Sign in  |  Register
Cart

CSJ Journals

▾
  • -

2020, Vol.93, No.12

1571-1582
Award AccountsOpen Access
The Chemical Society of Japan Award for Creative Work for 2019

Study on Proton-Coupled Electron Transfer in Transition Metal Complexes

Takahiko Kojima
Department of Chemistry, Faculty of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8571, Japan
https://doi.org/10.1246/bcsj.20200213

Tools
PDF (4099K)

Abstract

Section:
Next section

In this award account are described our achievements in proton-coupled electron transfer reactions performed by transition metal complexes. Topics include formation and reactivity of RuIV-oxo and RuIII-oxyl complexes, and hydrogen-atom transfer from substrates to RuIII-pterin complexes. Based on kinetic analysis of the reactions, mechanistic insights into those reactions are gained to shed light on the reactivity of metal complexes as hydrogen-atom acceptors. In addition, we present our efforts in artificial photosynthesis concerning efficient and photocatalytic water oxidation by a dinuclear CoIII complex as well as selective CO2 reduction by NiII complexes.

figure

Proton-coupled electron transfers (PCET) in metal complexes are described, mainly focusing on the formation and reactivity of RuIV-oxo, RuIII-oxyl, and RuIII-pterin complexes. Water oxidation by a CoIII complex and CO2 reduction by NiII complexes are also discussed in light of PCET.

1. Introduction
Section:
Previous sectionNext section

Proton-coupled electron transfer (PCET) is a fundamental and ubiquitous reaction process in many kinds of redox reactions.1 As the concept of PCET described in Scheme 1, in a PCET reaction, a proton is transferred to a proton-accepting site that should be basic and an electron is also transferred to an electron-accepting site that should have a hole, separately.2 PCET is a kind of hydrogen atom transfer (HAT); however, the hydrogen atom (H• = H+ + e−) is transferred as an electron and a proton to different sites. Thus, a HAT reaction from a X-H bond to, for example, an alkoxy radical is not PCET but called a “canonical HAT”, since the proton and electron go to the same site.3 PCET processes can be described as a thermochemical square scheme for a metal complex ((L)M(n+1)+) as shown in Scheme 2. PCET involves sequential proton transfer and electron transfer (PT/ET; upper path in Scheme 2), sequential electron transfer and proton transfer (ET/PT; lower path in Scheme 2), and concerted proton-electron transfer (CPET; middle path in Scheme 2), depending on the characteristics of the hydrogen donor as well as the hydrogen acceptor.4 In Scheme 2, L and M(n+1)+ in “(L)M(n+1)+” act as a proton acceptor (Ap in Scheme 1) and an electron acceptor (Ae in Scheme 1), respectively. For example, as shown in Scheme 3, a high valent metal-oxo complex (M(n+1)+=O) can act as a hydrogen atom acceptor. In this reaction, an electron is transferred to the metal center and a proton is accepted by the oxo ligand, separately. As seen in Scheme 2, the driving force of a PCET reaction is governed by the proton acceptability (basicity) of a proton acceptor and the electron acceptability (redox potential) of an electron acceptor. Thus, the driving force of a PCET reaction can be described by a bond dissociation free energy (BDFE) of the reaction as given in eq 1:5

\begin{equation} \text{BDFE} = 1.37 \text{p}K_{\text{a}} + 23.06E_{1/2} + C_{\text{sol}} \end{equation}
(1)
where Csol is a solvent-dependent constant and essentially ΔG° of the reaction, H• → H+ + e−, in a solvent. For example, Csol has been reported to be 59.4 kcal mol−1 for CH3CN (Fc+/Fc as a reference) and 55.8 kcal mol−1 for H2O (NHE as a reference).4

PCET reactions have been observed in oxidation of water to produce dioxygen in photosynthesis to obtain electrons and protons; electrons are used for the production of NADPH that acts as a ubiquitous reductant in biological systems and is required for CO2 fixation and protons generated contribute to production of ATP as the indispensable energy source of living cells.6 In the water oxidation at the reaction center of Photosystem II, a tetranuclear manganese-oxo cluster, which is known as “oxygen-evolving complex (OEC)”, acts as a catalyst undergoing 4-step photoinduced PCET processes through the Kok cycle involving the formation of a high-valent Mn-oxo species to afford dioxygen.7,8

Inspired by the water oxidation by OEC, PCET oxidation of a low-valent metal-aqua complex has been established as a method to generate a high-valent metal-oxo complex. In 1978, Meyer and Moyer reported a pioneering work to demonstrate the formation of a RuIV-oxo complex, [RuIV(O)(bpy)2(py)]2+, by oxidizing [RuII(bpy)2(py)(H2O)]2+ as a precursor complex with (NH4)2[CeIV(NO3)6] (CAN) as an oxidant.9 The PCET oxidation has been applied to water oxidation using a dinuclear μ-oxo RuIII-aqua complex, “the blue dimer” as a catalyst, which is converted to a O=RuV-O-RuV=O complex acting as a reactive species in water oxidation.10 The strategy has been applied to oxidation of organic substrates by high-valent Ru=O complexes and to gaining mechanistic insights into oxidation reactions performed by high-valent metal-oxo complexes.11,12

