Selective Aliphatic Carbon–Carbon Bond Activation by Rhodium

Article pubs.acs.org/accounts

Selective Aliphatic Carbon−Carbon Bond Activation by Rhodium Porphyrin Complexes Ching Tat To and Kin Shing Chan* Department of Chemistry, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, People’s Republic of China CONSPECTUS: The carbon−carbon bond activation of organic molecules with transition metal complexes is an attractive transformation. These reactions form transition metal−carbon bonded intermediates, which contribute to fundamental understanding in organometallic chemistry. Alternatively, the metal−carbon bond in these intermediates can be further functionalized to construct new carbon−(hetero)atom bonds. This methodology promotes the concept that the carbon−carbon bond acts as a functional group, although carbon−carbon bonds are kinetically inert. In the past few decades, numerous efforts have been made to overcome the chemo-, regio- and, more recently, stereoselectivity obstacles. The synthetic usefulness of the selective carbon−carbon bond activation has been significantly expanded and is becoming increasingly practical: this technique covers a wide range of substrate scopes and transition metals. In the past 16 years, our laboratory has shown that rhodium porphyrin complexes effectively mediate the intermolecular stoichiometric and catalytic activation of both strained and nonstrained aliphatic carbon−carbon bonds. Rhodium(II) porphyrin metalloradicals readily activate the aliphatic carbon(sp3)−carbon(sp3) bond in TEMPO ((2,2,6,6-tetramethylpiperidin-1-yl)oxyl) and its derivatives, nitriles, nonenolizable ketones, esters, and amides to produce rhodium(III) porphyrin alkyls. Recently, the cleavage of carbon−carbon σ-bonds in unfunctionalized and noncoordinating hydrocarbons with rhodium(II) porphyrin metalloradicals has been developed. The absence of carbon−hydrogen bond activation in these systems makes the reaction unique. Furthermore, rhodium(III) porphyrin hydroxide complexes can be generated in situ to selectively activate the carbon(α)−carbon(β) bond in ethers and the carbon(CO)−carbon(α) bond in ketones under mild conditions. The addition of PPh3 promotes the reaction rate and yield of the carbon−carbon bond activation product. Thus, both rhodium(II) porphyrin metalloradical and rhodium(III) porphyrin hydroxide are very reactive to activate the aliphatic carbon−carbon bonds. Recently, we successfully demonstrated the rhodium porphyrin catalyzed reduction or oxidation of aliphatic carbon−carbon bonds using water as the reductant or oxidant, respectively, in the absence of sacrificial reagents and neutral conditions. This Account presents our contribution in this domain. First, we describe the chemistry of equilibria among the reactive rhodium porphyrin complexes in oxidation states from Rh(I) to Rh(III). Then, we present the serendipitous discovery of the carbon− carbon bond activation reaction and subsequent developments in our laboratory. These aliphatic carbon−carbon bond activation reactions can generally be divided into two categories according to the reaction type: (i) homolytic radical substitution of a carbon(sp3)−carbon(sp3) bond with a rhodium(II) porphyrin metalloradical and (ii) σ-bond metathesis of a carbon−carbon bond with a rhodium(III) porphyrin hydroxide. Finally, representative examples of catalytic carbon−carbon bond hydrogenation and oxidation through strategic design are covered. The progress in this area broadens the chemistry of rhodium porphyrin complexes, and these transformations are expected to find applications in organic synthesis. long carbon chain in hydrocarbons.2 CCA in homogeneous media gradually has become a research field of much interest because it promotes the efficient use of organic feedstocks. In 1955, Tipper communicated the formation of platinacyclobutane via the CCA of cyclopropane with PtCl2 as the first homogeneous CCA reaction.3 During the past few decades, there have been many examples of stoichiometric and catalytic CCA reactions, which contribute both mechanistic understanding and synthetic usefulness in catalytic transformation.4 In particular, aliphatic C−C bonds are commonly associated with high bond dissociation energies, are surrounded by abundant C−H bonds, and lack coordination properties, which make them inert. The emerging strategies to achieve

1. INTRODUCTION The carbon−carbon bond activation (CCA) of organic substrates with a transition metal complex involves the cleavage of the carbon−carbon bond and the formation of a transition metal−carbon bond (Scheme 1).1 The concept of CCA has been industrially applied since the last century. It is related to the hydrocracking process, which converts petroleum reserves to more demanding commodity fuels by breaking down the Scheme 1. Carbon−Carbon Bond Activation with a Transition Metal Complex

Received: March 28, 2017 Published: June 13, 2017 © 2017 American Chemical Society

1702

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research

Me (tmp = 5,10,15,20-tetramesitylporphyrinato dianion) produces monomeric RhII(tmp) as the permanent metalloradical, where the unpaired electron occupies the axial dz2 orbital.8 However, the sterically less hindered RhII(por) (por = general abbreviation for dianionic porphyrin ligand), such as RhII(oep) (oep = 2,3,7,8,12,13,17,18-octaethylporphyrinato dianion) or RhII(ttp) (ttp = 5,10,15,20-tetra-p-tolylporphyrinato dianion), which will be discussed below, equilibrates with its dimer, RhII2(por)2, as the major species through the reversible cleavage of the weak Rh−Rh bond (12−16.5 kcal mol−1).9−11 Alternatively, the rapid ligand substitution of RhIII(ttp)X (X = Cl, Br, or I) with KOH also yields RhII2(ttp)2 via the reductive elimination of the RhIII(ttp)OH intermediate.12 The hydrolysis of the Rh−C bond in RhIII(ttp)Me provides a neutral condition to access RhII2(ttp)2 via the RhIII(ttp)OH intermediate.13 The equilibrium between the oxidative addition of RhII2(ttp)2 with H2O2 to yield RhIII(ttp)OH and the reverse reductive elimination is quasi-reversible. These complexes are responsible for various types of bond activation chemistry. The metalloradical character of RhII(tmp) can be enhanced through coordination with PPh3 to reversibly form (PPh3)RhII(tmp) by mixing the σ-donor orbital of PPh3 with the half-filled dz2 orbital of RhII(tmp).14 However, a strong σ-donor ligand such as pyridine (py) causes the disproportionation of RhII2(oep)2 to yield a (py)2RhIII(oep)+ cation and RhI(oep)− anion.15 The chemistry of RhIII(por)OH is rarely investigated because of its reactive nature for isolation. The water-soluble [Rh(tspp)] species (tspp = 5,10,15,20-tetra-p-sulfonatophenyl porphyrinato dianion), which includes its [Rh−OD] derivative in aqueous solutions, have been well studied by Wayland.16 However, the observation of lipophilic RhIII(ttp)OH in benzene remains possible,12 and its bond activation chemistry is well supported.

