New highly brominated Mn-porphyrin: a good catalyst for activation of inert C–H bonds

Abstract

This work describes the synthesis and characterization of a novel third-generation catalyst 5,10,15,20-tetrakis-(4′-bromine-3′,5′-dimethoxyphenyl)-2,3,7,8,12,13,17,18-octabromoporphyrinatomanganese chloride [Mn^IIIBr₁₂T3,5DMPP]Cl (Cat.2). The catalytic activity of Cat.2 in cyclohexane, adamantane, and n-hexane oxidation by iodosylbenzene (PhIO) or iodobenzene diacetate (PhI(OAc)₂) was compared with the catalytic activity of [Mn^IIIT3,5DMPP]Cl (Cat.1), a second generation catalyst. The Cat.2/PhI(OAc)₂ system led to higher yields of cyclohexane oxidation products (65%) with high selectivity for cyclohexanol (86%) as compared with Cat.1 (19% and 74%, respectively) and addition of water essentially did not alter total product yield. Addition of a small amount of imidazole to the Cat.1/PhIO system gave superior yields of cyclohexane oxidation products (64%) as compared with Cat.2 (52%). In all systems Cat.2 afforded significantly higher yields of 2-adamantanol, a product with great commercial value compared with 1-adamantanol. n-Hexane oxidation gave low total product yield; Cat.2 was more selective for alcohol products (2-hexanol and 3-hexanol).

---

Introduction

Manganese and iron porphyrins (MnPs and FePs, respectively) have been investigated mainly as structural and functional models of cytochrome P450-dependent monooxygenases.^1–6 These metalloporphyrins (MP) are efficient and selective catalysts for alkane and alkene oxidation. The first investigations into these complexes focused on alkane hydroxylation and alkene epoxidation by simple metalloporphyrins such as tetraphenylporphyrinatoiron(III) chloride ([Fe^IIITPP]Cl)⁷ and (tetramesitylporphyrinato)-manganese(III) chloride ([Mn^IIITMP]Cl).⁸ Oxidative degradation rapidly deactivated these catalysts, which has motivated researchers to develop metalloporphyrins bearing aryl groups with substituents like nitro (–NO₂),^9,10 chloro (–Cl),¹¹ fluoro (–F),¹² methoxy (–OCH₃),¹⁰ carbomethoxy (–COOCH₃),^13,14 and amino (–NH₂),^15,16 among others, in the meso positions of the porphyrin macrocycle. Later, researchers verified that metalloporphyrins with bulky electron withdrawing halogen substituents (e.g., –Cl and –Br)¹⁷ in the β-pyrrolic positions of the porphyrin ring modified the reduction potential of the metal center^16,18 and the macrocycle conformation (which acquired a saddle-shaped structure instead of a planar structure),¹⁹ affording more robust complexes. Meunier classified porphyrins and metalloporphyrins as first-, second-, and third-generation catalysts depending on the nature of the substituent and its position in the macrocycle.²⁰ Second-(substituents in the aryl group) and third-(substituents in the β-pyrrole positions) generation metalloporphyrins proved to be more efficient and stable catalysts for organic substrate oxidation than their first-generation counterparts.^21,22

New strategies to improve metalloporphyrin catalyst efficiency in organic substrate oxidation emerged after elucidation of the structure of cytochrome P450 – the prosthetic group heme, which consists of iron protoporphyrin IX, is axially coordinated to a sulfur atom of residual amino acids and contains a free site for later molecular oxygen activation.²³ This axial ligand is believed to affect the reactivity of the prosthetic group directly.^24,25 Therefore, countless studies have developed biomimetic models based on FePs and MnPs using pyridine or imidazole ligands as additives.^2,26 These ligands can act as Lewis bases and increase the metalloporphyrin catalytic efficiency in the system. Indeed, they prevent metal ion displacement from the porphyrin macrocycle (which could decrease superposition between the orbitals of the metal ion and the porphyrin), thereby avoiding macrocycle oxidation. These ligands also donate electrons to manganese in the Mn-oxo species via σ orbitals, which weakens the Mn=O bond and renders the oxygen atom more nucleophilic.²⁷

Despite all the efforts made by the scientific community, only in 2009 were iron and cobalt porphyrins tested as catalysts for cyclohexane oxidation on an industrial scale.²⁸ In this same line, in 2012 Liu and Guo conducted a theoretical study on the viability of implementing the oxidation of little reactive bonds (C–H) by molecular oxygen catalyzed by complexes that mimic cytochrome P450 in the industry.²⁹ Therefore, alkane oxidation is a potential area in chemistry and has stimulated new investigations, with great perspectives.

In this context, this work has focused on organic substrate oxidation reactions catalyzed by metalloporphyrins (MP) that are more stable in the reaction medium and allow for higher reaction rate, selectivity, and stereospecificity with respect to value-added products. To this end, the metalloporphyrins [Mn^IIIT3,5DMPP]Cl (Cat.1) and [Mn^IIIBr₁₂T3,5DMPP]Cl (Cat.2, novel), Fig. 1, were used as catalysts for cyclohexane, adamantane, and n-hexane oxidation by PhIO (iodosylbenzene) or PhI(OAc)₂ (iodobenzene diacetate). The influence of additives like imidazole and water on cyclohexane and adamantane oxidation was also evaluated.

Fig. 1 Structural representation of the metalloporphyrin catalysts synthesized in this work.

Results and discussion

A. Synthesis

The porphyrin macrocycle was obtained according to the method proposed by Gonsalves et al.³⁰ This method employs nitrobenzene to avoid the formation of chlorine, acting as oxidant. The porphyrin emerges in a single step and is easier to purify. Addition of methanol favors porphyrin precipitation and separation from the reaction mixture. Polypyrrole, which is difficult to separate from the porphyrin by column chromatography, remains in the reaction mixture. ¹H NMR helped to elucidate the structure of the porphyrin macrocycle (Fig. S2†). The chloroform/methanol method³¹ yielded the second-generation MnP (Cat.1). This method involves solubilization of the free-base porphyrin in CHCl₃ and manganese salt in methanol. The basic nature of the acetate anion from manganese(II) acetate tetrahydrate facilitates deprotonation of the pyrrole nitrogens in the porphyrin macrocycle, thereby favoring porphyrin metallation.