Other than high-valent metal-oxo complexes, Mayer and coworkers have reported PCET oxidation of organic substrates by metal complexes having proton-accepting sites. For example, they have reported on long-range CPET from phenols to a RuIII complex having a terpy derivative with a carboxylate group that acts as a proton acceptor.13 In the course of CPET from a phenol derivative to the RuIII complex, a phenolic proton goes to the carboxylate group and an electron is accepted at the RuIII center, even though the RuIII center is 7 Å away from the proton-accepting carboxylate oxygen.13

In this article, we would like to describe out achievements in PCET in transition metal complexes. The topics include formation and reactivity of RuIV=O and electronically equivalent RuIII-oxyl (RuIII-O•) complexes, and PCET oxidation of organic compounds with RuIII-pterin complexes as non-oxo hydrogen-acceptors. Also, as attempts at artificial photosynthesis, photocatalytic water oxidation by a dinuclear bis(μ-oxo) CoIII complex and photocatalytic hydrogen evolution and selective CO2 reduction to afford CO using NiII complexes are presented.

2. Formation and Reactivity of RuIV-Oxo Complexes
Section:
Previous sectionNext section

Formation of an oxidatively reactive high-valent metal-oxo complex through PCET in water should have advantages in catalytic oxidation of organic substrates: We can achieve selective and clean formation of reactive species and construction of an environmentally benign catalytic system using water not only as a clean solvent but also as the sole oxygen source.

We have prepared and characterized RuIV=O complexes generated by PCET oxidation of the corresponding RuII-aqua complexes using CAN as an oxidant in water (Figure 1). In the case of [RuII(H2O)2(TPA)]2+ (1), a RuIV=O complex in the S = 1 state, [RuIV(O)(TPA)(H2O)]2+ (2), is formed.14 The complex 2 shows high reactivity in oxidation of hydrocarbons in acidic water. For example, cyclohexene is catalytically oxidized by 1 as a catalyst and CAN as an oxidant to afford adipic acid in a high turnover number (TON) and high selectivity. In this process, the double bond of cyclohexene is epoxidized and then hydrolyzed to afford cyclohexane diol under acidic conditions. The diol is further oxidized to afford adipic acid.14

When a carboxyl group is introduced to the 6-position of one of three pyridine rings of the TPA ligand (6-COOH-TPA), the ligand acts as a pentadentate ligand including the coordination of the carboxylate moiety to form [RuII(6-COO−-TPA)(H2O)]+ (3).15 PCET oxidation of 3 with CAN in water allows us to obtain a 7-coordinate RuIV-oxo complex in the S = 0 state, [RuIV(O)(6-COO−-TPA)(H2O)]+ (4), which is the first example of a RuIV=O complex in the low-spin state.15 The 7-coordinate structure has been revealed to be stabilized by hydrogen bonding with water molecules. Furthermore, a pentadentate pyridylamine ligand, N4Py, affords [RuII(N4Py)(H2O)]2+ (5), which is oxidized by CAN to form another diamagnetic RuIV=O complex, [RuIV(O)(N4Py)(H2O)]2+ (6) in water.16 DFT calculations have suggested that those diamagnetic 7-coordinate RuIV=O complexes 4 and 6 should be in a pentagonal bipyramidal structure with an extra aqua ligand and the structures are stabilized by hydrogen bonding with water molecules in water as shown in Figure 2.15,16

Reactivity of the three RuIV=O complexes (2, 4, 6) in C-H oxidation has been examined in water to compare the reactivity.17 Alcohols are oxidized to the corresponding aldehydes and ketones; the starting RuII-aqua complexes (1, 3, 5) are recovered quantitatively. For each RuIV=O complex, pseudo-first-order rate constants of alcohol oxidation have exhibited saturation relative to the substrate concentrations, indicating that the RuIV=O complexes form adducts with alcohols prior to the oxidation reactions in water. The driving force of the adduct formation has been analyzed using 1,1,1,3,3,3-hexafluoro-2-propanol (hfp) as an oxidatively tolerant substrate. Spectroscopic titration by 19F NMR spectroscopy has allowed us to conclude that the adduct is formed by hydrogen bonding of the aqua ligand. The equilibrium constant of the adduct formation between 2 and hfp has been determined to be 3.6 × 102 M−1. In the hydrogen-bonded adduct, the hydrogen atom transfers from the interacting alcohol to the RuIV=O complex in accordance with the first-order kinetics, showing kinetic isotope effect (KIE) for the corresponding C-deuterated substrate such as CD3OH but not for CH3OD.

The kinetic analysis on the substrate oxidation reactions by the RuIV=O complexes (2, 4, 6 in Figure 1) clearly demonstrates that the spin states do not exert impact on the reactivity of the species. Thermochemical square schemes for 2 and 6 are shown in Figure 3, together with obtained pKa and E1/2 values, which have been determined by spectroscopic titration and electrochemical measurements in Britton-Robinson buffer solutions. The BDEs of the O-H bonds for the RuIII-OH complexes have been calculated on the basis of their Pourbaix diagrams and eq 1 to be 82.7 kcal mol−1 for 2 and 84.7 kcal mol−1 for 6.18 Thus, the thermodynamic hydrogen acceptability of the two complexes are comparable showing similar reactivity in the PCET oxidation of substrates. Arguments on the impact of the spin states of high-valent metal-oxo complexes on the reactivity have been of importance,19,20 and have been expanded from those on FeIV=O complexes: High-spin FeIV=O complexes (S = 2) have been demonstrated to show higher reactivity in substrate oxidation than the low-spin counterparts (S = 1) in terms of rate constants. In light of our results, the reactivity seems to be controlled by the BDE value of an O-H bond of a M(n−1)+-hydroxo complex as a product of hydrogen atom abstraction by a Mn+-oxo complex.