intermolecular and selective CCA in organic substrates generally include ring strain relief, carbonyl functionality, and chelation assistance.5−7 Moreover, the use of low-valent late transition metals to facilitate the oxidative addition of the C−C bond is critical in these systems. We started our studies on CCA reactions in 2000 with a serendipitous discovery. In this Account, we present the interconnected and chronicled development of rhodium porphyrin complexes (Scheme 2) to mediate the activation of Scheme 2. General Structure for Rhodium Porphyrin Complexes, Rh(tmp)L and Rh(ttp)L

aliphatic C−C bonds in various organic substrates developed in this laboratory. This CCA chemistry includes homolytic radical substitution with rhodium(II) porphyrin metalloradicals in nonpolar media and σ-bond metathesis with rhodium(III) porphyrin hydroxides in polar media (Schemes 3 and 4). The presented mechanistic insights were derived from extensive kinetic experiments and findings via DFT calculations.

2. PREPARATION AND INTERCONVERSION OF RHODIUM PORPHYRIN COMPLEXES Rh(I), Rh(II), and Rh(III) are three commonly encountered oxidation states for rhodium porphyrin complexes. Rhodium(III) porphyrin complexes are the common precursors to other Rh(III), Rh(II), and Rh(I) species (Scheme 5). Wayland has reported that the photolysis of the Rh−C bond in RhIII(tmp)-

3. SERENDIPITOUS DISCOVERY OF ALIPHATIC CARBON−CARBON BOND ACTIVATION During the investigation of the 1,2-rearrangement in rhodium porphyrin alkyls,17 TEMPO ((2,2,6,6-tetramethylpiperidin-1yl)oxyl) was used as a mechanistic probe to test the

Scheme 3. Overview of Rhodium(II) Porphyrin Metalloradical-Mediated CCA

1703

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research Scheme 4. Overview of Rhodium(III) Porphyrin Hydroxide-Mediated CCA

Scheme 5. Preparations and Equilibria of Rhodium Porphyrin Species

involvement of free organic radicals. Unexpectedly, RhIII(por)Me was formed (Scheme 6).18 We have also shown that

Table 1. PPh3 Promoted CCA of Non-enolizable Ketones with RhII(tmp)

Scheme 6. Discovery of CCA from 1,2-Rearrangement Investigations

RhII(tmp) (1) cleaves the C−C bond in TEMPO or its deuterated methyl and ethyl derivatives to produce the corresponding RhIII(tmp)Me (2), RhIII(tmp)CD3 (2-d), and RhIII(tmp)Et (3), respectively (Scheme 7).19 We envisioned the CCA reactivity of rhodium(II) porphyrin metalloradical and extended the substrate scope.

Table 2. PPh3 Promoted CCA of Esters with RhII(tmp)

II

Scheme 7. CCA of Nitroxy Radical with Rh (tmp)

4. CCA WITH RHODIUM(II) PORPHYRIN METALLORADICAL: FROM COORDINATING TO NONCOORDINATING SUBSTRATES RhIII(tmp)Me (2). The reactivity order is approximately amide > ester ≈ ketone. The preferential activation of the terminal methyl group suggests that the reaction is sterically sensitive. These clues help us propose that RhII(tmp) (1) abstracts the terminal methyl group from the substrate via a bimolecular homolytic radical substitution.23,24 The α-radical coproduct was not successfully detected or trapped.

4.1. Methyl Abstraction from Ketones, Esters, and Amides

Initially, we examined several coordinating substrates such as ketones, esters, and amides, which were capable of precoordination to rhodium(II) porphyrin prior to CCA. Enolizable ketones are susceptible to α-C−H activation and are consequently not examined.20 Indeed, RhII(tmp) (1) effected the activation of the C(sp3)−C(sp3) bond in nonenolizable ketones, esters, and amides to yield RhIII(tmp)Me (2) (Tables 1−3).21,22 The addition of PPh3 as a promoting ligand increased the yield of RhIII(tmp)Me (2). Substrates of the same class with a weaker C−C bond gave higher yields of

4.2. C(α)−C(β) Bond Activation of Nitriles

Nitriles are popular substrates for CCA studies because the facile cleavage of the sterically less hindered C(α)−C(CN) bond has been extensively used for the transfer hydro1704