Cat.1 was characterized by thin layer chromatography with chloroform as solvent. Cat.1 remained close to the base line (R_f = 0.0032), whilst the corresponding free-base porphyrin moved along the chromatographic plate (R_f = 0.44). The presence of a heavy atom (manganese) in Cat.1 led to loss of the red fluorescence typical of free-base porphyrins upon irradiation with UV light.³²

UV-vis spectroscopy (Fig. 2) evidenced a bathochromic shift of the Soret band of Cat.1 as compared with the Soret band of the free-base porphyrin. The UV-vis spectrum of Cat.1 also presented fewer bands and resembled the spectra of other type-d hyperporphyrins: it displayed the ligand–metal charge transfer (LMCT) band at 382 nm.³³ This happened because the e_g (d_xz, d_yz) orbital of Mn(III) in a complex with D_4h symmetry has adequate energy and symmetry to interact with the e^*_g (π) orbital of the porphyrin macrocycle.³⁴

Fig. 2 UV-vis absorption spectra of H₂T3,5DMPP (3.02 × 10⁻⁶ mol L⁻¹), [Mn^IIIT3,5DMPP]Cl (Cat.1, 8.59 × 10⁻⁶ mol L⁻¹) and [Mn^IIIBr₁₂T3,5DMPP]Cl (Cat.2, 1.68 × 10⁻⁵ mol L⁻¹) in CH₂Cl₂.

The novel third-generation metalloporphyrin (Cat.2) was synthesized by adaptation of the method described by Richards et al. (Fig. 1). In this method, DMF (as solvent) and excess bromine favor complete polybromination of the porphyrin macrocycle.³⁵ Considering that metalloporphyrin macrocycles are highly conjugated aromatic systems, their β-pyrrolic positions are very reactive and therefore more susceptible to aromatic electrophilic substitution than the ortho-, meta-, and para-aryl substituents in the meso position of the porphyrin ring, especially in the case of aryl groups that do not bear activating groups. As for porphyrins with electron donating groups in the ortho-, meta-, and para-mesoaryl positions of the macrocycle, halogenation reactions introduce halogens into the ortho-, meta-, and para-positions relative to the electron donating groups.^16,18,36,37 Here, the activating character of the methoxy (–OCH₃) groups promoted bromine addition in the ortho-position relative to –OCH₃, apart from bromine addition in the β-pyrrole positions of the porphyrin ring. Sankar et al. also observed introduction of 15 or 16 bromine atoms into Zn^IIT3,5DMPP (considering the β-pyrrole positions and the mesoaryl groups).³⁷ In the present work, bromination was monitored by UV-vis spectroscopy (Fig. 2). Compared with Cat.1, there was a bathochromic shift in the Soret band of Cat.2. This observation agreed with reports published for other β-octabrominated MnPs and their non-brominated counterparts.^{13,14,16,19,38,39} The bulky electron withdrawing bromine substituents in the β-pyrrole positions of Cat.2 caused the bathochromic shift. These bromine atoms distort the porphyrin macrocycle, destabilize the HOMO orbitals, and decrease the porphyrin π orbitals conjugation, at the same time that the HOMO and LUMO orbitals stabilize via an inductive effect. The final energy balance reduces the HOMO–LUMO energy difference and prompts a bathochromic shift of the Soret band.^13,18,40,41

Metalloporphyrins bearing bromine atoms in the β-pyrrole positions favor lower metal ion oxidation states (e.g., Mn(II)) as a consequence of the electron withdrawing effect of bromine.^12,42–45 To verify whether Mn(II) existed in Cat.2, solid-state Cat.2 and Cat.2 in DMF solution were submitted to Electron Paramagnetic Resonance (EPR) at 77 K (Fig. 3, frequency X, band 9.5 GHz) by perpendicular polarization with microwave X-band. EPR signals were absent in the spectrum of solid-sate Cat.2, suggesting that Cat.2 only contained Mn(III), which is EPR silent in the measurement conditions.^42,46 Indeed, the Mn(III) ion is paramagnetic and has four unpaired d electrons (S = 2); it typically exhibits pronounced Jahn–Teller distortion, which results in substantial spin–orbit coupling.⁴⁷ The final outcome of this coupling and rapid Mn(III) spin relaxation processes⁴⁸ is the absence of signals in conventional Mn(III) EPR spectra conducted in the X-band (microwave field (H1) perpendicular to the static magnetic field (H0)) even at 77 K,⁴⁷ even though spin transitions may arise during measurements in the parallel mode (H1 ∥ H0).⁴⁹ Already in the spectrum of Cat.2 in DMF solution, a signal appeared in g = 2.03, which may be associated with Mn(II).^50–53 The hyperfine lines were not well defined, which prevented determination of the value of parameter A. The coordinating character of DMF could explain the presence of Mn(II) in the case of Cat.2 in DMF solution, since this coordination could favor Mn(III) reduction to Mn(II).^14,44 In fact, the UV-vis spectrum of Cat.2 in DMF was typical of Mn(II)Ps (Fig. S3†).

Fig. 3 EPR spectra of Cat.2 in (a) the solid state and (b) DMF solution.

FTIR results confirmed formation of the free-base porphyrin, Cat.1, and Cat.2 (Fig. S4†). All the spectra displayed the bands attributed to ν C–O–C (1249 cm⁻¹), ν OCH₃ (1173 cm⁻¹), and δ N–H (pyrrole) (966 cm⁻¹). The band due to N–H (pyrrole) appeared in the spectrum of H₂T3,5DMPP.

The spectra of Cat.1 and Cat.2 exhibited the band relative to ν Mn–N (pyrrole) (1009 cm⁻¹).⁵⁴

The mass spectrum obtained for Cat.1 by electrospray ionization (ESI) in the positive mode revealed a peak at m/z 907.35, with 100% relative intensity. Chloride anion loss by [Mn^IIIT3,5DMPP]⁺ accounts for this peak. As for Cat.2, the mass spectrum (Fig. S5†) showed that this metalloporphyrin presented 12 bromine atoms. The peak at m/z 1886.02 (relative intensity = 100%) is probably associated with [Mn^IIIBr₁₂T3,5DMPP + CH₃OH]⁺, which originates from loss of one chloride by Cat.2 and addition of methanol, the solvent, to the metalloporphyrin.

The preparation of Cat.2 was accomplished via a two-step procedure starting from the H₂T3,5DMPP (a symmetric porphyrin). This procedure is environmentally and economically more viable than that used for obtain non-symmetrical metalloporphyrin catalysts¹⁶ because in the latter MPs, more steps are necessary to obtain the free base porphyrin (precursor).

B. Catalytic studies

Investigation into the oxidation of poorly reactive substrates such as cyclohexane, adamantane, and n-hexane (Scheme 1) enables comparison of catalyst efficiency during oxidation of linear and cyclic alkanes.^2,3,55,56 This work used two distinct oxygen donors in the MnP-catalyzed alkane oxidation reactions: PhIO and PhI(OAc)₂.^16,57 It also examined how addition of water and imidazole affected cyclohexane and adamantane oxidation.

Scheme 1 Structural representation of n-hexane, cyclohexane, and adamantane oxidation reactions.