The complexes 1, 3, 5 have been applied as catalysts to highly efficient and selective photocatalytic oxidation of organic substrates in water, using [RuII(bpy)3]2+ as a photosensitizer and [CoIIICl(NH3)5]2+ as a weak oxidant,21 in Britton-Robinson buffer (pH = 1.8) under visible-light irradiation (λ > 380 nm).22 Photoinduced ET from the triplet excited state of [RuII(bpy)3]2+ to the CoIII complex occurs to form [RuIII(bpy)3]3+ (Ered = 1.25 V vs NHE) as a real oxidant to afford the corresponding RuIV=O complexes 2, 4, and 6 as reactive species in oxidation reactions (Scheme 4). In the photocatalytic oxidation of alcohols, the three complexes showed high efficiency relative to the oxidant used up to 100% in the oxidation of 4-methyl-benzylalcohol. The quantum yields of the reaction have been determined to be 0.35 for 1, 0.33 for 3, and 0.31 for 5, respectively.

In the course of C-H hydroxylation of organic substrates, an “oxygen-rebound” mechanism has been proposed and accepted for a long time (Scheme 5);23 however, no direct evidence has been provided. We prepared a RuIV=O complex from a coordinatively saturated RuII complex, [RuII(TPA)(bpy)]2+,24 through the oxidation with CAN in water. In the course of the PCET oxidation of the RuII precursor, one of pyridines of TPA dissociates from the Ru center to afford [RuIV(O)(η3-H+TPA)(bpy)]3+ (7), which has been spectroscopically and crystallographically characterized (Figure 4).25 The reactivity of 7 in C-H oxidation has been investigated using CH3CN as a solvent. When cumene is oxidized by 7, cumyl alcohol is obtained as a main product and [RuII(η3-H+TPA)(bpy)(CH3CN)]3+ (8). ESI-MS analysis of the reaction mixture including 7 and cumene has revealed the formation of a RuIII-cumyl alkoxo complex, which is assumed to be derived from oxidation of the corresponding RuII-cumyl alcohol complex as the oxygen-rebound product by 7. The RuII-cumyl alcohol complex cannot be formed by a reaction of 8 with excess cumyl alcohol in CH3CN. The C-H hydroxylation proceeds in a two-step mechanism involving second-order PCET, which occurs in an intermolecular manner, followed by the oxygen-rebound mechanism to form a RuII-alcohol complex; the subsequent first-order ligand substitution of the alcohol complex, in which the alcohol ligand is substituted by a CH3CN molecule from the solvent, proceeds to generate 8. As for the first PCET process, we can observe KIE of 12 for the cumene oxygenation; however, no KIE has been observed for the second step.

The PCET process has been kinetically analyzed to obtain a linear relationship between logarithms of normalized rate constants and BDEs of C-H bonds to be cleaved in substrates. The linear relationship can be discussed on the basis of Bell-Evans-Polanyi (BEP) equation as follows:26

\begin{equation} E_{\text{a}} = \alpha\Delta H + \text{C} \end{equation}
(2)
\begin{equation} \Delta H = \text{BDE}\ (\text{to be formed}) - \text{BDE}\ (\text{to be cleaved}) \end{equation}
(3)
The slope of a BEP plot for 7 has been determined to be −0.44 in CH3CN (Figure 5, red line), which is comparable to that (−0.36) obtained for [RuIV(O)(bpy)2(py)]2+ in CH3CN.27 In sharp contrast, complex 2 shows a linear BEP plot with α = −0.08 in water (Figure 5, blue line).17 The difference of the α values between 2 in water and 7 in CH3CN indicates that of the transition states of CPET between in water and in CH3CN. The first-order PCET from a substrate to 2 in the hydrogen-bonded adduct is elucidated to proceed in a substrate-like early transition state. On the other hand, the second-order PCET from a substrate to 7 in CH3CN proceeds via a mid-point transition state.

ET properties of a RuIV=O complex, [RuIV(O)(MeBPA)-(bpy)]2+ (9; Figure 6a), which can be obtained by PCET oxidation of [RuII(MeBPA)(bpy)(H2O)]2+ with CAN in water, has been scrutinized in CH3CN using various electron donors on the basis of kinetic analysis in light of the Marcus theory of ET.28 The driving-force dependence of rate constants of ET (kET) and PCET (kPCET) for 9 and [RuIV(O)(bpy)2(py)]2+ (Figure 6b) in the presence of trifluoroacetic acid has been analyzed using the Marcus equation of adiabatic ET as follows:29

