Keywords

1 Introduction

Among clean, renewable alternatives to fossil fuels, hydrogen is one of the most attractive candidates for the future sustainable energy system, because hydrogen is produced via water splitting by directly using solar energy or by electrolysis with a solar cell [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Hydrogen is used to fix carbon dioxide and to furnish fuels, chemicals, and biomass [22,23,24,25]. A photovoltaic (PV) electrolysis of water has already achieved an average solar-to-hydrogen (STH) efficiency as high as 30% under continuous operation [26]. In industry, however, only 4% of hydrogen is currently produced by water electrolysis, whereas steam methane reforming and coal gasification make more than 95% of the whole hydrogen production, because the cost of hydrogen production by electrolysis is still much higher than that produced by fossil fuels [26]. The high cost of hydrogen production by electrolysis results from the use of precious metal such as platinum as the hydrogen evolution catalysts for electrolysis of water [26]. In order to develop more efficient catalysts composed of earth-abundant metals for the hydrogen evolution, it is highly desired to clarify the molecular mechanism of the catalytic hydrogen evolution. In general, the catalytic hydrogen evolution consists of several components as shown in Fig. 46.1, where D = sacrificial electron donor, P = photocatalyst (photosensitizer), M = electron mediator or relay, and Cat. = hydrogen evolving catalyst. Although both photocatalyst and photosensitizer have been used to describe molecules, which participate in light-driven chemical processes without being consumed, the term “photocatalyst” is frequently used to avoid the usage of term “photosensitizer,” which is normally used to describe a molecule that participates in energy transfer processes [27,28,29,30,31,32,33]. Thus, the term “photocatalyst” instead of “photosensitizer” is used in this chapter. Photocatalysts have been extensively varied ranging from [RuII(bpy)3]2+ [34,35,36,37] to Zn(II) porphyrins [37,38,39], cyclometalated Ir(III) complexes [40,41,42], organic dyes [43, 44], Pt(II) terpyridyl complexes [45, 46], Re(I) complexes [47], and Cu(I) complexes [48,49,50,51]. Various sacrificial electron donors such as aliphatic and aromatic amines, 1-benzyl-1,4-dihydronicotinamide (BNAH), 1,4-dihydronictinamide adenine dinucleotide (NADH), dimethylphenylbenzimidazoline (BIH), ascorbic acid, oxalate, and thiols have been used for photocatalytic hydrogen evolution [52, 53]. Hydrogen evolving catalysts have also been varied ranging from a colloidal Pt to hydrogenase, hydrogenase model complexes [54,55,56,57], iron complexes [58, 59], Ni(II) complexes [60,61,62,63,64], Pt(II) complexes [12, 65, 66], Co(III) complexes [67,68,69,70], and Rh(III) complexes [71,72,73,74]. This chapter focuses on molecular mechanisms of photocatalytic hydrogen evolution to clarify how photons are converted to two electrons that are required for hydrogen production from two protons.

Fig. 46.1
scheme 1

A generalized three component photocatalytic system for hydrogen production where D sacrificial electron donor, P photocatalyst, M electron mediator or relay, and Cat. hydrogen evolving catalyst