B.1. Cyclohexane oxidation. Metalloporphyrin-catalyzed cyclohexane oxidation usually yields cyclohexanol (Cy-ol) and cyclohexanone (Cy-one) (Scheme 1). In most cases, Cy-ol is the major product. [Mn^IIITPP]Cl,⁷ a first-generation metalloporphyrin, is the classic reference in studies on the catalytic activity of MnPs.²⁰

In the presence of PhIO as oxidant (Fig. 4), Cat.1 and [Mn^IIITPP]Cl gave practically the same total product yield (Cy-ol yield = 14%).¹⁶ Compared with the MnP bearing the carbomethoxy group (–COOCH₃) in the para-meso aryl positions, [Mn^IIITCMPP]Cl,¹³ the Cat.1/PhIO system was less efficient. This might have resulted from the electron withdrawing character of the –COOCH₃ groups as compared with the electron donating character of the –OCH₃ groups.⁵⁸ Cat.1 was less efficient than a second-generation MnP bearing electron donating groups like –NH₂ in the mesoaryl positions. In the latter case, the substituents containing nitrogen atoms may coordinate with the metal ion of another MnP complex and donate electrons to the manganese atom in the Mn-oxo species via σ orbitals.^15,16,59 This donation weakens the Mn=O bond and renders the oxygen more nucleophilic.

Fig. 4 Cyclohexane oxidation by PhIO or PhI(OAc)₂ catalyzed by MnP in CH₂Cl₂ under aerobic condition: cyclohexane (Cy-ol) and cyclohexanone (Cy-one) yields, selectivity for Cy-ol, and degree of catalyst destruction. Reactions in the absence of catalyst did not yield any significant amount of the products. Reaction conditions: [MnP] = 5 × 10⁻⁴ mol L⁻¹, [oxidant] = 5 × 10⁻³ mol L⁻¹, MnP/oxidant/cyclohexane/CH₂Cl₂ molar ratio = 1 : 10 : 4650 : 15 550, 25 °C, magnetic stirring.

This work also evaluated the efficiency of the Cat.1/PhI(OAc)₂ and Cat.2/PhI(OAc)₂ systems. PhI(OAc)₂ is an alternative oxidant to PhIO. Although PhIO gives rise to the catalytically active species, it has some disadvantages: (1) it is poorly soluble in most of the organic solvents, (2) it is potentially explosive, and (3) it undergoes slow but progressive disproportionation.⁶⁰ Moreover, few works have used PhI(OAc)₂ as oxygen donor during alkane oxidation catalyzed by MnPs.^15,16,38,61 Therefore, the use of this oxidant might contribute to investigations into other reaction mechanisms.⁶²

The Cat.1/PhI(OAc)₂ system afforded total product yield similar to those obtained with the Cat.1/PhIO system, evidencing that PhI(OAc)₂ can substitute PhIO, as reported in other works published by our group.^16,63

In the presence of PhIO or PhI(OAc)₂, the third-generation Cat.2 was much more efficient than its second-generation counterpart, Cat.1. The higher efficiency of Cat.2 systems derived from introduction of bromine atoms into the β-pyrrole positions of the macrocycle, which withdraw electron density from the metal center and destabilize the high-valent active species.^21,42,64 Nevertheless, Cat.2 underwent more extensive destruction as compared with Cat.1, probably because the structure of Cat.2 is more distorted due to the presence of bulky bromine atoms in the β-pyrrole positions of the porphyrin ring. Indeed, metalloporphyrins containing bromine atoms in these positions acquire a saddle-shaped conformation.⁶⁵

Cat.2 gave higher Cy-ol yields and selectivity than other third-generation catalysts.^14,16,38,42 Cy-one may originate from Cy-ol oxidation.⁶⁶ Cat.2 bears substituents in the meta-(–OCH₃) and para-(–Br) mesoaryl positions of the porphyrin macrocycles, as well as bromine atoms in the β-pyrrole positions. This should make the approach between Cy-ol and the metal center difficult, which would justify the lower Cy-one yield and the higher selectivity for the alcohol in the case of Cat.2.

PhI(OAc)₂ as oxidant provided higher total product yield as compared with PhIO, which contrasted with the results published by In et al.⁶⁷ This behavior had already emerged in other investigations conducted by our group.⁶³ These contrasting results may be associated with the distinct active species generated during the oxidation reactions when different oxygen donors are used to oxidize the same substrate, as proposed in a kinetic study reported by Collman et al.⁶²

Addition of imidazole (Im) at a MnP/Im ratio of only 1 : 0.5 to the systems based on Cat.1 elicited over 200% increase in total product yield. Catalyst destruction declined, in accordance with literature results.^12,16 Imidazole addition may have prevented metal ion displacement from the porphyrin macrocycle plane and destabilized the high-valent active species, making it more nucleophilic.²⁷ As for the Cat.2/PhIO system, imidazole addition did not alter total product yield significantly, but it diminished catalyst destruction drastically, as reported in the literature.⁶³ On the other hand, imidazole addition to the Cat.2/PhI(OAc)₂ system lowered the catalytic activity as compared with the system without added imidazole. In this case, the bromine atoms and the distorted structure of Cat.2 may have hindered the interaction between the imidazole-coordinated MnP and the substrate/oxidant complex, accounting for the lower total product yield.⁶⁸ It is worth highlighting that second-generation catalysts like Cat.1, which are easier to obtain than third-generation catalysts such as Cat.2, display high catalytic activity in the presence of small amounts of additives like imidazole because the planar structure of second-generation metalloporphyrins favor interaction with this ligand.

Considering that water can act as a Lewis base, addition of water to systems using metalloporphyrins as catalysts could also improve catalyst efficiency. However, no consensus exists on how water acts during metalloporphyrin-catalyzed biomimetic oxidation. A theoretical study by Balcells et al. showed that water may act as an axial ligand.⁶⁹ On the other hand, In et al.,⁶⁷ Kwong et al.⁷⁰ and Chen et al.⁷¹ reported that water hydrolyzes PhI(OAc)₂ to generate PhIO in situ.⁶⁷ Da Silva et al.^15,16 verified that water enhances catalyst efficiency during cyclohexane oxidation by PhIO and PhI(OAc)₂.⁶⁷ Nevertheless, this same research team did not detect any rise in total product yield upon addition of water to systems that employ PhIO as oxidant.¹⁴ Hence, we decided to evaluate how water addition affected our systems.

Water (0.5 μL) decreased total product yield slightly in the case of the Cat.2/PhI(OAc)₂ and Cat.2/PhIO systems (from 56 to 53% and from 48 to 46%, respectively), as reported by da Silva et al.¹⁴ Methoxy groups may establish intermolecular interactions with water and prevent water molecules from approaching the metal center in the MnP.

