alcohol moieties (2u) were tolerated. The bulkier binaphthyl-derived substrate was not reactive, likely due to the steric hindrance around of the C(aryl)–C(aryl) bond (entry 23). Regarding the scope of DGs, substituted pyridines with various electronic properties exhibited similar reactivity (entries 24–26). Gratifyingly, other heteroarenes, including pyrimidine (entry 27), five-membered pyrazole (entry 28), and quinoline (entry 29), were found as competent DGs. More labile oxazoline (1ac), oxazole (1ad), and nitrile (1ae) were ineffective. Finally, attempts to cleave an aryl–pyridyl bond or use a monobidentate DG were unfruitful at this stage (entries 33 and 34).

Table 3.

Substrate Scope^a

^aCondition 1: biaryl 1 (0.1–0.2 mmol), Ru(COD)Cl₂ (10 mol %), 1,4-dioxane (0.075 M), 130 °C, 18 h, Q-tube filled with 150 psi H₂ gas. Condition 2: biaryl 1 (0.1–0.2 mmol), Ru(COD)Cl₂ (10 mol %), 5.0 equiv of Ph₂MeSiH, 5.0 equiv of H₂O, 1,4-dioxane (1.0 mL/0.1 mmol 1), 130 °C, 24 h, sealed vial. Condition 3: biaryl 1 (0.1–0.2 mmol), Ru(COD)Cl₂ (10 mol %), 50 equiv of cyclopentanol, toluene (1.0 mL/0.1 mmol 1), 130 °C, 24 h, sealed vial. All yields are isolation yields.

^bReaction time was 6 h.

^cReaction time was 3 h.

^dReaction time was 11 h.

^eThe total yields are isolation yields, and the ratio of the two products was determined by ¹H NMR.

^fThe two products were both observed and isolated from the reaction system.

^gRu(COD)Cl₂ (20 mol %), 160 °C.

^hRu(COD)Cl₂ (20 mol %), 150 °C.

ⁱ(EtO)₃SiH (5.0 equiv) was used instead of Ph₂MeSiH. n.d. = not detected.

The limits of the catalyst loading and reaction temperature under the hydrogenation condition were further investigated (Table 4). Reducing the Ru loading from 10 to 2.5 mol % only marginally affected the yield (entry 1, Table 4); further lowering the catalyst loading to 1 mol % still afforded 55% yield of the product (entry 2, Table 4). It was surprising that, at a lower temperature (110 °C), a higher yield (90%) was obtained (entry 4, Table 4). Further decreasing the temperature to 70 °C still showed moderate reactivity (entries 5 and 6, Table 4). The hydrogen pressure could be further reduced to 70 psi without affecting the reaction efficiency (entries 7 and 8, Table 4). A lower yield (65%) was obtained when 30 psi of hydrogen was used (entry 9, Table 4).

Table 4.

Exploring the Limits of Hydrogenolysis of the C(aryl)–C(aryl) Bonds

entrya	X (mol %)	T (°C)	Y (psi)	yieldb
1c	2.5	130	150	80%
2d	1.0	130	150	55%
3e	10	130	150	83%
4	10	110	150	90%
5	10	90	150	59%
6	10	70	150	33%
7	10	130	110	90%
8	10	130	70	89%
9	10	130	30	65%

^aConditions: 1a (0.2 mmol), Ru(COD)Cl₂, 1,4-dioxane (C cat = 0.0075 M), 18 h, Q-tube filled with H₂ gas.

^bIsolated yield.

^c0.4 mmol of 1a was used.

^d1.0 mmol of 1a was used.

^e0.1 mmol of 1a was used.

In addition, a one-pot C–C activation/C–C formation approach has also been established (eq 1). After the hydrogenolysis of the aryl–aryl bond, the ruthenium catalyst was found to remain active. Subsequent addition of acrylate allowed for mono ortho alkylation of the C–C cleavage product in a high yield.¹⁵ This result shows the potential to couple aryl–aryl bond activation with subsequent functionalization using a single catalyst.

(1)

MECHANISTIC STUDIES

The mechanism of the Ru-catalyzed aryl–aryl bond activation was explored using a combination of computational and experimental efforts. Three possible reaction pathways are proposed (Figure 1). Path a involves insertion of a Ru(II) dichloride species (“RuCl₂”) into the aryl–aryl bond to give a Ru(IV) intermediate, which then undergoes hydrogenolysis to give the monomer product. Path b is initiated by a Ru(II) monohydride monochloride species (“RuHCl”), generated via monohydrogenation of the ruthenium dichloride precursor.¹⁶ Oxidative addition of the “RuHCl” into the aryl–aryl bond followed by C–H reductive elimination affords one monomer product, and the resulting ruthenium aryl intermediate then reacts with H₂ to deliver the other monomer product and regenerate the “RuHCl” catalyst. Path c is based on a Ru(II) dihydride (“RuH₂”) species, generated from double hydrogenation of the “RuCl₂” precursor.¹⁷ Similarly, insertion of the “RuH₂” intermediate into the aryl–aryl bond, followed by double C–H reductive elimination, should afford two monomer products. The resulting Ru(0) can then react with H₂ to regenerate the “RuH₂” species (for a discussion of an alternative Ru(0)-initiated pathway, see the Supporting Information).