Full size image

2 Disproportionation Following Photoinduced Electron Transfer

How photoinduced electron transfer of a photocatalyst (a one-electron process) leads to hydrogen evolution (a two-electron process) was clarified for photocatalytic hydrogen evolution with water-soluble transition-metal complexes [RhIII(Cp*)(bpy)(H2O)](SO4) (1: Cp* = η5-C5Me5, bpy = 2,2′-bipyridine) [75] and [IrIII(Cp*)(H2O)(bpm)RuII(bpy)2](SO4)2 (2: bpm = 2,2′-bipyrimidine) [76], as proton reduction catalysts, [RuII(bpy)3]2+ as a photocatalyst and sodium ascorbate (HA−) as a sacrificial electron donor [77]. Upon photoexcitation of an aqueous solution containing [RuII(bpy)3]2+, HA− and [RhIII(Cp*)(bpy)]2+, electron transfer from HA− to [RuII(bpy)3]2+ occurred to produce HA• and [RuI(bpy)3]+, which were detected as transient absorption spectra [77]. The decay of [RuI(bpy)3]+ (λmax = 510 nm) was accompanied by formation of [RhII(Cp*)(bpy)]+ (λmax = 750 nm), indicating that electron transfer from [Ru(bpy)3]+ to [RhIII(Cp*)(bpy)]2+ occurred to produce [RhII(Cp*)(bpy)]+, accompanied by regeneration of [RuII(bpy)3]2+ (Fig. 46.2a) [77]. The decay of transient absorption spectrum at 510 nm due to [RuI(bpy)3]+ obeyed pseudo-first-order kinetics, and the pseudo-first-order rate constant increases linearly with concentration of 1(SO4) and the rate constant of electron transfer from [Ru(bpy)3]+ to 1(SO4) was determined to be 1.3 × 109 M−1 s−1 [77]. In contrast, the decay of [RhII(Cp*)(bpy)]+ (λmax = 750 nm) obeyed second-order kinetics (Fig. 46.2b), which indicates that disproportionation of [RhII(Cp*)(bpy)]+ occurred to produce [RhIII(Cp*)(bpy)]2+ and the RhI complex, [RhI(Cp*)(bpy)] [77]. The resulting RhI complex is immediately protonated at pH 3.6 to produce the Rh(III)-hydride complex, ([RhIII(Cp*)(H)(bpy)]+), which reacts with a proton to generate hydrogen, accompanied by regeneration of [RhIII(Cp*)(bpy)]2+ [77]. Thus, photoinduced electron transfer from HA− to [RuII(bpy)3]2+* (a one-electron process) leads to the two-electron reduction of protons to produce H2 by disproportionation of the one-electron-reduced species of [RhIII(Cp*)(bpy)]2+ to afford the two-electron-reduced metal complexes ([RhI(Cp*)(bpy)]), which is protonated to give the metal hydride complex ([RhIII(Cp*)(H)(bpy)]+), as shown in Fig. 46.3 [77]. H2 is produced by the reaction of [RhIII(Cp*)(H)(bpy)]+ with H+ to regenerate [RhIII(Cp*)(bpy)]2+ [77]. Such a disproportionation of Rh(II) complexes has been well established for Rh(II) porphyrins to produce Rh(III) and Rh(I) porphyrins [78]. Rh(I) porphyrins react with proton to afford Rh(III)-hydride porphyrins that react further with proton to produce hydrogen and Rh(III) porphyrins [79]. It was also reported that [RhI(dmbpy)2]+ (dmbpy = 4,4′-dimethyl-2,2′-bipyridine) reacts with protons to form a Rh(III) hydride intermediate that can, in turn, release H2 [79, 80].

Fig. 46.2
figure 2

(a) Transient absorption spectra of 1 (1.6 × 10−4 M), [RuII(bpy)3]2+ (8.0 × 10−5 M), H2A (0.8 M), and NaHA (0.3 M) in deaerated H2O at pH 3.6 at 298 K after laser excitation at 455 nm. (b) Decay time profile of absorbance at 750 nm due to [RhII(Cp*)(bpy)]+. Inset shows second-order plot of 1/ΔAbs versus time. (Reprinted with permission from Ref. [77]. Copyright 2011, Wiley-VCH)

Full size image
Fig. 46.3
scheme 3

Photocatalytic mechanism for hydrogen evolution form ascorbate (HA−) with [RuII(bpy)3]2+ as a photocatalyst and [RhIII(Cp*)(bpy)(H2O)](SO4) (1) or [IrIII(Cp*)(H2O)(bpm)RuII(bpy)2](SO4)2 (2) as a hydrogen evolution catalyst via disproportionation of the one-electron reduced species of 1 or 2. (Reprinted with permission from Ref. [77]. Copyright 2011, Wiley-VCH)

Full size image

When [IrIII(Cp*)(H2O)(bpm)RuII(bpy)2]3+ (2) was employed as a hydrogen evolution catalyst, photoinduced electron transfer from HA− to [RuII(bpy)3]2+ was also followed by electron transfer from [RuI(bpy)3]+ to [IrIII(Cp*)(H2O)(bpm)RuII(bpy)2]3+ and disproportionation of [IrII(Cp*)(H2O)(bpm)RuII(bpy)2]3+ to produce [IrI(Cp*)(H2O)(bpm)RuII(bpy)2]2+ that reacts with protons to release H2 [77]. The maximum quantum yield of the photocatalytic hydrogen evolution with 2 (0.015) was obtained at pH 3.6. This is because HA− acts as an electron donor at pH < 4.0 and the Ir(III)-hydride complex ([IrIII(Cp*)(H)(bpm)RuII(bpy)2]3+) is deprotonated at pH >4.0 to produce the low-valent iridium complex [IrI(Cp*)(H2O)(bpm)RuII(bpy)2]2+, which has no catalytic activity for hydrogen evolution [77].