The use of Cat.2 in the cyclohexane oxidation by PhIO or PhI(OAc)₂, allows to obtain cyclohexanol in high yield (48 and 56%) and selectivity (96 and 86%), respectively, without additives (water or imidazole) addition. These values are higher than those obtained when using the Mn^IIIBr₉APTPPCl¹⁶ catalyst. We report here that the addition of imidazole to Cat.1 systems causes an increase in yield and selectivity for cyclohexanol, a similar behavior was observed for Mn^IIIAPTPPCl.¹⁶ Moreover, the addition of a small amount of water (0.5 μL) barely affects the catalytic systems containing Cat.1 and Cat.2, probably because water interacts with methoxy (–OCH₃) substituents of the catalysts. This shows the important role of the substituents in the aryl group on the catalytic activity of metalloporphyrins.¹⁴

B.2. Adamantane oxidation. Adamantane is a cyclic alkane. Oxidation of its tertiary carbon gives 1-adamantanol (1-Adol), whereas oxidation of its secondary carbon affords 2-adamantanol (2-Adol) and 2-adamantanone (2-Adone). Here, regioselectivity for the adamantane tertiary carbon was normalized to consider the statistical probability of hydroxylating one of the four tertiary C–H bonds (which generates 1-Adol) versus one of the 12 secondary C–H bonds (which yields 2-Adol).

Groves et al. were the first to report on metalloporphyrin-catalyzed adamantane oxidation by PhIO.⁷ These authors employed [Fe^IIITPP]Cl, a first-generation metalloporphyrin,²⁰ as catalyst and observed preferential formation of 1-Adol; 2-Adone did not emerge during the reaction. Compared with [Fe^IIITPP]Cl, Cat.1 (Fig. 5) gave slightly higher total product yield than those verified by Groves et al. (12% 1-Adol and 1% 2-Adol).⁷ Cat.1 was also more efficient than the catalyst cis-[Mn^IIIDAPDPP]Cl proposed by Silva et al.⁷² Compared with Cat.1, the catalyst [Mn^IIIPFTDCPP]Cl obtained by Doro et al. was much more efficient (35% 1-Adol, 10% 2-Adol, and 5% 2-Adone).¹² In the latter MnP, the electron withdrawing substituents (fluoro and chloro) in the ortho-, meta-, and para-mesoaryl positions of the macrocycle must have rendered the high-valent active species more reactive. Also compared with Cat.1, the catalyst [Mn^IIIT2,6DMeOPP]Cl obtained by Baciocchi et al. furnished higher 1-Adol yield (30%).⁵⁸ These different results may be associated with the substrate concentration employed by the aforementioned authors. Indeed, Ucoski et al.⁷³ described that higher substrate concentrations afford higher total product yields.

Fig. 5 Adamantane oxidation by PhIO or PhI(OAc)₂ catalyzed by MnP in CH₂Cl₂ under aerobic condition: 1-adamantanol (1-Adol) and 2-adamantanol (2-Adol) yields, regioselectivity for 1-Adol, and degree of catalyst destruction. Reactions in the absence of catalyst did not yield any significant amount of the products. Reaction conditions: [MnP] = 5 × 10⁻⁴ mol L⁻¹, [oxidant] = 5 × 10⁻³ mol L⁻¹, MnP/oxidant/adamantane/CH₂Cl₂ molar ratio = 1 : 10 : 100 : 15 550, 25 °C, magnetic stirring.

Regarding Cat.1, regioselectivity for 1-Adol was higher than 95%, because the tertiary carbon is more reactive than the secondary carbon.^74–76 2-Adone arose in small amounts (∼0.5%) in all the studied MnP/oxidant systems.

When PhIO was the oxidant, the third-generation Cat.2 afforded slightly higher 1-Adol and 2-Adol yields and less catalyst degradation as compared with Cat.1. Therefore, introducing bromine atoms into the β-pyrrole positions of the porphyrin macrocycle had an important effect on adamantane oxidation. Although Cat.2 was less selective than Cat.1, higher 2-Adol yields did not constitute a drawback: its commercial value ($47.60 – Sigma-Aldrich/USA) is substantially higher than the commercial value of 1-Adol ($26.20 – Sigma-Aldrich/USA). The increased 2-Adol yield in reactions catalyzed by Cat.2 may have resulted from the steric hindrance posed by the bromine atoms in the β-pyrrole positions of the porphyrin ring, what allowed for substrate oxidation at preferential, albeit less thermodynamically favorable positions.^75,76 In a previous study conducted by our group, addition of small quantities of imidazole during metalloporphyrin-catalyzed organic substrate oxidation increased catalyst efficiency.⁶³

Here, addition of imidazole to the MnP/PhIO systems at a MnP/Im ratio of 1 : 0.5 augmented 1-Adol yield, especially in the case of Cat.1; catalyst destruction decreased. Recent papers have shown that imidazole addition to systems based on third-generation metalloporphyrins does not improve the catalytic results.^16,63 Nonetheless, when Doro et al. used the third-generation [Mn^IICl₈PFTDCPP] to catalyze adamantane oxidation by PhIO in the presence of imidazole, they achieved higher yield of hydroxylated products (61% 1-Adol and 14% 2-Adol) as compared with the Cat.2/PhIO system.¹² This difference may be related to the MnP/Im ratio (1 : 30) that these authors used as compared with the lower MnP/Im ratio employed here (1 : 0.5). Furthermore, [Mn^IICl₈PFTDCPP] bears electron withdrawing substituents (Cl and F) in the ortho-, meta-, and para-mesoaryl positions of the macrocycle, which probably make the active species of this MnP more reactive than the active species of Cat.2.

Baciocchi et al., who used [Mn^IIICl₈T2,6DMeOPP]Cl as catalyst during adamantane oxidation, reported a result similar to the result described by Doro et al.:¹² the total product yield was 75% for a MnP/Im ratio of 1 : 25.⁶⁸ Hence, the amount of imidazole used in the present work was not enough to ensure quantitative formation of the pentacoordinated MnP-Im species that would be necessary to raise the catalytic efficiency.

Addition of water at a MnP/H₂O molar ratio of 1 : 139 to the MnP/PhIO systems did not improve catalyst efficiency. Water must have interacted with the –OCH₃ groups in the MnP, being little available for coordination to the metal center.

MnP-catalyzed adamantane oxidation by PhI(OAc)₂ led to virtually the same total product yield as the MnP/PhIO systems, but selectivity for 1-Adol was higher. Considering the oxidants PhI(OAc)₂ and PhIO, Cat.2 was slightly more efficient than Cat.1, in contrast with the behavior observed during cyclohexane oxidation, where Cat.2 was much more efficient than Cat.1. Adamantane approach to the bulkier catalyst Cat.2 was probably more difficult, accounting for the catalytic results.⁷⁷

Addition of imidazole to the MnP/PhI(OAc)₂ systems raised the total product yield (>20%) for Cat.1 and Cat.2. Moreover, in the presence of imidazole, the MnP/PhI(OAc)₂ systems furnished slightly better total product yield than the MnP/PhIO systems. Li and Xia used [Mn^IIITPP]Cl and [Mn^IIITSPP] in the presence of ionic liquids and verified that addition of imidazole at a MnP/Im molar ratio of 1 : 50 increases the 1-Adol yield.⁶¹

Most importantly, addition of imidazole reduced catalyst destruction. Imidazole coordination to the metal center in the porphyrin macrocycle prevents metal ion displacement and catalyst oxidative destruction.