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research Table 3. PPh3 Promoted CCA of Amides with RhII(tmp)

increased with the decrease in C(Me)−C(α) bond dissociation energy (BDE).27 Nonselective cleavage of C−C bonds was observed when α-alkylphenylacetonitriles were used as the substrates, which gives a mixture of RhIII(tmp)Me (2), RhIII(tmp)Et (3), and RhIII(tmp)nPr (4) (Table 4, entries 7 and 8). The preference for terminal methyl abstraction is consistent with the CCA of nonenolizable ketones, esters, and amides. The fate of the organic coproduct, presumably via an αcyano alkyl radical intermediate, was unclear. Nevertheless, the possibilities of a radical recombination via either alkyl−alkyl homocoupling or alkyl−RhII(tmp) coupling were experimentally eliminated for sterically bulky nitriles. A hydrogen atom transfer from the parent nitrile to the α-cyano alkyl radical to produce demethylated nitrile is a possible decay pathway in some cases (Table 4, entries 2, 3, and 5). The reaction order of RhII(tmp) (1) with tBuCN under saturated PPh3 conditions was measured, and the observed rate law was obtained: rate = k + kobs[RhII(tmp)][tBuCN]. The k term was attributed to the PPh3-induced thermal decomposition of (PPh3)RhII(tmp) at 130 °C, which is likely via disproportionation.14,28 According to the proposed mechanism in Scheme 8, the kobs[RhII(tmp)][tBuCN] term suggests that two parallel processes are operating: (1) a nitrile-independent process because of the PPh3-promoted disproportionation at elevated temperature from RhII to RhIII and RhI and (2) a nitrile-precoordination assisted aliphatic CCA of nitrile. The first-order dependence on RhII(tmp) (1) and tBuCN provides the first kinetic evidence for the homolytic bimolecular radical substitution with RhII(tmp) (1).

carbylation or cyanation reaction.25 Therefore, we have examined the CCA of more coordinating nitriles with RhII(tmp) (1). Delightfully, RhII(tmp) (1) selectively cleaved the aliphatic C(α)−C(β) bond in nitriles to give good yields of RhIII(tmp)Me (2) (Table 4, entries 2−6).26 When acetonitrile Table 4. PPh3 Promoted CCA of Nitriles with RhII(tmp)

4.3. Methyl Abstraction from TEMPO

After quantitatively understanding the aliphatic CCA of nitrile with RhII(tmp) (1), we examined the detailed mechanism of the high-yielding aliphatic CCA of TEMPO with RhII(tmp) (1). RhII(tmp) (1) reacted with TEMPO to produce 85% yield of RhIII(tmp)Me (2) as the organometallic product (Scheme 9). The corresponding organic coproduct nitrone was too unstable to be detected. Nonetheless, when we attempted CCA using 2,5-dimethyl-2,5-diphenylpyrrolidin-1-oxyl (DMPNO) as the substrate, a close analogue of TEMPO, 2-methyl-2,5diphenyl-3,4-dehydronitrone was found with the formation of RhIII(tmp)Me. The reaction stoichiometry of CCA was supported. Meanwhile, RhIII(tmp)H (5), which was produced from the minor C−H activation (CHA) channel, undergoes fast hydrogen atom transfer (HAT) with TEMPO to regenerate RhII(tmp) (1) and the observed TEMPO-H in 4% yield.29 The coordination of TEMPO to RhII(tmp) (1) is evidenced by the spectrophotometric titration with binding constant K = (2.24 ± 1.2) × 104 M−1 at 20 °C. The temperature-dependent

yield (%) entry

nitrile

1 2 3 4 5 6 7 8

MeCN EtCN Me2CHCN t BuCN PhMeCHCN PhMe2CCN PhCH(Et)CN PhCH(nPr)CN

a

C(Me)−C(α) BDE (kcal mol−1) a 83.2 79.5 74.7 63 59.9 a a

27

time (h)

2

3

4

60 48 48 24 36 24 48 48

10 38 41 52 48 61 30 27

23 16

9

Not applicable.

without a C(α)−C(β) group was used, the C(α)−C(CN) bond cleavage in CH3CN occurred inefficiently to give a low yield of RhIII(tmp)Me (2) (Table 4, entry 1). The reaction rates and yields of the CCA product RhIII(tmp)Me (2)

Scheme 8. Proposed Mechanism for the CCA of Nitriles with RhII(tmp)

1705

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research Scheme 9. CCA and CHA of TEMPO with RhII(tmp)

Scheme 10. CCA of TEMPO via an Initial Electron Transfer

Scheme 11. CCA of Cyclooctane with a Rhodium Porphyrin Complex

Table 5. Test for Rhodium Porphyrin Intermediates in CCA of Cyclooctane

yield (or recovery) (%) entry

X

time

RhIII(ttp)(c-octyl)

RhIII(ttp)(n-octyl)

RhII2(ttp)2

RhIII(ttp)H

1 2 3

Rh(ttp) H H

15 h 15 h 12 d

41 0 0

4 21 94

0 0 2

46 73 0

mol for ionic), the ionic pathway with electron transfer is unlikely to operate in a nonpolar benzene solvent.

equilibrium constants yield the enthalpy and entropy values of ΔH = −14.7 ± 1.4 kcal mol−1 and ΔS = −30.6 ± 4.6 cal mol−1, respectively. η1-O-bound [RhIII(tmp)(TEMPO)] was the lowest-energy adduct in the DFT calculations. The first-order dependence on both RhII(tmp) (1) and TEMPO was spectrophotometrically measured to obtain an overall rate law of rate = kobs[RhII(tmp)][TEMPO], where kobs = 0.11 L mol−1 s−1 at 70 °C. The corresponding activation enthalpy and entropy were ΔH⧧ = 15.02 ± 0.71 kcal mol−1 and ΔS⧧ = −20.93 ± 0.88 cal mol−1, respectively. However, we could not distinguish whether the CCA process proceeded with or without the precoordination of TEMPO. Then, we interpreted a parallel major CCA process with a minor CHA process based on the product distribution. Moreover, the hydrogen atom transfer from RhIII(tmp)H 1 to TEMPO to produce TEMPOH is not rate-limiting, for example, kHAT ≫ (kCCA + kCHA). In addition to the proposed radical pathway, DFT calculations suggest that an ionic CCA pathway is viable. This process involves the initial electron transfer from TEMPO to RhII(tmp) to produce a RhI(tmp) anion and a TEMPO cation and the nucleophilic substitution at the methyl carbon to form RhIII(tmp)Me with nitrone as the leaving group (Scheme 10). Although the calculated activation energies for both processes are low (+11.7 kcal/mol for radical and +17.4 kcal/