\begin{equation} k_{\text{ET or PCET}} = Z\exp [ - (\lambda /4)(1 + \Delta G/\lambda )^{2}/k_{\text{B}}T] \end{equation}
(4)
where Z is the collision frequency (1 × 1011 M−1 s−1 in CH3CN), λ is the reorganization energy of ET, ΔG is the Gibbs energy change in an ET reaction (−ΔG is the driving force of ET), kB is the Boltzmann constant, and T is the absolute temperature. We have succeeded in the first determination of reorganization energy (λ) of ET and PCET of RuIV=O complexes using isolated [RuIV(O)(bpy)2(py)]2+ and 9. The λ value of ET has been determined to be 1.70 ± 0.06 eV and that of PCET to be 1.26 ± 0.04 eV for 9 on the basis of Marcus plots as shown in Figure 7. The λ value of ET is much smaller than those of CrV=O,30 FeIV=O,31 and MnIV=O complexes,32 due to the smaller structural change of 9 after ET as suggested by DFT calculations. This smaller structural change is because of the lack of electrons in dσ orbitals in the 1e−-reduced RuIII state. In addition, the smaller λ value of PCET than that of ET indicates that PCET is more kinetically favorable than ET.

3. Formation and Reactivity of a RuIII-Oxyl Complex
Section:
Previous sectionNext section

As an ancillary ligand of a RuII-aqua complex, we have used an N-heterocyclic carbene (NHC) ligand to control the reactivity of a RuIV=O complex, in which the NHC ligand binds to the trans position of the oxo ligand. The tridentate NHC ligand, 1,3-bis(2-pyridyl-methyl)imidazol-2-ylidene (BPIm),33 binds to a RuII(bpy) unit in a facial mode to form [RuII(BPIm)(bpy)(H2O)]2+ (10; Scheme 6).34 The complex shows one reversible redox wave at 0.86 V vs. NHE in an acidic aqueous solution at pH 2.5 and also an irreversible wave at 1.6 V, which is comparable to the reduction potential of CAN under the conditions. The complex can be oxidized by CAN to form an RuIII-aqua complex, [RuIII(BPIm)(bpy)(H2O)]3+ (11; Scheme 6), whose structure has been determined by X-ray crystallography. The RuIII-aqua complex undergoes 1e−/2H+ PCET oxidation to afford not a RuIV=O complex but an electronically equivalent RuIII-oxyl complex, [RuIII(O•)(BPIm)(bpy)]2+ (12; Scheme 6).34 The PCET oxidation processes have been clarified in light of the Pourbaix diagram of 10 under acidic conditions.34

The RuIII-oxyl (RuIII-O•) complex has been characterized by various spectroscopic methods. XANES spectroscopy has indicated that the Ru center of 12 is in the almost RuIII state rather than the RuIV state (Figure 8a). Resonance Raman spectroscopy has been applied to observe a Raman scattering assigned to the Ru-O bond of 12 at 732 cm−1, which shifts to 696 cm−1 with use of H218O as a solvent (Figure 8b). The Raman data indicate that the Ru-O bond of 12 is much weaker than that of a typical RuIV=O complex, which shows a Raman band in the range of 780–840 cm−1. The bond order of the Ru-O bond in 12 has been calculated to be 1.3, supporting the oxyl character of the oxo ligand. EXAFS analysis on 12 has been performed to estimate the Ru-O bond length to be 1.77(1) Å. Note that the Ru-O bond length is not a reliable parameter to deduce any oxidation state of a Ru center having an oxo ligand.11a

The RuIII-O• complex 12 shows unique reactivity, which is totally different from that of RuIV=O complexes. Kinetic analysis of oxidation reactions of benzaldehyde derivatives in acidic water sheds light on the reactivity of 12 as represented by the ρ value of −0.07 in a Hammett plot, which is much smaller than that (−0.65) reported for oxidation reactions of benzaldehyde derivatives with a RuIV=O complex. The extremely small ρ value strongly indicates strong radical character of 12 in the hydrogen abstraction from the formyl group.34

Further exploration of the reactivity of 12 has reached unique oxidative cracking of aromatic rings in acidic water to afford formic acid, which can be converted to H2 as an energy source using a RhIII catalyst,35 as shown in Scheme 7.36 The oxidative cracking of aromatic rings can be observed for various benzene derivatives and naphthalene. It should be emphasized that the reactive species 12 shows electrophilic reactivity on an aromatic ring to show the ρ value of −1.41 in a Hammett plot for the normalized total efficiency. Catalytic oxidation of electron-rich benzene derivatives having alkyl substituents has been examined using 10 as the catalyst and CAN as an oxidant. In the reactions, the alkyl groups are not oxidized to afford formic acid and carboxylic acids having the corresponding alkyl substituents. For example, ethylbenzene is oxidized under the conditions to afford formic acid and propionic acid. Thus, the reactivity of the RuIII-O• complex 12 has been shown to be totally different from that of RuIV=O complexes, which should oxidize the benzylic C-H bonds selectively, in terms of strong radical character. A proposed mechanism of the oxidative aromatic ring cracking is depicted in Scheme 8. After the electrophilic addition of the oxyl radical ligand in 12 to an aromatic ring to afford an equilibrium mixture of arene oxide and oxepin,37 the mixture is further oxidized to form a muconaldehyde derivative followed by a muconic acid derivative, the latter of which is further oxidized to afford formic acid and a carboxylic acid. In the course of the reaction, we can observe the formation of a p-quinone derivative, which has been confirmed to be oxidized to form formic acid as well.

Metal-oxyl complexes have been recently proposed as reactive species in oxidation reactions.38 Accumulation of information of the metal-oxyl complexes should be provided to tackle harder oxidation reactions including methane oxidation.