Addition of water to the MnP/PhI(OAc)₂ systems increased 1-Adol yields (63 and 72% total product yield for Cat.1 and Cat.2, respectively). As proposed by In et al., water may hydrolyze PhI(OAc)₂.⁶⁷ According to these authors, water accelerated cyclohexene epoxidation and yielded maximum alkene conversion within 15 min, whereas complete cyclohexene epoxidation took 5 h in the absence of water. Therefore, these authors proposed that PhI(OAc)₂ hydrolysis generated PhIO in situ.⁶⁷ To a lesser extent, these results may also have been associated with the coordination of water to the central metal in the MnPs, as proposed by Balcells et al.⁶⁹

B.3. n-Hexane oxidation. Linear alkane oxidation is a chemical transformation that has attracted a lot of interest from the scientific community – it is a highly complex process, and the alkane substrates are highly stable.^56,77 An important aspect to study during n-alkane oxidation is selectivity toward the oxidation of primary, secondary, and tertiary carbons, which present the following energy bonds: 435.4, 396.1, and 391.1 kJ mol⁻¹, respectively.⁷⁸ Hence, n-hexane oxidation reactions are little selective for 1-hexanol (1-ol), whereas 2-hexanol (2-ol) and 3-hexanol (3-ol) are the preferred products. Overall, MnP-catalyzed n-hexane oxidation affords low yields because the substrate is little reactive.

The third-generation Cat.2/PhIO system gave higher total product yield than the second-generation Cat.1/PhIO system (Fig. 6); the bromine atoms in the β-pyrrole positions of the macrocycle may account for this outcome.^42,68 Cat.2 was more selective for the alcohols than Cat.1 (∼81 and 48%, respectively). Compared with Cat.1, the second-generation cis-[Mn^IIIDAPDPP]Cl, which bears amino groups in the para-mesoaryl positions of the porphyrin ring, is more selective for the alcohols (∼71%).⁷² The amino groups in cis-[Mn^IIIDAPDPP]Cl may have coordinated with the metal center of another cis-[Mn^IIIDAPDPP]Cl, to render the high-valent active species Mn^V(O)P more reactive and to elevate the selectivity toward the alcohol.²⁷ Compared with the third-generation catalyst derived from cis-[Mn^IIIDAPDPP]Cl,⁷² Cat.2 was more selective (∼74 and ∼81%, respectively).

Fig. 6 Oxidation of n-hexane by PhIO or PhI(OAc)₂ catalyzed by MnP in CH₂Cl₂ under aerobic conditions: 1-hexanol (1-ol), 2-hexanol (2-ol), 3-hexanol (3-ol), 2-hexanone (2-one) and 3-hexanone (3-one) yields and degree of catalyst destruction. Reactions in the absence of catalyst did not yield any significant amount of the products. Reaction conditions: [MnP] = 5 × 10⁻⁴ mol L⁻¹, [oxidant] = 5 × 10⁻³ mol L⁻¹, MnP/oxidant/n-hexane/CH₂Cl₂ molar ratio = 1 : 10 : 3800 : 15 550, 25 °C, magnetic stirring.

The MnP/PhI(OAc)₂ and the MnP/PhIO systems gave similar total product yield and selectivity, but Cat.2 was always slightly more prone to oxidative destruction than Cat.1. Regarding the MnP/PhI(OAc)₂ systems, again Cat.2 was more efficient than Cat.1. Therefore, PhI(OAc)₂ can be used as an alternative to PhIO in these reactions.

Cat.2 furnished a higher amount of 2-ol as compared with 3-ol. Again, Cat.2 favored formation of a product with superior commercial value. This different behavior of Cat.2 as compared with Cat.1 and other catalysts described in the literature,⁷² may originate from the great steric hindrance provided by the bromine atoms in the β-pyrrole positions of the macrocycle and of the methoxy groups in the meta-mesoaryl positions of the porphyrin ring in Cat.2. Access of n-hexane to the active metal center should be restricted to the more exposed C–H bonds, giving rise to shape selectivity as proposed by Cook et al.⁷⁷

Experimental

A. Reagents

Analytical grade CH₃OH, CH₂Cl₂, CHCl₃ and pyrrole were obtained from Aldrich Chemical Co and distilled prior to use. PhIO was prepared according to a literature procedure,⁷⁹ stored at −20 °C in a freezer, and assayed periodically by iodometric titrations. All the other reagents and solvents were of analytical grade and were used without further purification, unless stated otherwise.

B. Equipment

UV-vis spectra (190–1100 nm) were recorded on an HP-8453A diode-array spectrophotometer. Infrared (IR) spectra were registered on a Perkin Elmer spectrometer model BXFTIR; the samples were prepared in KBr pellets. Room-temperature (25 °C) ¹H NMR spectra were obtained in CDCl₃ on a Bruker DPX-200 Advance spectrometer operating at 200 MHz; tetramethylsilane (TMS) was the internal standard. Gas chromatography was conducted on a Shimadzu GC-17A chromatograph equipped with a flame ionization detector and a Carbowax capillary column (measuring 30.0 m × 0.32 mm, with a film thickness of 0.25 μm). The ultrasound equipment Unique® MaxiClean 1400, 40 kHz, was also employed in the experiments. The ESI-MS analyses were conducted on an LCQFleet (ThermoScientific, San Jose, CA, USA) mass spectrometer equipped with an electrospray ionization (ESI) source operating in the positive ion mode; CH₃OH was used as solvent. Electron paramagnetic resonance (EPR) measurements of the powder materials were accomplished on an EPR BRUKER EMX microX spectrometer (frequency X, band 9.5 GHz), at room temperature and 77 K (using liquid N₂), by using perpendicular microwave polarization X-band. Elemental analysis (carbon, hydrogen and nitrogen) were carried out on a Perkin Elmer 2400 Series II CHN Analyser.