3 Photoinduced Electron Transfer Combined with Thermal Electron Transfer

A one-photon two-electron process was made possible in photocatalytic H2 evolution with ascorbate (HA−) and a cobalt(II) chlorin complex (CoII(Ch)) via electron transfer from ascorbate to the excited state of [RuII(bpy)3]2+ followed by electron transfer from [RuI(bpy)3]+ to [CoII(Ch)] with proton to give the hydride complex (CoIII(H)(Ch)), which reacts with proton to produce H2 as shown in Fig. 46.4. [CoIII(Ch)]+ produced by the reaction of (CoIII(H)(Ch)) with proton was reduced by ascorbate to regenerate CoII(Ch) to complete the catalytic cycle [81]. Thus, once one photon is used to produce [CoI(Ch)]− from [CoII(Ch)] and another electron is provided by HA− to regenerate [CoII(Ch)] from [CoIII(Ch)]+. Photoexcitation of [RuII(bpy)3]2+ resulted in electron transfer from AH− to [RuII(bpy)3]2+* to produce [RuI(bpy)3]+, followed by electron transfer from [RuI(bpy)3]+ to CoII(Ch) to produce [CoI(Ch)]−, which reacts with AH2 to produce [CoIII(H)(Ch)(AH)]− [81]. The rate constant of electron transfer from [RuI(bpy)3]+ to CoII(Ch) was determined to be 2.5 × 109 M−1 s−1. Hydrogen is produced by the reaction of [CoIII(H)(Ch(AH))]− with AH2 via the Co–H bond heterolysis to produce [CoIII(Ch)(AH)], which is reduced by AH− to regenerate CoII(Ch) [81]. The rate constant of electron transfer from AH− to [CoIII(Ch)]+ that was prepared by the electron-transfer oxidation of CoII(Ch) by (p-BrC6H4)3N•+SbCl6− in H2O/MeCN was determined to be 1.5 × 103 M−1 s−1. The Co-H bond heterolysis is involved in the rate-determining step, because the kinetic isotope effect (kH/kD = 1.8) was observed for the photocatalytic hydrogen evolution, when H2O was replaced by D2O [81]. Such heterolytic cleavage of the Co-H bond by proton affords H2 [82].

Fig. 46.4
scheme 4

Mechanism of photocatalytic hydrogen evolution from ascorbate (AH−) and ascorbic acid (AH2) with [RuII(bpy)3]2+ and CoII(Ch). (Reprinted with permission from Ref. [81]. Copyright 2015, Royal Society of Chemistry)

Full size image

Virtually the same mechanism as Fig. 46.4 is applied to the photocatalytic H2 evolution from AH− with [RuII(bpy)3]2+ and a water-soluble cobaltous meso-tetrakis (p-sulfonylphenyl)porphyrin complex (CoIITPPS) [83]. The best quantum yield of photocatalytic H2 evolution was obtained as 56% using a cobalt(II) tripodal iminopyridine complex [CoII(tachpy3)](ClO4)2 (tachpy3 = cis,cis-1,3,5-tris(pyridine-2- carboxaldimino)cyclohexane) as a catalyst, a cyclometallated Ir complex as photosensitizer and triethylamine as a sacrificial electron donor in aqueous acetonitrile [84] .