4. PCET from X-H Bonds to RuIII-Pterin Complexes
Section:
Previous sectionNext section

As described in Scheme 1, PCET reactions require a proton-accepting site and an electron-accepting site for a hydrogen atom (H• = H+ + e−) acceptor. This means that high-valent metal-oxo complexes are not necessarily required for hydrogen-atom abstraction. We have adopted RuIII complexes having pterins, which are heteroaromatic coenzymes, as ligands. In a RuIII-pterin complex, a pterin ligand acts as a proton acceptor and the RuIII center acts as an electron acceptor in PCET reactions.39,40 We have used RuIII-pterin complexes, [RuIII(dmp)(TPA)]2+ (13), [RuIII(dmdmp)(TPA)]2+ (14), and [RuIII(dmdmp)(Cl)(MeBPA)]+ (15), which are depicted in Figure 9, as hydrogen-atom acceptors. The RuIII-pterin complexes have been prepared by oxidation of the corresponding RuII-pterin complexes41 using [RuIII(bpy)3]3+ as an oxidant in CH3CN. Reactions of 13–15 with organic substrates have been kinetically analyzed to gain detailed mechanistic insights into PCET from phenolic O-H bonds and aliphatic C-H bonds as shown in Scheme 9.

The difference between 13 and 14 is the presence or absence of two methyl groups on the 2-amino group of the pterin ligands. Protonation of the pterin ligand occurs at the 1-N-position as evidenced by the crystal structure of the protonated form of the corresponding RuII precursor complexes.42 Determination of pKa values of the pterin ligands and redox potentials of the Ru centers in CH3CN has allowed us to provide thermochemical square schemes involving 13 (Figure 10a) and 14. Based on the square schemes, we have calculated the BDE values of PCET processes using eq 1 to be 85 kcal mol−1 for 13 and 89 kcal mol−1 for 14. When 13 and 14 react with phenol derivatives in CH3CN, [RuII(Hdmp)(TPA)]2+ from 13 and [RuII(Hdmdmp)(TPA)]2+ from 14 are formed as products, respectively; in the protonated RuII-pterin products, the nitrogen atom of 1-position in the pterin framework is protonated as shown in Scheme 9.40,42 When 2,4,6-tri-tert-butylphenol is used as a substrate, formation of stable 2,4,6-tri-tert-butylphenoxyl radical as a PCET product can be observed by ESR spectroscopy. Apparently, the reactions obey the second-order kinetics under pseudo-first-order conditions. In the case of the reaction of 13 with 4-nitrophenol, an adduct composed of both compounds can be detected by ESI-MS spectrometry at 233 K, although no adduct formation is observed for 14.39 The adduct formation for 13 strongly suggests that the 2-amino group of the dmp− ligand in 13 is operative in the adduct formation, probably through the hydrogen bonding with the OH group of the phenol as a substrate. KIE has been observed for the phenol oxidation, indicating that PCET from phenols to 13 and 14 should be involved in the rate-determining step. Correlations of logarithm of the second-order rate constants (log k) with pKa values, redox potentials and BDEs of O-H bonds of phenol derivatives have been scrutinized to indicate that the oxidation of phenolic O-H bonds should proceed via CPET, rather than ET/PT or PT/ET mechanisms.39

We have also investigated PCET reactions from C-H bonds to two RuIII-pterin complexes, 14 and 15, to elucidate the impact of proton acceptability and electron acceptability of a hydrogen-atom acceptor on the reaction mechanism and the transition state of a PCET process.40 As shown in Figure 10, complexes 14 and 15 show different pKa values of the pterin ligands and redox potentials of the Ru centers. The pKa value of 14 has been calculated to be 9.9 and that of 15 to be 11.3 on the basis of the corresponding square schemes and eq 1. On the contrary, the E1/2 values have been determined by cyclic voltammetry to be +0.52 V (vs. Fc+/Fc) for protonated 14 and +0.14 V for protonated 15. Thus, complex 14 should exhibit weaker proton acceptability and higher electron acceptability than 15. This trend should be derived from the difference of the charge between dicationic 14 and monocationic 15.

We have examined C-H oxidation reactions of a series of organic substrates using 14 and 15 as hydrogen-atom acceptors in CH3CN. Kinetic analysis of the reactions has allowed us to determine the second-order rate constants and the corresponding activation parameters. Analysis of the rate constants in light of BEP plots has shed light on the difference of the transition states of the CPET from substrates to the RuIII-pterin complexes: the α values in eq 2 have been determined to be −0.27 for 14 and −0.44 for 15. These results suggest that the CPET reactions from C-H bonds of substrates to 14 and 15 proceed through different transition states in terms of the proton position. In the case of 14, the proton transferred to the pterin ligand is rather close to the attached carbon atom; for 15, the proton is in the middle of the carbon atom of a substrate and nitrogen atom (Scheme 10).