C. Metalloporphyrin catalysts synthesis

C.1. 5,10,15,20-Tetrakis-(3,5-dimethoxyphenyl)porphyrin – H₂T3,5DMPP (1). H₂T3,5DMPP was synthesized via the procedure described by Gonsalves et al.³⁰ To this end, 3.3463 g (19.735 mmol) of 3,5-dimethoxybenzaldehyde, 1.40 mL (19.775 mmol) of pyrrole, and 50 mL of a propionic/nitrobenzene mixture (7 : 3) were submitted to reflux for 1 h. After this period, the reaction mixture was cooled to ambient temperature (∼20 °C) in an ice bath, and 50 mL of methanol was added to the mixture, which was left to stand for approximately 1 h. The precipitate (porphyrin) was filtered in a Buchner funnel and washed with methanol until the solid became colorless. The solid was then solubilized and collected in dichloromethane. The solvent was removed in a rotary evaporator. The product was percolated in a neutral alumina chromatographic column (diameter = 2.7 cm and height = 26.0 cm) containing a 2.0 cm-thick upper layer of basic alumina and eluted with dichloromethane. Next, the fractions with the porphyrin were collected, and the solvent was eliminated in a rotary evaporator. The product was kept in a desiccator with silica gel. Yield: 26% (1.1192 g, 1.3091 mmol). UV-vis. (CH₂Cl₂), λ_max, nm (log ε): 421 (5.60); 514 (4.31); 549 (3.81); 589 (3.83); 645 (3.59). FTIR in KBr (cm⁻¹): (1604) δ C=C; (1294) δ of the porphyrin skeleton; (1249) ν C–O–C; (1173) ν OCH₃; (966) δ N–H (pyrrole). ¹H NMR: (CDCl₃, 25 °C, TMS) δ: −2.81 (s, 2H); 3.98 (s, 24H); 6.92 (s, 4H); 7.42 (m, 8H); 8.95 (s, 8H).

C.2. 5,10,15,20-Tetrakis-(3,5-dimethoxyphenyl)porphyrinatomanganese(III) chloride – [Mn^IIIT3,5DMPP]Cl (Cat.1). The metalloporphyrin (Cat.1) (Fig. 1) was synthesized in a similar manner to that used to prepare Mn^IIIAPTPPCl, according to the protocol described by Silva et al.¹⁶ which employed the free-base porphyrin (514.43 mg; 0.60171 mmol) and Mn(OAc)₂·4H₂O (1.4778 g; 6.0296 mmol). Yield: 98% (567.71 mg, 0.58734 mmol). UV-vis. (CH₂Cl₂), λ_max,⁸⁰ nm (log ε): 377 (4.68); 481 (5.01); 581 (3.95); 617 (3.88). FTIR in KBr (cm⁻¹)⁸⁰: (1593) δ C=C; (1206) ν C–O–C; (1155) ν OCH₃; (1009) δ Mn–N (pyrrole). ESI-TOF [Mn^IIIT3,5DMPP]⁺ m/z 907.35 (100%).

C.3. 5,10,15,20-Tetrakis-(4-bromine-3,5-dimethoxyphenyl)-2,3,7,8,12,13,17,18-octabromoporphyrinatomanganese chloride – [Mn^IIIBr₁₂T3,5DMPP]Cl. The new metalloporphyrin [Mn^IIIBr₁₂T3,5DMPP]Cl (Fig. 1) was synthesized in a similar manner to that used to prepare Mn^IIIBr₉APTPPCl, as described by Silva et al.¹⁶ [Mn^IIIT3,5DMPP]Cl (15.20 mg, 0.01572 mmol) and liquid bromine (∼0.2 mL, 3.7 mmol, 100-fold molar excess) were employed and the reaction lasted 24 h. The yield was 85% (25.44 mg, 0.08005 mmol). Elemental analysis. Found: C, 31.1; H, 1.8; N, 2.3. Calc. for C₅₂H₃₂Br₁₂ClMnN₄O₈. 1.5CHCl₃: C, 31.1; H, 1.6, N, 2.7%. UV-vis. (CH₂Cl₂), λ_max, nm (log ε): 394 (4.71); 502 (4.58); 611 (4.03); 656 (4.04). FTIR in KBr (cm⁻¹): (1569) δ C=C; (1277) δ C_β–Br; (1210) ν C–O–C, (1155) ν OCH₃; (1022) δ Mn–N (pyrrole). ESI-TOF [Mn^IIIBr₁₂T3,5DMPP + CH₃OH]⁺ m/z 1886.02 (100%).

D. Alkane oxidation reactions

All the catalytic reactions were performed in 2 mL Wheaton® vials sealed with Teflon-faced silicon septa. Reactions were conducted under magnetic stirring, at 25 °C, for 90 min, according to procedures adapted from de Sousa et al.⁸¹ Oxidation was carried out in air; either PhI(OAc)₂ or PhIO was used as oxygen donor. Reaction mixtures comprised 2.0 × 10⁻⁴ mmol of the catalyst (MnP); 2.0 × 10⁻³ mmol of the oxidant (PhIO or PhI(OAc)₂); 100 μL of cyclohexane (0.93 mmol), 100 μL of n-hexane (0.76 mmol), or 100 μL of adamantane solution in dichloromethane (0.2 mol L⁻¹, 0.02 mmol); and 200 μL of dichloromethane. The catalyst/oxidant/substrate molar ratio was 1 : 10 : 4650 for cyclohexane, and 1 : 10 : 3800 for n-hexane. As for adamantane, the catalyst/oxidant/substrate molar ratio was 1 : 10 : 100, because adamantane was poorly soluble in dichloromethane. When deemed necessary, the reaction was quenched by addition of sulfite and borax.⁸² The reaction mixtures were analyzed by capillary gas chromatography; bromobenzene was the internal standard. The retention times of the products were confirmed by comparison with the retention times of authentic product samples.⁸¹ The yields were based on either initial PhIO or PhI(OAc)₂. Each reaction was accomplished at least three times, and the reported data represent the average of the results of these reactions. Errors in yields and selectivity were calculated on the basis of the reproducibility of the reactions. The degree of MnP destruction (bleaching) was determined by UV-vis spectroscopy at the end of the catalytic run. Control reactions were conducted in the absence of the catalyst. The effect of imidazole was studied by adding a 10 μL aliquot of an imidazole (Im) 1.0 × 10⁻² mol L⁻¹ solution in dichloromethane to the reaction medium. To test the effect of water on cyclohexane/adamantane oxidation, 0.5 μL of water was added to the reaction mixture.

Conclusions

The third-generation MnP (Cat.2) catalyzed cyclohexane, adamantane, and n-hexane oxidation by PhIO or PhI(OAc)₂ more efficiently than the second-generation MnP (Cat.1).

Regarding cyclohexane oxidation by PhIO or PhI(OAc)₂, Cat.2 provided high Cy-ol yields with high selectivity. On the other hand, addition of imidazole to these systems significantly increased the Cy-ol yields for Cat.1 but not for Cat.2. Moreover, addition of water did not improve total product yield in any of the cases.

Concerning adamantane oxidation by PhIO or PhI(OAc)₂, Cat.2 gave significantly higher 2-Adol yield as compared with Cat.1. This less thermodynamically favorable product has superior commercial value than 1-Adol, so Cat.2 displays a more interesting behavior than other catalysts described in the literature.^7,12,58 Addition of imidazole or water to the MnP/PhI(OAc)₂ systems significantly improved total product yield. Nevertheless, addition of water did not alter the total product yield for MnP/PhIO systems.

As for n-hexane oxidation, Cat.1 and Cat.2 presented low catalytic activity. However, the Cat.2/PhIO and the Cat.2/PhI(OAc)₂ systems displayed increased selectivity for the alcohols (2-ol and 3-ol). The selectivity of Cat.2 for 2-ol is noteworthy, because 2-ol has higher commercial value.