4.4. RhII-Catalyzed 1,2-Addition of Cyclooctane with RhIII(ttp)H

During the course of our investigation on the C−H activation of unfunctionalized hydrocarbons with RhII(por), 30 we serendipitously discovered the CCA of cyclooctane. Unexpectedly, an attempt of C−H activation of cyclooctane using RhIII(ttp)Cl (6) with added K2CO3 yielded RhIII(ttp)(n-octyl) (7) in 33% yield (Scheme 11).31 The C−C bond of cyclooctane was activated. Although the C−H activation product RhIII(ttp)(c-octyl) (8) was observed, it was not the active intermediate that caused RhIII(ttp)(n-octyl) (7) because of its inefficient conversion to 7. Hence, we investigated two other possible intermediates, RhIII(ttp)H (9) and RhII2(ttp)2 (10). Independent reactions of 9 or 10 with cyclooctane gave only low yields of RhIII(ttp)(n-octyl) (7) in 15 h, which suggests that they were also not the active species in cleaving the C−C bond of cyclooctane (Table 5, entries 1 and 2). Nonetheless, prolonged heating of RhIII(ttp)H (9) with cyclooctane for a period of 12 d produced the CCA product RhIII(ttp)(n-octyl) (7) in 94% yield (Table 5, entry 3).32 The formation of RhIII(ttp)(n-octyl) (7) is a formal 1,2addition of RhIII(ttp)H (9) across the C−C bond of cyclooctane. We reason that the RhII(ttp) metalloradical that 1706

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research Scheme 12. Proposed RhII(ttp)-Catalyzed 1,2 Addition of Cyclooctane with RhIII(ttp)H

dissociates from RhII2(ttp)2 (10), which is formed from the dehydrogenative dimerization of RhIII(ttp)H (9), catalyzes such 1,2-addition reaction in analogy to the RhII(oep)-catalyzed 1,2addition of RhIII(oep)H to styrene to produce RhIII(oep)CH2CH2Ph, as reported by Halpern.33 The proposed radical chain mechanism in Scheme 12 was ascertained by systematically varying the reagent concentration ratio between RhIII(ttp)H (9) and RhII2(ttp)2 (10). Indeed, the yield of RhIII(ttp)(noctyl) (7) increases with decreased loading of RhII2(ttp)2, while the C−H activation channel is suppressed (Table 6).

contrast to the CCA of cyclooctane, RhII(tmp) (1) alone cleaved the benzylic C−C bond in PCP to yield the dirhodium bis-benzylic product 11 with no observed intermediate (Scheme 13).13 Benzylic C−H activation was not observed, likely because of the sterically inhibited transition state (Scheme 14). Scheme 13. Bimetalloradical CCA of PCP with RhII(tmp)

Table 6. RhII(ttp)-Catalyzed CCA of Cyclooctane with RhIII(ttp)H

yield/% III

II

III

entry

Rh (ttp)H/Rh 2(ttp)2

Rh (ttp)(c-octyl)

RhIII(ttp)(n-octyl)

1 2 3 4

1:0 1:0.5 1:0.2 1:0.1

0 60 53 0

21 18 26 73

Scheme 14. Sterically Inhibited C−H Activation of PCP with RhII(tmp)

RhIII(ttp)(n-octyl) (7) was selectively formed in 73% yield from the reaction with 10:1 ratio of RhIII(ttp)H (9) and RhII2(ttp)2 (10). The control of C−C over the C−H activation is naturally kinetic. The bimetalloradical C−H activation channel is suppressed to a greater extent by decreasing the RhII(ttp) concentration because of its second-order dependence, which makes the CCA channel dominant. The calculated activation barrier of RhII(ttp)-metalloradicalmediated CCA (ΔG⧧ = 36.8 kcal mol−1) is reasonably consistent with the experimental conditions.32 We did not consider any electron transfer pathway in this case because both substrate and solvent medium are nonpolar in nature. The direct reaction of RhIII(ttp)H (9) with cyclooctane to yield RhIII(ttp)(n-octyl) (7) also has unreasonably high barriers from the calculations. 4.5. RhII Bimetalloradical CCA of [2.2]Paracyclophane

The CCA of cyclooctane demonstrates that the rhodium(II) porphyrin metalloradical is a promising reagent in cleaving the aliphatic C−C bond of hydrocarbons. To eliminate the downside of an interfering C−H activation channel, we selected [2.2]paracyclophane (PCP) in the next study. Using the thermodynamically unfavorable formation of [2.2]paracyclophana-1-ene from PCP via the formation of a RhIII(tmp)(cyclophanyl) intermediate by the benzylic C−H activation34 we fortunately discovered the CCA of PCP. In

Kinetic studies reveal that RhII(tmp) (1) is second-order dependent and PCP is first-order dependent, for example, rate = kobs[RhII(tmp)]2[PCP]. The precoordination of PCP to RhII(tmp) (1), probably via a π−π interaction, was not observed using 1H NMR spectroscopy. Thus, RhII(tmp) has two possible roles: (1) an active CCA reagent or (2) a slow radical-trapping reagent upon the self-homolysis of the benzylic C−C bond in PCP. Roth measured the rate of PCP homolysis 1707

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research

5. CCA WITH RHODIUM(III) PORPHYRIN HYDROXIDE: MILD AND SELECTIVE TOOL

by trapping with saturated NO or O2. The rate constant at 220 °C is approximately 10−3 s−1.35 However, the rate constants of CCA of PCP with RhII(tmp) (1) are 10−102 M−2 s−1 at 140− 170 °C. The latter values are greater by at least 4 orders of magnitude than the PCP homolysis. Hence, an active role for RhII(tmp) (1) in cleaving the benzylic C−C bond of PCP is supported. Since the overall reaction order is three, we propose that two RhII(tmp) molecules approach the benzylic C−C bond from both ends in the termolecular transition state in analogy to the bimetalloradical C−H activation with RhII(tmp) (Scheme 13).11 An estimation of this transition state geometry by the Spartan ’10 software at the PM3 level using meso-aryl omitted porphyrin as a simplified model suggests that the two RhII(por) molecules interact with the respective benzylic carbons in PCP at inclined angles (Scheme 15). The distortion of rhodium porphyrin from planarity is observed, which is probably a result of steric congestion.