This argument has been supported by DFT calculations on the transition states of CPET from indene to 14 and 15 (Figure 11); the C-H and N⋯H distances are calculated to be 1.27 and 1.47 Å, respectively, for 14 and 1.34 and 1.37 Å, respectively, for 15. Thus, it has been clarified that the transition state of CPET is governed by the proton acceptability and electron acceptability of a hydrogen-atom acceptor.40

5. Catalytic Water Oxidation by a Dinuclear CoIII Complex
Section:
Previous sectionNext section

As an important process in photosynthesis, photocatalytic water oxidation to evolve O2 has been investigated intensively using transition metal complexes, not only Ru complexes such as the “blue dimer” mentioned earlier10,43 but also basic metal complexes such as Mn complexes that are reminiscent of the natural oxygen evolving complex.7a,44 The 4e−-oxidation of water to O2 can be considered as a PCET reaction as follows:

\begin{equation*} 2\text{H$_{2}$O} \to \text{O}_{2} + 4\text{H}^{ + } + 4\text{e}^{ - } \end{equation*}

As a new candidate, we have prepared a bis-hydroxo-bridged dinuclear CoIII complex, [{CoIII(TPA)}2(μ-OH)2]4+ (16; Figure 12), as a water oxidation catalyst to evolve O2.45,46 The complex shows two-step deprotonation as represented by pKa1 = 6.8 and pKa2 = 10.9 in a Britton-Robinson (BR) buffer. In a cyclic voltammogram of 16 in a BR buffer solution, a catalytic current has been observed and the potential of the point at 0.1 mA shows pH dependence. A plot of the potential shows two kinks at pH values corresponding to the pKa values. On the basis of the gradients of the plot, it has been indicated that 16 undergoes a 2H+/2e− PCET process in the range of pH 5.0–7.5 showing a slope of −59 mV/pH, a 1H+/2e− PCET process in the range of pH 7.5–11.5 showing a slope of −27.5 mV/pH, and a 2e− ET-oxidation process in the range of pH >11.5 with no slope.45

Complex 16 has been applied to photocatalytic water oxidation using [RuII(bpy)3]2+ as a photosensitizer and Na2S2O8 as an oxidant under visible light irradiation. In this system, [RuIII(bpy)3]3+ is formed as a direct electron acceptor from the catalyst through photoinduced ET from the triplet excited state of [RuII(bpy)3]2+ to S2O82−. Catalytic turnover number has reached 742 and the quantum yield of O2 evolution is high enough to be 44%. The highest O2 yield based on the oxidant has been determined to be 72%.45

Mechanistic insights into water oxidation by 16 have been gained mainly to understand the O-O bond formation to propose a mechanism of water oxidation as shown in Scheme 11. The two-electron-oxidized species has been analyzed by DFT calculations at the B3LYP level of theory. The DFT-optimized structure should be described as a dinuclear bis(μ-oxyl) CoIII complex, [{CoIII(TPA)}2(μ-O•)2]4+ (17), having large spin densities on the two oxo bridge, rather than [{CoIV(TPA)}2(μ-O)2]4+. 18O-labelling experiments have been performed in 42% 18O-enriched BR buffer (pH 7.8) using [RuIII(bpy)3]3+ as an oxidant to elucidate the oxygen source of O2. Under the conditions, the isotropic distribution has been determined to be 16O2:16O18O:18O2 = 48:34:18; this result is fairly consistent with a calculated ratio of 16O2:16O18O:18O2 = 50:37:13 by assuming the first O2 molecule should be derived from the two bridging hydroxo ligands and intramolecular oxyl-oxyl coupling to form an O-O bond. Thus, we propose that the O-O bond formation occurs through intramolecular oxyl-oxyl coupling in 17, rather than intermolecular O-O bond formation through the reaction of 17 with outer-sphere water molecules. The oxyl-oxyl coupling affords a μ-peroxo dinuclear CoIII complex 18, which undergoes 2e−-oxidation to evolve O2 and to recover 16 as a catalyst. On the basis of the data obtained, we have proposed the reaction mechanism of photocatalytic water oxidation using 16 as a catalyst as depicted in Scheme 11.45,47

The μ-peroxo dinuclear CoIII complex, [{CoIII(TPA)}2(μ-peroxo)(μ-hydroxo)]3+ (18; Scheme 9),46,48 as a putative intermediate has been alternatively synthesized. We have found that two isomers are available for 18, i.e., cis-18 and trans-18, depending on the synthetic procedure, as shown in Figure 13. The complexes show two-step oxidation processes in CV measurements. ET oxidation of the two isomers has been investigated to shed light on the O2-evolving process as the final step of water oxidation.

1e−-oxidation of the isomers of 18 affords [{CoIII(TPA)}2(μ-superoxo)(μ-hydroxo)]4+, as characterized by ESR and resonance Raman spectroscopies. The dinuclear μ-superoxo CoIII complex can be further oxidized through PCET oxidation to form CoIIICoIV species, [{CoIII(TPA)}{CoIV(TPA)}(μ-oxo)(μ-superoxo)]5+, as suggested by DFT calculations. In the CoIIICoIV species, intramolecular ET from the O2•− ligand to the CoIV center generates O2 and recovers 16.48

6. Photocatalytic CO2 Reduction to CO by NiII Complexes
Section:
Previous sectionNext section

CO2 reduction to form CO can be also considered as an apparent PCET reaction as follows:

\begin{equation*} \text{CO}_{2} + 2\text{H}^{ + } + 2\text{e}^{ - } \to \text{CO} + \text{H$_{2}$O} \end{equation*}
In metal-catalyzed photochemical CO2 reduction to form CO, a metal complex as a catalyst is reduced by a reduced photosensitizer and the reduced complex reacts with CO2 to afford a metal-CO2 complex. Then the metal-CO2 complex accepts proton(s) from a proton donor at an oxygen atom of the CO2 ligand to cleave the C-O bond to form CO. In this context, CO formation from CO2 should be considered as a stepwise PCET reaction.