The good results obtained in the oxidation of alkanes shows the potential of the new catalyst for oxidation of others more reactive substrates (alkenes, drugs etc.).

Acknowledgements

Financial support from CAPES, CNPq, and FAPEMIG is gratefully acknowledged. Prof. Gulherme Sippel Machado, UFPR – Campus Litoral for the elemental analysis.

Notes and references

1. D. Mansuy, Coord. Chem. Rev., 1993, 125, 129–141.
2. M. Costas, Coord. Chem. Rev., 2011, 255, 2912–2932.
3. H. Lu and X. P. Zhang, Chem. Soc. Rev., 2011, 40, 1899–1909.
4. B. Meunier, S. P. de Visser and S. Shaik, Chem. Rev., 2004, 104, 3947–3980.
5. M. J. Gunter and P. Turner, Coord. Chem. Rev., 1991, 108, 115–161.
6. C. Zheng and S.-L. You, RSC Adv., 2014, 4, 6173–6214.
7. J. T. Groves, T. E. Nemo and R. S. Myers, J. Am. Chem. Soc., 1979, 101, 1032–1033.
8. J. T. Groves and Y. Watanabe, Inorg. Chem., 1986, 25, 4808–4810.
9. M. J. Bazin, H. Shi, J. Delaney, B. Kline, Z. Zhu, C. Kuhn, F. Berlioz, K. A. Farley, G. Fate, W. Lam, G. S. Walker, L. Yu and M. P. Pollastri, Chem. Biol. Drug Des., 2007, 70, 354–359.
10. S. Zakavi, F. Heidarizadi and S. Rayati, Inorg. Chem. Commun., 2011, 14, 1010–1013.
11. S. Nakagaki, K. A. D. F. Castro, G. M. Ucoski, M. Halma, V. Prevot, C. Forano and F. Wypych, J. Braz. Chem. Soc., 2014, 25, 2329–2338.
12. F. G. Doro, J. R. L. Smith, A. G. Ferreira and M. D. Assis, J. Mol. Catal. A: Chem., 2000, 164, 97–108.
13. E. do Nascimento, G. D. Silva, F. A. Caetano, M. A. M. Fernandez, D. C. da Silva, M. E. M. D. de Carvalho, J. M. Pernaut, J. S. Reboucas and Y. M. Idemori, J. Inorg. Biochem., 2005, 99, 1193–1204.
14. D. C. da Silva, G. DeFreitas-Silva, E. do Nascimento, J. S. Reboucas, P. Barbeira, M. E. M. D. de Carvalho and Y. M. Idemori, J. Inorg. Biochem., 2008, 102, 1932–1941.
15. B. R. S. Lemos, D. CarvalhoDa-Silva, D. Z. Mussi, L. d. S. Santos, M. M. da Silva, M. E. M. D. de Carvalho, J. S. Reboucas and Y. M. Idemori, Appl. Catal., A, 2011, 400, 111–116.
16. V. S. da Silva, L. I. Teixeira, E. do Nascimento, Y. M. Idemori and G. DeFreitas-Silva, Appl. Catal., A, 2014, 469, 124–131.
17. T. G. Traylor and S. Tsuchiya, Inorg. Chem., 1987, 26, 1338–1339.
18. M. Autret, Z. P. Ou, A. Antonini, T. Boschi, P. Tagliatesta and K. M. Kadish, J. Chem. Soc., Dalton Trans., 1996, 2793–2797.
19. O. Horvath, Z. Valicsek, G. Harrach, G. Lendvay and M. A. Fodor, Coord. Chem. Rev., 2012, 256, 1531–1545.
20. B. Meunier, Chem. Rev., 1992, 92, 1411–1456.
21. T. Wijesekera, A. Matsumoto, D. Dolphin and D. Lexa, Angew. Chem., Int. Ed. Engl., 1990, 29, 1028–1030.
22. P. E. Ellis and J. E. Lyons, Coord. Chem. Rev., 1990, 105, 181–193.
23. I. G. Denisov, M. M. Thomas, G. S. Stephen and I. Schlichting, Chem. Rev., 2005, 105, 2253–2277.
24. T. Higuchi, J. Synth. Org. Chem., Jpn., 2009, 67, 134–142.
25. J. S. Reboucas, B. O. Patrick and B. R. James, J. Am. Chem. Soc., 2012, 134, 3555–3570.
26. E. Mesbahi, N. Safari and M. Gheidi, J. Porphyrins Phthalocyanines, 2014, 18, 354–365.
27. M. J. Gunter and P. Turner, J. Mol. Catal., 1991, 66, 121–141.
28. C.-C. Guo, X.-Q. Liu, Q. Liu, Y. Liu, M.-F. Chu and W.-Y. Lin, J. Porphyrins Phthalocyanines, 2009, 13, 1250–1254.
29. Q. Liu and C. Guo, Sci. China: Chem., 2012, 55, 2036–2053.
30. A. Gonsalves, J. Varejao and M. M. Pereira, J. Heterocycl. Chem., 1991, 28, 635–640.
31. T. P. Wijesekera and D. A. Dolphin, Metalloporphyrins in Catalytic Oxidations, Marcel Dekker, New York, 1994.
32. A. Harriman, J. Chem. Soc., Faraday Trans. 1, 1981, 77, 369–377.
33. M. Gouterman, in The Porphyrins, ed. D. Dolphin, Academic Press, New York, 1978, vol. 3, ch. 1, pp. 1–165.
34. H. Kobayash, Y. Yanagawa, H. Osada, S. Minami and M. Shimizu, Bull. Chem. Soc. Jpn., 1973, 46, 1471–1479.
35. R. A. Richards, K. Hammons, M. Joe and G. M. Miskelly, Inorg. Chem., 1996, 35, 1940–1944.
36. P. Bhyrappa, B. Purushothaman and J. J. Vittal, J. Porphyrins Phthalocyanines, 2003, 7, 682–692.
37. M. Sankar, P. Bhyrappa, B. Varghese, K. K. Praneeth and G. Vaijayanthimala, J. Porphyrins Phthalocyanines, 2005, 9, 413–422.
38. G. d. F. Silva, D. C. da Silva, A. S. Guimaraes, E. do Nascimento, J. S. Reboucas, M. P. de Araujo, M. E. Moreira Dai de Carvalho and Y. M. Idemori, J. Mol. Catal. A: Chem., 2007, 266, 274–283.
39. G. Hariprasad, S. Dahal and B. G. Maiya, J. Chem. Soc., Dalton Trans., 1996, 3429–3436.
40. P. Bhyrappa and B. Purushothaman, J. Chem. Soc., Perkin Trans. 2, 2001, 238–242.