Figure 1.

Proposed possible reaction pathways.

DFT Calculation.

To differentiate the three possible pathways, density functional theory (DFT) calculations were performed. It was found that the “RuHCl” pathway (path b) was the most favorable. The computed energy profile in Figure 2 shows that “RuHCl” complex 5b is the active catalyst species in the catalytic cycle, which is formed from the endothermic reaction of RuCl₂ species 5a with H₂ via TS1 with a barrier of 21.3 kcal/mol. In both 5a and 5b, the two pyridine DGs adopt a trans geometry. This places the target aryl–aryl bond in closer proximity to the Ru, which is evidenced by the short Ru···C distances of 2.3–2.6 Å in 5a and 5b (vide infra, Figure 5, the X-ray structure of 5ah). This agostic C–C/Ru coordination leads to a low barrier of 12.2 kcal/mol for the subsequent C(aryl)–C(aryl) oxidative addition transition state TS2b with respect to 5b. The overall activation free energy of TS2b is 24.5 kcal/mol with respect to the resting state 5a. In contrast, the experimentally observed low reactivity of biaryl substrates with only one pyridine substituent (e.g., 1af) can be attributed to the lack of the agostic C–C coordination with the Ru (see Figure S7.2.2 for detailed computational results). The necessity of two DGs for the C(aryl)–C(aryl) bond cleavage has also been demonstrated in the catalytic activation of the C(aryl)–C(aryl) bonds of 2,2′-biphenols by installing phosphinites as DGs in our prior study.¹⁰ After the C–C cleavage step, the ensuing C–H reductive elimination (TS3) and σ-bond metathesis with H₂ (TS4) both occur with low barriers, leading to two monomer products (2a) and regenerating the “RuHCl” catalyst (5b).

Figure 2.

DFT-computed reaction energy profile of the C(aryl)–C(aryl) bond activation of substrate 1a catalyzed by a Ru monohydride monochloride catalyst (path b).

Figure 5.

Capture of the resting state of the catalyst.

The possibility of the “RuCl₂” and “RuH₂” pathways is also considered, and the key results are summarized in Figure 3. In the “RuCl₂” pathway (path a, Figure 3A), although the oxidative addition of C(aryl)–C(aryl) (TS2a) requires a relatively low barrier of 20.8 kcal/mol with respect to the “RuCl₂” species 5a, the resulting octahedral Ru intermediate 6a is coordinatively saturated and thus incapable of binding with H₂ and undergoing hydrogenolysis of the Ru–C(aryl) bonds. The transition state TS5 for H₂ cleavage has a barrier of 28.6 kcal/mol, even higher than that of the C(aryl)–C(aryl) cleavage (TS2b) in the “RuHCl” pathways (Figure 2). Our calculations indicated several other possible pathways of 6a reacting with H₂, including the one via dissociation of one of the pyridine DGs or one chloride ligand,¹⁸ all require high barriers (see details in Figure S7.2.3). Figure 3B shows that the formation of the RuH₂ complex 5c from 5a is endergonic by 34.8 kcal/mol. This results in a highly disfavored C(aryl)–C(aryl) oxidative addition transition state TS2c (ΔG^‡ = 53.0 kcal/mol with respect to 5a) via the ruthenium dihydride complex (see details in Figure S7.2.4). Taken together, these computational results indicate that the “RuCl₂” and “RuH₂” pathways are both dis-favored. In addition, our DFT calculations show that, although the reductive elimination of RuH₂ (5c) to form a Ru(0) species is energetically feasible, the Ru(0) pathway requires very high activation barriers for the C(aryl)–C(aryl) oxidative addition and the further hydrogenolysis steps (see details in Figure S7.2.5). Therefore, the DFT calculations suggested the “RuHCl” pathway (path b, Figure 1) is the most feasible.

Figure 3.

DFT computed pathways for the C(aryl)–C(aryl) bond activation of substrate 1a catalyzed by Ru dichloride and Ru dihydride catalysts. Gibbs free energies and enthalpies (in parentheses) are in kcal/mol with respect to the RuCl₂ complex 5a.

Kinetic Studies.

To validate the computational results that favor the “RuHCl”-mediated C–C activation pathway, the following kinetic studies were performed. First, the fate of 1,5-cyclooctadiene (COD) on the Ru precatalyst was determined. It was found that the COD ligand was hydrogenated to cyclooctene and cyclooctane in high efficiency at the beginning of the reaction (Scheme 3, eq 2). This result indicated that COD is likely not involved in the catalytic cycle. Second, the kinetic profile of the reaction with substrate 1a was measured. The initial-rate method was employed to determine the reaction order of each component. The reaction was found to exhibit first-order dependence on the concentration of Ru(COD)Cl₂, zero-order on [substrate 1a] and [COD], and pseudo zero-order on [H₂] under a higher pressure, but some rate dependence on [H₂] was observed under a relatively low H₂ gas pressure (<40 psi) (Scheme 3, eq 3, and see Table S5.2.3 for details).¹⁹ These kinetic data are consistent with the DFT calculation (vide supra, Figure 2), which suggests that the oxidative addition step is the turnover-limiting step (TLS).