Photocatalytic CO2 reduction has been intensively investigated using Ru and Re complexes as catalysts.49 However, we have focused on Ni complexes as catalysts for CO2 reduction because carbon monoxide dehydrogenase (CODH) that catalyzes interconversion between CO2 and CO contains a Ni center embedded in an iron-sulfur cluster as a reactive site.50 Actually, Ni complexes have been used as catalysts for photocatalytic CO2 reduction for a long time.51 Recently, Chang and coworkers have reported on photocatalytic CO2 reduction to CO using NiII-NHC complexes as catalysts; their system shows excellent TON and selectivity in the reaction but in a very low quantum yield (0.01%).52 Inspired by the structure of the active site of CODH, we have prepared NiII complexes having N2S2-type coordination environments as catalysts.

In search of a good candidate for the catalyst, we have found that a NiII complex having S,S′-bis(2-pyridylmethyl)-1,2-dithiaethane (bpet) as a ligand exhibits a good performance in the photocatalytic CO2 reduction. The complex, [NiII(bpet)(CH3CN)2]2+ (19; Figure 14), shows a high TON more than 700 in 55 h, excellent selectivity in CO formation (>99% over H2 evolution), and a good quantum yield of 1.42% at 450 nm, in the presence of [RuII(bpy)3]2+ as a photosensitizer and BIH (Figure 14; right) as a reductant in a dimethylacetamide (DMA)/H2O (9:1, v/v) mixed solvent.53 Kinetic analysis of the reaction has revealed that coordination equilibrium between a Ni0 species formed by photoinduced ET from BIH and CO2 is operative to determine the efficiency of the reaction (Scheme 12). In addition, the use of a DMA/D2O (9:1, v/v) mixed solvent retards the CO formation, indicating the solvent KIE (1.2). This result suggests that a water molecule is involved in the RDS in the CO formation, probably through proton transfer to the Ni-bound CO2 ligand to remove one oxygen atom from the CO2 ligand. The CO2 reduction by 19 can be considered as an ET/PT reaction, in which the NiII center is an electron acceptor and a CO2 ligand is a proton acceptor. Note that complex 19 can also act as a catalyst for hydrogen evolution in ascorbate buffer (pH 4), through PCET reaction involving a putative NiII-hydrido intermediate.54

In order to improve the efficiency of photocatalytic CO2 reduction, stabilization of Ni-CO2 complex in the CO2 binding equilibrium (see Scheme 12) is assumed to be important. Thus, we have introduced two pyridine pendants to the 5-positions of the two pyridine rings of the bpet ligand as Lewis-acid binding sites. The ligand, bpet-py2, forms a NiII complex, [NiII(bpet-py2)(H2O)2]2+ (20).55 We have examined various Lewis acids to find that divalent alkaline-earth metal ions, such as Mg2+ and Ca2+, are effective to enhance CO production using 20 as a catalyst. In the presence of 150 eq of Mg2+ ion, [RuII(bpy)3]2+ as a photosensitizer, and BIH as a reductant, the quantum yield of photocatalytic CO production by 20 has reached 11.1%, which is 8-times higher than that of 19 in the absence of a Lewis acid, and the selectivity is 99.8% over H2 evolution.55 It should be noted that the efficiency of CO production is maintained even at the CO2 concentration of 5% in the gas phase, which should be derived from the stabilization of the Ni-CO2 species in the coordination equilibrium to increase its concentration for the enhancement of the following processes. A proposed mechanism of the photocatalytic CO2 reduction using 20 with Mg2+ is depicted in Scheme 13. DFT-optimized structure has been obtained for a Mg2+-bound NiII-CO2 complex derived from 20 as an intermediate (Figure 15). In the intermediate, the Mg2+ ion binds to the two oxygen atoms of the NiII-bound CO2 ligand to stabilize the intermediate. The choice of Lewis acid cocatalyst is very important for the enhancement of CO production: Monovalent alkaline metal ions such as Li+ and Na+ ions do not show any enhancement in the reaction; in sharp contrast, trivalent ions such as Sc3+ are extremely Lewis-acidic and thus activate coordinated water molecules to increase proton concentration for enhancement of H2 evolution rather than CO formation. The excellent performance of Mg2+-bound 20 should be related to the fact that RuBisco (ribulose bisphosphate carboxylase/oxygenase), which is a CO2 fixation enzyme, contains a Mg2+ ion at the reactive site.56

The Lewis-acid-assisted photocatalytic CO2 reduction using 20 would provide a new strategy to make progress in related fields.