5.1. C(α)−C(β) Bond Activation of Ethers

During the investigation of the C−H activation of various substrates, RhIII(por)X (X = Cl, Br, I) was usually used as a convenient precursor to RhIII(por)OH via ligand substitution with a base. When we examined the attempted C−H activation of aliphatic ethers with RhIII(tmp)I (13) in the presence of KOH, the C(α)−C(β) bond of ethers was selectively cleaved with the β-alkyl group transferred to rhodium porphyrin to yield the corresponding RhIII(tmp)R (R = alkyl) (Table 7, Table 7. Selective C(α)−C(β) Bond Activation of Ethers with Rhodium Porphyrin Complexes

Scheme 15. Computed Transition State Geometry of CCA of PCP with RhII(por)

yield (%) entry

(RCH2)2O

RhIII(tmp)R

conditions 1

1 2 3 4 5 6

(MeCH2)2O (EtCH2)2O (nPrCH2)2O (nBuCH2)2O (AmylCH2)2O (Me2CH)2O

RhIII(tmp)Me RhIII(tmp)Et RhIII(tmp)nPr RhIII(tmp)nBu RhIII(tmp)Amyl RhIII(tmp)Me

30 86 88 58 32

conditions 2 10, 62, 83, 79, 54, 85,

1d 10 min 10 min 10 min 2h 10 min

conditions 1).36 Alternatively, the in situ generation of RhIII(tmp)OH through the disproportionation of RhII(tmp) (1) with H2O promoted by PPh3 under phase transfer conditions effectively mediated the same transformation at room temperature (Table 7, conditions 2).37 Alkyl formate or alkyl acetate was formed as the organic coproduct from the CCA of linear or branched ethers, respectively. This result establishes the reaction stoichiometry in Scheme 17. The cleavage of the C(α)−C(β) bond by RhIII(tmp)OH through σbond metathesis is consistent with the incorporation of an extra

By regenerating RhII2(ttp)2 (10) from the hydrolysis of RhIII(ttp)Me, the thermal reaction of PCP in the presence of excess H2O and 10 mol % RhIII(ttp)Me causes the catalytic formation of 4,4′-dimethylbibenzyl 12 (Scheme 16). This Scheme 16. Catalytic Hydrogenation of the C−C σ-Bond in PCP with H2O

Scheme 17. Proposed Mechanism for the C(α)−C(β) Bond Activation of Ethers

spontaneous transformation is further supported by calculation at the Hartree−Fock/3-21G level to obtain the reaction enthalpy, entropy, and free energy of −14.5 kcal mol−1, −11.5 cal mol−1 K−1, and −9.0 kcal mol−1, respectively, at 200 °C. The ring strain relief contributes to the driving force of the reaction. Therefore, the benzylic C−C bond in PCP is reduced by H2O to form two new C−H bonds without the addition of a co-reductant, which successfully couples catalytic CCA and water splitting. 1708

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research oxygen atom into the organic coproduct. A mild-temperature selective oxidation of the C(α)−C(β) bond of ethers with H2O was achieved. The origin of the regioselective C(α)−C(β) bond cleavage is attributed to its weaker bond compared to the less steric but stronger terminal R−CH3 bond.38 The transfer of the β-alkyl group to rhodium is proposed to be directed by the charge interaction between the [Rh−OH] moiety and ether in a preorganized structure (Scheme 18). The sterically less

Table 8. Selective C(CO)−C(α) Bond Activation of Ketones with RhIII(ttp)Me

Scheme 18. Proposed Transition State for the σ-Bond Metathesis of Ethers with RhIII(tmp)OH

hindered R group compared to the CH2OR′ group minimizes the repulsion to the porphyrin ligand. The σ-bond metathesis pathway was also proposed for the benzylic C−H activation of toluene with RhIII(ttp)Me to yield RhIII(ttp)(benzyl) and CH4.39 In addition, the axial coordination of two ligands in the cis-manner in metalloporphyrin complexes is sterically accessible.40

a

32% recovery of RhIII(ttp)Me. b39% recovery of RhIII(ttp)Me.

Scheme 19. CCA of 2,6-Dimethylcyclohexanone as an “Intramolecular Trap”