41. A. Ghosh, I. Halvorsen, H. J. Nilsen, E. Steene, T. Wondimagegn, R. Lie, E. van Caemelbecke, N. Guo, Z. P. Ou and K. M. Kadish, J. Phys. Chem. B, 2001, 105, 8120–8124.
42. G. R. Friedermann, M. Halma, K. A. Dias de Freitas Castro, F. L. Benedito, F. G. Doro, S. M. Drechsel, A. S. Mangrich, M. d. D. Assis and S. Nakagaki, Appl. Catal., A, 2006, 308, 172–181.
43. G. DeFreitas-Silva, J. S. Reboucas, I. Spasojevic, L. Benov, Y. M. Idemori and I. Batinic-Haberle, Arch. Biochem. Biophys., 2008, 477, 105–112.
44. J. S. Reboucas, G. DeFreitas-Silva, I. Spasojevic, Y. M. Idemori, L. Benov and I. Batinic-Haberle, Free Radical Biol. Med., 2008, 45, 201–210.
45. K. Prakash, R. Kumar and M. Sankar, RSC Adv., 2015, 5, 66824–66832.
46. S. Konishi, M. Hoshino and M. Imamura, J. Phys. Chem., 1982, 86, 4537–4539.
47. M. T. Caudle, C. K. Mobley, L. M. Bafaro, R. LoBrutto, G. T. Yee and T. L. Groy, Inorg. Chem., 2004, 43, 506–514.
48. E. P. Talsi and K. P. Bryliakov, Mendeleev Commun., 2004, 14, 111–112.
49. K. A. D. d. F. Castro, M. M. Q. Simões, M. d. G. P. M. S. Neves, J. A. S. Cavaleiro, R. R. Ribeiro, F. Wypych and S. Nakagaki, Appl. Catal., A, 2015, 503, 9–19.
50. C. J. Weschler, B. M. Hoffman and F. Basolo, J. Am. Chem. Soc., 1975, 97, 5278–5280.
51. B. M. Hoffman, T. Szymanski, T. G. Brown and F. Basolo, J. Am. Chem. Soc., 1978, 100, 7253–7259.
52. K. Ozette, P. Battioni, P. Leduc, J. F. Bartoli and D. Mansuy, Inorg. Chim. Acta, 1998, 272, 4–6.
53. S. M. d. M. Romanowski, S. P. Machado, G. R. Friedermann, A. S. Mangrich, M. d. F. Hermann, H. O. Lima and S. Nakagaki, J. Braz. Chem. Soc., 2010, 21, 842–850.
54. L. J. Boucher and J. J. Katz, J. Am. Chem. Soc., 1967, 89, 1340–1345.
55. C.-M. Che, V. K.-Y. Lo, C.-Y. Zhou and J.-S. Huang, Chem. Soc. Rev., 2011, 40, 1950–1975.
56. E. Roduner, W. Kaim, B. Sarkar, V. B. Urlacher, J. Pleiss, R. Glaeser, W.-D. Einicke, G. A. Sprenger, U. Beifuss, E. Klemm, C. Liebner, H. Hieronymus, S.-F. Hsu, B. Plietker and S. Laschat, ChemCatChem, 2013, 5, 82–112.
57. Z. Li, C. G. Xia and M. Ji, Appl. Catal., A, 2003, 252, 17–21.
58. E. Baciocchi, T. Boschi, C. Galli, A. Lapi and P. Tagliatesta, Tetrahedron, 1997, 53, 4497–4502.
59. M. Gardner, A. J. Guerin, C. A. Hunter, U. Michelsen and C. Rotger, New J. Chem., 1999, 23, 309–316.
60. V. V. Zhdankin and J. D. Protasiewicz, Coord. Chem. Rev., 2014, 275, 54–62.
61. Z. Li and C. G. Xia, J. Mol. Catal. A: Chem., 2004, 214, 95–101.
62. J. P. Collman, A. S. Chien, T. A. Eberspacher and J. I. Brauman, J. Am. Chem. Soc., 2000, 122, 11098–11100.
63. V. S. da Silva, A. M. Meireles, D. C. da Silva Martins, J. S. Reboucas, G. DeFreitas-Silva and Y. M. Idemori, Appl. Catal., A, 2015, 491, 17–27.
64. A. A. Guedes, A. C. M. A. Santos and M. D. Assis, Kinet. Catal., 2006, 47, 555–563.
65. Z. Valicsek, O. Horvath, G. Lendvay, I. Kikas and I. Skoric, J. Photochem. Photobiol., A, 2011, 218, 143–155.
66. G. R. Karimipour, H. A. Shadegan and R. Ahmadpour, J. Chem. Res., 2007, 252–256.
67. J.-H. In, S. E. Park, R. Song and W. Nam, Inorg. Chim. Acta, 2003, 343, 373–376.
68. E. Baciocchi, T. Boschi, L. Cassioli, C. Galli, A. Lapi and P. Tagliatesta, Tetrahedron Lett., 1997, 38, 7283–7286.
69. D. Balcells, C. Raynaud, R. H. Crabtree and O. Eisenstein, Inorg. Chem., 2008, 47, 10090–10099.
70. K. W. Kwong, T.-H. Chen, W. Luo, H. Jeddi and R. Zhang, Inorg. Chim. Acta, 2015, 430, 176–183.
71. T.-H. Chen, K. W. Kwong, A. Carver, W. Luo and R. Zhang, Appl. Catal., A, 2015, 497, 121–126.
72. V. S. da Silva, Y. M. Idemori and G. DeFreitas-Silva, Appl. Catal., A, 2015, 498, 54–62.
73. G. M. Ucoski, K. A. Dias de Freitas Castro, K. J. Ciuffi, G. P. Ricci, J. A. Marques, F. S. Nunes and S. Nakagaki, Appl. Catal., A, 2011, 404, 120–128.
74. R. H. Crabtree, J. Chem. Soc., Dalton Trans., 2001, 2951.
75. A. Sorokin, A. Robert and B. Meunier, J. Am. Chem. Soc., 1993, 115, 7293–7299.
76. Y. Shiota, N. Kihara, T. Kamachi and K. Yoshizawa, J. Org. Chem., 2003, 68, 3958–3965.
77. B. R. Cook, T. J. Reinert and K. S. Suslick, J. Am. Chem. Soc., 1986, 108, 7281–7286.
78. J. A. Kerr, Chem. Rev., 1966, 66, 465–500.
79. H. Saltzman and J. G. Sharefkin, Org. Synth., 1963, 43, 62–64.
80. S. Yanga, B. Suna, Z. Ou, D. Menga, G. Lua, Y. Fangb and K. M. Kadish, J. Porphyrins Phthalocyanines, 2013, 17, 857–869.
81. A. de Sousa, M. de Carvalho and Y. Idemori, J. Mol. Catal. A: Chem., 2001, 169, 1–10.
82. M. Ribani, C. Bottoli, C. Collins, I. Jardim and L. Melo, Quim. Nova, 2004, 27, 771–780.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra20690a

---