4 Photoinduced Electron Transfer Followed by Proton and Electron Transfer

A highly efficient photocatalytic hydrogen-evolution system has been constructed using 9-mesityl-10-methylacridinium ion (Acr+–Mes) [85], poly(N-vinyl-2- pyrrolidone)-protected platinum nanoclusters (Pt-PVP), and NADH (1,4-nicotinamide adenine dinucleotide), used as an organic photoredox catalyst, a hydrogen-evolution catalyst, and an electron donor, respectively [86]. Photoexcitation of Acr+–Mes results in efficient electron transfer from the Mes moiety to the singlet excited state of the Acr+ moiety to produce the electron-transfer (ET) state, Acr•–Mes•+, which has the strong oxidizing ability of the Mes•+ moiety with Ered vs. SCE = 2.06 V and the strong reducing ability of the Acr• moiety with Eox vs. SCE = −0.57 V [87,88,89]. NADH is oxidized by electron transfer from the Mes•+ moiety of Acr•–Mes•+ to produce NADH•+ and Acr•–Mes as shown in Fig. 46.5, where NADH•+ undergoes deprotonation to produce NAD• that has the strong reducing ability with Ered vs. SCE = −1.1 V [90]. The formation of Acr•–Mes is clearly seen as the transient absorption at λmax = 520 nm in Fig. 46.6a, where the increase in bleaching results from a decrease in absorption at 420 nm due to NAD•, accompanied by increase in absorbance at 520 nm due to Acr•–Mes (Fig. 46.6b) [86]. This indicates that electron transfer from NAD• to Acr+–Mes occurs to produce NAD+ and Acr•–Mes. The rate constant of electron transfer from NAD• to Acr+–Mes was determined to be 3.7 × 109 M−1 s−1, which is close to the diffusion-limited value, as expected from the large driving force (0.53 eV) of electron transfer from NAD• (Eox = −1.1 V vs. SCE) [90] to the Acr+ moiety in Acr+–Mes (Ered = −0.57 V vs. SCE) [86]. Thus, one photon used to excite Acr+–Mes is converted to two electrons that are used to reduce two equivalents of Acr+–Mes to produce two equivalents of Acr•-Mes (Fig. 46.5) [86]. The quantum yield for the formation of Acr•–Mes was determined to be 0.52 from the absorbance at 520 nm due to the Acr• moiety [86].

Fig. 46.5
scheme 5

One-photon two-electron processes started from photoinduced electron transfer from the Mes moiety to the singlet excited state of Acr+-moiety in Acr+–Mes as an organic photoredox catalyst and NADH as a source of two electrons and protons. (Reprinted with permission from Ref. [86]. Copyright 2007, Royal Society of Chemistry)

Full size image
Fig. 46.6
figure 6

(a) Transient absorption spectra of Acr+–Mes (0.10 mM) and NADH (1.0 mM) in deaerated H2O and MeCN (v/v 1:1) solution mixture (2.0 cm3) at 298 K taken at 1.0 μs (○) and 10 μs (●) after nanosecond laser excitation at 430 nm. (b) Time profiles of formation of Acr•–Mes at 520 nm and decay of NAD• at 420 nm. (Reprinted with permission from Ref. [86]. Copyright 2007, Royal Society of Chemistry)

Full size image

Electron injection from Acr•-Mes produced by the photochemical reduction of Acr+–Mes to Pt-PVP (PVP = poly(vinylpyrrolidone)) with protons resulted in generation of 0.5 equivalent of hydrogen (Eq. 46.1) as shown in Fig. 46.7 [86]. The hydrogen-evolution rate agrees with the rate of formation of Acr+–Mes (Fig. 46.8) [91]. This indicates that electron transfer from Acr•–Mes to Pt-PVP is the rate-determining step of the hydrogen-evolution reaction [91].

$$ 2{\mathrm{Acr}}^{\bullet}\hbox{--} \mathrm{Mes}+2{\mathrm{H}}^{+}\underset{\mathrm{Pt}\hbox{-} \mathrm{PVP}}{\to }2{\mathrm{Acr}}^{+}\hbox{--} \mathrm{Mes}+{\mathrm{H}}_2 $$
(46.1)
Fig. 46.7
figure 7

Comparison of the amount of evolved hydrogen and Acr•–Mes after laser excitation (λ = 430 nm) of a deaerated mixed solution (2.0 cm3) of phthalic acid buffer (pH 4.5; 50 mM) and MeCN [1:1 (v/v)] containing Acr+–Mes (0.10 mM), NADH (2.0 mM) and Pt–PVP (0.10 mg cm−3) at 298 K. (Reprinted with permission from Ref. [86]. Copyright 2007, Royal Society of Chemistry)

Full size image
Fig. 46.8
figure 8

(a) Decay time profile (black circles) of absorbance at 520 nm due to Acr•–Mes in electron transfer from Acr•–Mes to Pt-PVP (0.1 μg) in a (pH 5.0, 50 mM) CH3COOH/CH3COONa buffer and MeCN (v/v 1:1) solution mixture. Red circles show time course of hydrogen evolution. (b) Dependence of ket on [H+] (black circles) and [D+] (red circles) observed in electron transfer from Acr•–Mes to Pt-PVP in H2O/MeCN (v/v 1:1) containing CH3COOH/CH3COONa buffer (50 mM) and in D2O/MeCN (v/v 1:1) containing CH3COOD/CH3COONa buffer (50 mM) at 298 K, respectively. (Reprinted with permission from Ref. [91]. Copyright 2010, Wiley-VCH)