5.2. C(CO)−C(α) Bond Activation of Ketones

The C(CO)−C(α) bond activation of ketones was initially observed when acetophenone was reacted with IrIII(ttp)X (X = (CO)BF4, (CO)Cl, Me) to produce IrIII(ttp)COPh.41 However, the C−C bond cleavage was nonselective for aliphatic ketones, which produced a mixture of IrIII(ttp)(acyl) and IrIII(ttp)(alkyl).42 Hence, we turned our attention to rhodium porphyrin complexes because of their well-documented CCA reactivity. When RhIII(ttp)Me (14) was used, the selective C(CO)−C(α) bond activation of ketones was observed, and the acyl group transferred to the rhodium porphyrin to produce the corresponding RhIII(ttp)COR (15) (Table 8).43 The C−C bond cleavage of isopropyl methyl ketone occurred at the sterically more hindered but weaker C(CO)−C(iPr) bond to produce RhIII(ttp)COMe (Table 8, entry 5). Diisopropyl ketone was the most reactive substrate and yielded RhIII(ttp)COiPr and acetone as the organometallic and organic coproducts, respectively (Table 8, entry 6). The formation of acetone suggests that the incorporation of an extra oxygen atom originates from the RhIII(ttp)OH intermediate. Therefore, RhIII(ttp)Me (14) was allowed to react with 2,6-dimethylcyclohexanone as an intramolecular trap at 100 °C to give 85% yield of RhIII(ttp)(COCHMe(CH2)3COMe) (16) (Scheme 19). The rate-promoting effect of added H2O is consistent with the more rapid formation of RhIII(ttp)OH for CCA. The intermediacy of RhIII(ttp)OH is further supported when RhIII(ttp)CH2CH2OH, which is the precursor of RhIII(ttp)OH by β-hydroxyl elimination, was used as a starting material and produced the identical RhIII(ttp)COR products.44 Hence, the selective C(CO)−C(α) bond activation of ketones with RhIII(ttp)OH via σ-bond metathesis was proposed (Scheme 20). The exclusive formation of RhIII(ttp)COR (15) may be an outcome of the kinetic control by the attack of rhodium porphyrin to the sterically more accessible

Scheme 20. Proposed Mechanism for the C(CO)−C(α) Bond Activation of Ketones

sp2-hybridized carbonyl carbon instead of the sp3-hybridized αcarbon (Scheme 21). Then, we successfully demonstrated the mild temperature, visible-light photocatalytic C(CO)−C(α) bond anaerobic oxidation of ketone with H2O by RhIII(ttp)Me (14) and RhIII(ttp)COR (15). For example, 2,6-dimethylcyclohexanone was catalytically converted to 2-heptanone (Scheme 22).45 CO and H2 were formed as coproducts from the acyl radical 1709

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research Scheme 21. Proposed Transition State for the σ-bond Metathesis of Ketones with RhIII(ttp)OH

Author Contributions

The manuscript was written with the contributions of all authors. All authors have given approval to the final version of the manuscript. Funding

We thank the GRF No. 400212 of Hong Kong for the financial support. Notes

The authors declare no competing financial interest. Biographies

Scheme 22. Photocatalytic Oxidation of C−C Bond in Ketones

Ching Tat To received his B.Sc. degree (2010) and Ph.D. degree (2015) in chemistry from The Chinese University of Hong Kong under the supervision of Professor Kin Shing Chan. His work focused on catalytic carbon−carbon σ-bond hydrogenation using water. Currently, he is working as a postdoctoral fellow on the team of Professor Zhaomin Hou at RIKEN. Kin Shing Chan received his B.Sc. degree (1981) from the University of Hong Kong and his Ph.D. degree (1986) from the University of Chicago while working under the supervision of Professor William D. Wulff. After postdoctoral training (1986−1989) under Professor Jack Halpern at the University of Chicago, he accepted an associate professorship (1989−1990) at the National Taiwan University. Then, he moved to the Chinese University of Hong Kong (1990−now) and is currently a professor in chemistry.

decomposition and alcohol dehydrogenation. A labeling experiment using H218O produced 18O-enriched 2-heptanone. The photolytic and thermal 16O/18O exchange between H218O and 2-heptanone was not observed. Therefore, the C(CO)− C(α) bond was anaerobically oxidized by H2O through an RhIII(ttp)OH intermediate.

■

6. CONCLUSIONS Over the years, this laboratory has developed several protocols for the intermolecular and selective cleavage of aliphatic C−C bonds in both coordinating and noncoordinating organic substrates with rhodium porphyrin complexes. Two active intermediates, rhodium-(II) porphyrin metalloradical and rhodium(III) porphyrin hydroxide, effectively activate the aliphatic C−C bonds via homolytic radical substitution or σbond metathesis, respectively. Moreover, these reaction systems can bypass the C−H activation process or likely degenerate in nature. Based on these stoichiometric reactions, catalytic CCA using rhodium porphyrin complexes has been developed. The use of H2O as a reductant or oxidant for the C−C σ-bond under neutral conditions is possible using well-designed reaction systems. These catalysts remain in the early stage of development and therefore suffer from limited scope of substrates. Although we have disclosed the majority of CCA chemistry by rhodium porphyrin complexes, studies on other group-9 metalloporphyrin complexes are in progress in this laboratory. Recently, there have been notable advances on both stoichiometric and catalytic CCA using iridium and cobalt porphyrin complexes.46−48 Using easy modification of the electronic and steric properties of porphyrin ligands, we envision that the catalytic efficiency can be improved in the near future. The development of more practical catalytic C−C bond transformations, particularly using H2O, may provide a prominent direction for the greener hydrogenation of small carbon rings.

■

ACKNOWLEDGMENTS The authors thank the past and present Chan group members, whose work contributed to the knowledge and results in this Account, and their names have appeared in the cited papers.