Scheme 3. Kinetic Studies with the Model Substrate.

In addition, Hammett plot analysis was conducted to investigate the sensitivity of the reaction to electronic changes (Figure 4).²⁰ The results indicated that the electron-withdrawing substituents on the arenes could promote the reaction to some extent, while electron-donating groups slowed the reaction. This observation is also consistent with the DFT calculated results, in which the oxidative addition step is predicted to be the TLS, as the electron-deficient bonds typically promote oxidative addition.

Figure 4.

Hammett plot.

Resting State.

Considerable endeavors have been made to capture the DFT calculated resting state, “RuCl₂” species 5a. After numerous attempts, heating a mixture of substrate 1ah and 30 mol % of Ru(COD)Cl₂ under 150 psi H₂ atmosphere in 1,4-dioxane at 130 °C for 20 min afforded a dark green metal complex (5ah). The structure of complex 5ah was unambiguously determined by X-ray crystallography (Figure 5), which is consistent with the proposed “RuCl₂” resting state by DFT (vide supra, Figure 2). In complex 5ah, the metal center exhibits octahedral geometry with the two pyridine DGs adopting a trans spatial relationship. An interesting η⁴-coordination mode between two arene π bonds and the Ru center was observed;²¹ in particular, the bond lengths between the Ru center and the carbons to be cleaved are short: ca. 2.2 Å. Thus, this structure shows that the Ru(II) center is very close to the target C(aryl)–C(aryl) bond. As compared to d⁸ Rh(I) that favors a square planar geometry, the d⁶ Ru(II) can easily form a 18 electron complex through coordination with two arene π bonds. The agostic interaction with the C(aryl)–C(aryl) bond, as illustrated in the structure of 5ah, is anticipated to be important for the desired C–C bond activation.

In addition, the reaction of substrate 1ah was monitored by ¹H NMR, in which the “RuCl₂” species 5ah was observable from the very beginning to almost the end of the reaction by comparing the ¹H NMR spectrum of the crude mixture with that of the isolated 5ah (Figure 6).

Figure 6.

Monitoring the reaction of substrate 1ah.

Moreover, complex 5ah is catalytically active. When 10 mol % of the 5ah was employed as the catalyst under otherwise identical conditions, 91% yield of the desired product was obtained (eq 4). Therefore, all of the above observations are consistent with the proposal of having the “RuCl₂” species (5a) as the resting state of the catalyst.

(4)

Other Control Experiments.

To explore the “RuHCl”-mediated reaction pathway, a ruthenium monohydride monochloride complex 17 was synthesized and subjected to the reaction with substrate 1a in the absence of hydrogen gas (Scheme 4, eq 5).²² To our delight, 60% yield of the desired monomer product was obtained on the basis of the hydride complex.²³ For comparison, the analogous ruthenium dihydride complex 18²⁴ gave only trace product via LC–MS analysis under the same reaction conditions (Scheme 4, eq 6). These results suggest the important role of the chloride ligand in the Ru-catalyzed C(aryl)–C(aryl) bond activation.

Scheme 4. Further Control Experiments.

To further examine the possibility of the Ru(0)-initiated pathway, a control experiment was run with 60 mol % of Mn added to the reaction, as Mn metal is known to be capable of reducing Ru(II) to Ru(0) (Scheme 4, eq 7).²⁵ To our surprise, the reaction with Mn not only gave a lower overall yield on the C–C cleavage products, but also generated a significant amount of arene hydrogenation product 2a′. It is intriguing that the neutral benzene ring is selectively reduced instead of the more electron-deficient pyridine ring, implying a possible directed hydrogenation by a Ru(0) catalyst.²⁶ For comparison, such an over-reduction product was almost not observed under the standard reaction conditions. This experiment suggests that the Ru(0) is unlikely to be the actual catalyst for the activation of the C(aryl)–C(aryl) bonds.

CONCLUSION

In summary, to explore a fundamental reaction mode, we have conducted a detailed study of an unusual Ru-catalyzed activation of unstrained C(aryl)–C(aryl) bonds. The reaction limits and substrate scopes have been carefully examined. Besides hydrogen gas, a number of other reagents, such as Hantzsch ester, silanes, and alcohols, have also been found effective to serve as terminal reductants for the reductive cleavage. Various heterocycles, such as pyridine, quinoline, pyrimidine, and pyrazole, can be employed as DGs. In addition, a range of functional groups are compatible under the reaction conditions. Moreover, a one-pot C–C activation/C–C coupling has been realized. Finally, the reaction mechanism has been investigated through collaborative efforts between DFT calculations and experiments. The involvement of a ruthenium(II) monohydride-mediated C(aryl)–C(aryl) activation and a η⁴-coordinated ruthenium(II) complex as the resting state should have broad implications beyond this work. The knowledge obtained in this study may improve our understanding on activating strong, nonpolar, and unstrained chemical bonds. Efforts on expanding the reaction mode to nonreductive processes are ongoing in our laboratories.