7. Summary and Perspective
Section:
Previous section

We have briefly described our achievements in the development of functionality of transition metal complexes on the basis of PCET. Moreover, deeper mechanistic insights into HAT reactions are gained to understand the controlling factors in HAT. The formation of RuIV-oxo complexes has been achieved in water using ET oxidants such as CAN. The characterization of those species has been made to clarify the unique characteristics including low-spin RuIV=O complexes which are stabilized by hydrogen bonding of water molecules in water. Comparison of the reactivity of RuIV=O in different spin states (S = 0 or 1) allows us to conclude that the spin state of a RuIV=O complex is not determinant but the BDE of the O-H bond of the corresponding RuIII-OH complex should be important for the argument on the reactivity. We have also exemplified the “oxygen-rebound” mechanism in C-H hydroxylation, which has been proposed and accepted without direct evidence for a long time, by detecting oxygen-rebound intermediates in hydroxylation reactions of C-H bonds. As for the characteristics of RuIV=O complexes, the first determination of reorganization energies of ET and PCET has been made to clarify the kinetic advantage of PCET over ET in oxidation reactions by RuIV=O complexes. Through the PCET oxidation of a RuII-OH2 complex having a NHC moiety at the trans position of the aqua ligand, we have succeeded in the formation of the first RuIII-O• complex, which exhibits unique reactivity on the basis of its strong radical character, such as oxidative cracking of aromatic rings to afford formic acid as an H2 source. Using RuIII-pterin complexes as H-atom acceptor in PCET without oxo ligands, it has been clarified that the transition states of HAT are regulated by controlling the electron-acceptability and proton-acceptability of an H-atom acceptor.

PCET is also involved in water oxidation and CO2 reduction, which are important for the development of artificial photosynthesis, to facilitate those reactions kinetically and thermochemically. A dinuclear bis(μ-OH) CoIII-TPA complex can serve as an efficient water oxidation catalyst through the formation of a dinuclear bis(μ-oxyl) CoIII-TPA intermediate, in which the intramolecular O-O bond formation occurs to generate a μ-peroxo intermediate. In the CO2 reduction to form CO selectively, ET from a reduced photosensitizer to a NiII catalyst to make CO2 activation possible and the Ni-bound CO2 undergo PT from a water molecule to remove one oxygen atom to afford CO.

Since PCET is a ubiquitous and fundamental process, a wide range of redox reactions can be in placed in focus. Accumulating the understanding of PCET processes should be indispensable for further development of efficient redox reactions and catalysis using metal complexes to produce valuable chemicals and chemical energy for construction of a sustainable society.

The author appreciates his excellent coworkers including students and prominent experts who appear as co-authors in the references. He would like to share the Award for Creative Work of The Chemical Society of Japan in 2019 with his coworkers, especially Dr. Tomoya Ishizuka and Dr. Hiroaki Kotani (University of Tsukuba) as colleagues with respect to their great efforts to promote our research into the present shape. He also sincerely appreciates Prof. Shunichi Fukuzumi (Osaka University, Japan; at present, Ewha Womans University, Korea and Meijo University, Japan) for his strong support and encouragement that have been indispensable for achievements described in this paper. His sincere thanks go to Prof. Yoshihisa Matsuda (Kyushu University) and Prof. Hiroshi Kitagawa (Kyushu University; at present, Kyoto University) for kindly providing opportunities to start and perform research presented here. The works described herein have been supported by Japan Society of Promotion of Science (JSPS), MEXT, Japan, Japan Science and Technology Agency (JST), The Mitsubishi Foundation, Asahi Glass foundation, Yazaki Memorial Foundation for Science and Technology, The Iwatani Naoji Foundation, and Tsukuba Innovation Arena (TIA).

Takahiko Kojima

Takahiko Kojima was born in Nagoya, Japan, in 1962. He received Bachelor (1986) and Doctor of Engineering (1991) from the University of Tokyo. After working as a postdoc at University of Minnesota with Prof. Lawrence Que, Jr., he joined Kyushu University as a research associate in 1994, working with Prof. Yoshihisa Matsuda followed by Prof. Hiroshi Kitagawa. In 2005, he moved to Osaka University as an associate professor working with Prof. Shunichi Fukuzumi. Since 2008, he has been a professor of Department of Chemistry, University of Tsukuba. His research interest is mainly focused on redox reactivity of transition metal complexes to develop novel functionality.

<< Previous High Sensitive Electrochemical Nitrite Sensor Using Fe2O3/MoS2 Nanocomposites Synthesized by Facile Method Next >> Spin Crossover between the High-Spin and Low-Spin States and Dielectric Switching in the Ionic Crystals of a Fe(II) [2 × 2] Molecular Grid

Select Issues

▾
  • -

Search for Articles

Advanced

Saved Search

Please login to view your saved searches.

Top Accessed Articles

Most Cited

Related Articles

  1. Quinone–Hydroquinone π-Conjugated Redox Reaction Involving Proton-coupled Electron Transfer Plays an Important Role in Scavenging Superoxide by Polyphenolic Antioxidants
  2. Oxidation of Guanosine to the Imidazolone Derivative via Proton-coupled Electron Transfer to Hydroperoxy Radical Derived from Superoxide
  3. Oxidation by Transition Metal Complexes. I. Oxidation of Styrene and Triphenylphosphine Catalyzed by O2-Ir Complex
  4. Extracellular Electron Transport Scarcely Accumulates Proton Motive Force in Shewanella oneidensis MR-1
  5. Transition Metal Complexes of 1,1′-Bis(2-pyridyl)ferrocene. A Cationic Rhodium(I) and a Silver(I) Complex. Preparation and Structure
MANAGE YOUR ALERTS
RECOMMEND TO LIBRARY