■

REFERENCES

(1) Rybtchinski, B.; Milstein, D. Metal Insertion into C-C Bonds in Solution. Angew. Chem., Int. Ed. 1999, 38, 870−883. (2) Raseev, S. Thermal and Catalytic Processes in Petroleum Refining; Marcel Dekker, Inc., New York, 2003. (3) Tipper, C. F. H. Some Reactions of cyclopropane, and a Comparison with the Lower Olefins. Part II. Some PlatinouscycloPropane Complexes. J. Chem. Soc. 1955, 2045−2046. (4) Chen, F.; Wang, T.; Jiao, N. Recent Advances in TransitionMetal-Catalyzed Functionalization of Unstrained Carbon-Carbon Bonds. Chem. Rev. 2014, 114, 8613−8661. (5) Gozin, M.; Weisman, A.; Ben-David, Y.; Milstein, D. Activation of a Carbon-Carbon Bond in Solution by Transition-Metal Insertion. Nature 1993, 364, 699−701. (6) Jun, C.-H. Transition Metal-Catalyzed Carbon-Carbon Bond Activation. Chem. Soc. Rev. 2004, 33, 610−618. (7) Murakami, M.; Ito, Y. Cleavage of Carbon-Carbon Single Bonds by Transition Metals. Top. Organomet. Chem. 1999, 3, 97−129. (8) Sherry, A. E.; Wayland, B. B. Metalloradical Activation of CO: Formation and Carbonyl Coupling of a Bent 17-Electron M-CO Unit. J. Am. Chem. Soc. 1989, 111, 5010−5012. (9) Setsune, J.-I.; Yoshida, Z.-I.; Ogoshi, H. Syntheses and Reactions of Hydridorhodium(III) Octaethylporphyrin and Rhodium(II) Octaethylporphyrin Dimer. J. Chem. Soc., Perkin Trans. 1 1982, 983−987. (10) Wayland, B. B.; Coffin, V. L.; Farnos, M. D. Estimation of the Rh-Rh Bond Dissociation Energy in the (Octaethylporphyrinato)rhodium(II) Dimer by 1H NMR Line Broadening. Inorg. Chem. 1988, 27, 2745−2747. (11) Wayland, B. B.; Ba, S.; Sherry, A. E. Activation of Methane and Toluene by Rhodium(II) Porphyrin Complexes. J. Am. Chem. Soc. 1991, 113, 5305−5311.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Kin Shing Chan: 0000-0002-1112-6819 1710

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711

Article

Accounts of Chemical Research (12) Choi, K. S.; Lai, T. H.; Lee, S. Y.; Chan, K. S. Reduction of Rhodium(III) Porphyrin Hydroxide to Rhodium(II) Porphyrin. Organometallics 2011, 30, 2633−2635. (13) To, C. T.; Choi, K. S.; Chan, K. S. Catalytic Carbon-Carbon σBond Hydrogenation with Water Catalyzed by Rhodium Porphyrins. J. Am. Chem. Soc. 2012, 134, 11388−11391. (14) Wayland, B. B.; Sherry, A. E.; Bunn, A. G. EPR Studies of 1:1 Complexes of Rhodium(II) and Cobalt(II) Porphyrins with σ Donor and π Acceptor Ligands: Origins of Rhodium(II) Metalloradical Reactivity. J. Am. Chem. Soc. 1993, 115, 7675−7684. (15) Wayland, B. B.; Balkus, K. J.; Farnos, M. D. Organometallic Reactions of Rhodium Octaethylporphyrin Species in Pyridine: Heterolytic Cleavage of [(OEP)Rh]2 and Metalloanion Activation of CO. Organometallics 1989, 8, 950−955. (16) Fu, X.; Wayland, B. B. Equilibrium Thermodynamic Studies in Water: Reactions of Dihydrogen with Rhodium(III) Porphyrins Relevant to Rh-Rh, Rh-H, and Rh-OH Bond Energetics. J. Am. Chem. Soc. 2004, 126, 2623−2631. (17) Mak, K. W.; Chan, K. S. Novel 1,2-Rearrangement of Porphyrinatorhodium(III) Alkyls: Cis β-Hydride Elimination/Olefin Metal-Hydride Insertion Pathway. J. Am. Chem. Soc. 1998, 120, 9686− 8687. (18) Mak, K. W.; Yeung, S. K.; Chan, K. S. Nonradical Trapping Pathway for Reactions of Nitroxides with Rhodium Porphyrin Alkyls Bearing β-Hydrogens and Subsequent Carbon-Carbon Bond Activation. Organometallics 2002, 21, 2362−2364. (19) Tse, M. K.; Chan, K. S. Activation of Unstrained Aliphatic Carbon-Carbon Bonds by a Transition Metal Complex. J. Chem. Soc., Dalton Trans. 2001, 510−511. (20) Oayama, Y.; Tanaka, Y.; Yoshida, Y.; Toi, H.; Ogoshi, H. Catalytic Reactions of Metalloporphyrins.: II. Activation and Catalytic Aldol Condensation of Ketone with Rhodium(III)-Porphyrin Perchlorate under Neutral and Mild Conditions. J. Organomet. Chem. 1987, 329, 251−266. (21) Zhang, L.; Chan, K. S. Aliphatic Carbon-Carbon Bond Activation of Ketones by Rhodium(II) Porphyrin Radical. J. Organomet. Chem. 2006, 691, 3782−3787. (22) Zhang, L.; Chan, K. S. Activation of Aliphatic Carbon-Carbon Bonds of Esters and Amides by Rhodium(II) Porphyrin. J. Organomet. Chem. 2007, 692, 2021−2027. (23) Johnson, M. D. Bimolecular Homolytic Displacement of Transition-Metal Complexes from Carbon. Acc. Chem. Res. 1983, 16, 343−349. (24) Schiesser, C. H.; Wild, L. M. Free-Radical Homolytic Substitution: New Methods for Formation of Bonds to Heteroatoms. Tetrahedron 1996, 52, 13265−13314. (25) Tobisu, M.; Chatani, N. Catalytic Reactions Involving the Cleavage of Carbon-Cyano and Carbon-Carbon Triple Bonds. Chem. Soc. Rev. 2008, 37, 300−307. (26) Chan, K. S.; Li, X. Z.; Zhang, L.; Fung, C. W. Metalloradical Activations of Aliphatic Carbon-Carbon Bonds of Nitriles: Scope and Mechanism. Organometallics 2007, 26, 2679−2687. (27) Luo, Y. R. Comprehensive Handbook of Chemical Bond Energies; CRC Press, Boca Raton, FL, 2007. (28) Collman, J. P.; Boulatov, R. Synthesis and Reactivity of Porphyrinatorhodium(II)-Triethylphosphine Adducts: The Role of PEt3 in Stabilizing a Formal Rh(II) State. J. Am. Chem. Soc. 2000, 122, 11812−11821. (29) Chan, K. S.; Li, X. Z.; Dzik, W. I.; de Bruin, B. Carbon-Carbon Bond Activation of 2,2,6,6-Tetramethyl-piperidine-1-oxyl by a RhII Metalloradical: A Combined Experiment and Theoretical Study. J. Am. Chem. Soc. 2008, 130, 2051−2061. (30) Chan, Y. W.; Chan, K. S. Base-Promoted Carbon-Hydrogen Bond Activation of Alkanes with Rhodium(III) Porphyrin Complexes. Organometallics 2008, 27, 4625−4635. (31) Chan, Y. W.; Chan, K. S. Metalloradical-Catalyzed Aliphatic Carbon-Carbon Activation of Cyclooctane. J. Am. Chem. Soc. 2010, 132, 6920−6922.

(32) Chan, Y. W.; de Bruin, B.; Chan, K. S. Metalloradical-Catalyzed Selective 1,2-Rh-H Insertion into the Aliphatic Carbon-Carbon Bond of Cyclooctane. Organometallics 2015, 34, 2849−2857. (33) Paonessa, R. S.; Thomas, N. C.; Halpern, J. Insertion and Oxidative Addition Reactions of Rhodium Porphyrin Complexes. Novel Free Radical Chain Mechanisms. J. Am. Chem. Soc. 1985, 107, 4333−4335. (34) de Meijere, A.; Kozhushkov, S. I.; Rauch, K.; Schill, H.; Verevkin, S. P.; Kümmerlin, M.; Beckhaus, H.-D.; Rüchardt, C.; Yufit, D. S. Heats of Formation of [2.2]Paracyclophane-1-ene and [2.2]Paracyclophane-1,9-diene − An Experimental Study. J. Am. Chem. Soc. 2003, 125, 15110−15113. (35) Roth, W. R.; Hopf, H.; de Meijere, A.; Hunold, F.; Börner, S.; Neumann, M.; Wasser, T.; Szurowski, J.; Mlynek, C. Zur Rekombinationsenthalpie von Radikalen − Thermochemie der Ringöffnung von Cyclophanen. Liebigs Ann. 1996, 2141−2154. (36) Lai, T. H.; Chan, K. S. Base-Promoted, Selective Aliphatic Carbon-Carbon Bond Cleavage of Ethers by Rhodium(III) Porphyrin Complexes. Organometallics 2009, 28, 6845−6846. (37) Lee, S. Y.; Lai, T. H.; Choi, K. S.; Chan, K. S. RoomTemperature Selective Aliphatic Carbon-Carbon Bond Activation and Functionalization of Ethers by Rhodium(II) Porphyrin. Organometallics 2011, 30, 3691−3693. (38) Lee, S. Y.; Lai, T. H.; Hui, Y. Y.; Chan, K. S. HydroxidePromoted Selective C(α)-C(β) Bond Activation of Aliphatic Ethers by Rhodium(III) Porphyrins. J. Organomet. Chem. 2014, 762, 88−93. (39) Choi, K. S.; Chiu, P. F.; Chan, K. S. Selective Activation of Benzylic Carbon-Hydrogen Bonds of Toluenes with Rhodium(III) Porphyrin Methyl: Scope and Mechanism. Organometallics 2010, 29, 624−629. (40) Brand, H.; Arnold, J. Recent Developments in the Chemistry of Early Transition Metal Porphyrin Compounds. Coord. Chem. Rev. 1995, 140, 137−168. (41) Li, B. Z.; Song, X.; Fung, H. S.; Chan, K. S. Cleavage of Carbonyl Carbon and α-Carbon Bond of Acetophenones by Iridium(III) Porphyrin Complexes. Organometallics 2010, 29, 2001− 2003. (42) Li, B. Z.; Fung, H. S.; Song, X.; Chan, K. S. Scope and Mechanism of Carbonyl Carbon and α-Carbon Bond Cleavage of Ketones by Iridium(III) Porphyrin Complexes. Organometallics 2011, 30, 1984−1990. (43) Fung, H. S.; Li, B. Z.; Chan, K. S. Mild and Selective C(CO)C(α) Bond Cleavage of Ketones by Rhodium(III) Porphyrins: Scope and Mechanism. Organometallics 2012, 31, 570−579. (44) Chan, C. S.; Lee, S. Y.; Chan, K. S. Mild and Selective C(CO)C(α) Bond Activation of Ketones with Rhodium(III) Porphyrin βHydroxyethyl. Organometallics 2013, 32, 151−156. (45) Lee, S. Y.; Chan, K. S. Photocatalytic Carbon-Carbon σ-Bond Anaerobic Oxidation of Ketones with Water by Rhodium(III) Porphyrins. Organometallics 2013, 32, 5391−5401. (46) To, C. T.; Tam, C. M.; Chan, K. S. Iridium-Catalyzed CarbonCarbon σ-Bond Hydrogenation with Water: Rate Enhancement with Iridium Hydride. ACS Catal. 2015, 5, 4333−4336. (47) Tam, C. M.; To, C. T.; Chan, K. S. Carbon-Carbon σ-Bond Transfer Hydrogenation with DMF Catalyzed by Cobalt Porphyrins. Organometallics 2016, 35, 2174−2177. (48) Lee, S. Y.; Fung, H. S.; Feng, S.; Chan, K. S. Visible Light Photocatalysis of Carbon-Carbon σ-Bond Anaerobic Oxidation of Ketones with Water by Cobalt(II) Porphyrins. Organometallics 2016, 35, 2480−2487.

1711

DOI: 10.1021/acs.accounts.7b00150 Acc. Chem. Res. 2017, 50, 1702−1711