Full size image

When a CH3COOH/CH3COONa buffer (pH 4.5, 50 mM) in H2O is replaced by a CH3COOD/CH3COONa buffer in D2O, a substantial inverse kinetic isotope effect (KIE = ket(H)/ket(D) = 0.47) is observed in the rate constant of electron transfer from Acr•–Mes to Pt-PVP (red circles for D+ and black circles for H+ in Fig. 46.8b) [91]. Such an inverse kinetic isotope effect results from the higher zero point energy of the Pt–H bond formation than the Pt–D bond formation on Pt-PVP [91]. Because the ET rate increases linearly with increasing [H+] and [D+] (Fig. 46.8), ET from Acr•–Mes to PtNPs is coupled with proton transfer (PT) and the proton-coupled electron transfer (PCET) results in formation of a Pt–H bond on the Pt-PVP surface as shown in Fig. 46.9, where H2 is evolved by reductive elimination from two Pt–H species [91].

Fig. 46.9
scheme 9

Hydrogen evolution via PCET from Acr•–Mes to Pt-PVP. (Reprinted with permission from Ref. [91]. Copyright 2010, Wiley-VCH)

Full size image

Pt nanoparticle catalysts can be replaced by non-Pt metal nanoparticles (MNPs: M = Ru or Ni) for the photocatalytic hydrogen evolution with 2-phenyl-4-(1-naphtyl)- quinolinium ion (QuPh+–NA ) as an organic photocatalyst [92] and NADH as a sacrificial electron donor as shown in Fig. 46.10 [93, 94]. Electron transfer from the photogenerated QuPh•–NA to MNPs results in hydrogen evolution even under basic conditions (pH 10) [93, 94]. The catalytic activity of RuNPs is as high as that of PtNPs in the photocatalytic hydrogen evolution [93]. The hydrogen evolution rate with the most active Ni nanoparticles (NiNPs; hexagonal close-packed structure, 6.6 nm) examined reached 40% of that with commercially available Pt nanoparticles (2 nm) using the same catalyst weight [94]. In this case as well, a one-photon two-electron process occurs via photoinduced electron transfer from NADH to QuPh•–NA•+, followed by electron transfer from NAD•, which is formed by deprotonation of NADH•+, to QuPh+–NA to produce two equivalents of QuPh•–NA that inject two electrons to MNPs (M = Ru and Ni) with two protons to evolve H2 (Fig. 46.10) [93, 94]. The rate constants of electron transfer from NADH to QuPh•–NA•+ and from NAD• to QuPh+–NA were determined to be 5.7 × 109 M−1 s−1 and 2.5 × 109 M−1 s−1, respectively. The more negative one-electron reduction potential of the QuPh+ moiety in QuPh+–NA (Ered = −0.90 V vs. SCE) [92] than that of the Acr+ moiety in Acr+–Mes (Ered = −0.57 V vs. SCE) [86] resulted in efficient H2 evolution even under basic conditions such as pH 10 [93].

Fig. 46.10
scheme 10

Structure of QuPh+–NA (a) and the overall catalytic cycle (b) for the photocatalytic hydrogen evolution with NADH and QuPh+–NA and metal nanoparticles (MNPs, M = Ru and Ni). (Reprinted with permission from Ref. [94]. Copyright 2012, Royal Society of Chemistry)

Full size image

5 Photoinduced Electron Transfer Followed by Bond Cleavage

Oxalate is also used as a two-electron donor for photocatalytic H2 generation with QuPh+–NA and Pt-PVP [95]. As the case of oxidation of NADH by Acr•–Mes•+ (Fig. 46.5), a one photon-two electron process occurs via photoinduced electron transfer from (COO)22− to QuPh•–NA•+ to produce (COO)2•– and QuPh•–NA (Fig. 46.11) [95]. The C–C bond cleavage of (COO)2•– occurs rapidly to produce CO2 and CO2•– that reduced another molecule of QuPh+–NA to produce QuPh•–NA (Fig. 46.11) [95]. Thus, two equivalents of QuPh•–NA obtained by the ET reduction of QuPh+–NA with two electrons released from (COO)22− (Eq. 46.2) are used to produce H2 with metal nanoparticles (MNPs) as shown in Fig. 46.12 [95]. The maximum turnover number of H2 based on QuPh+–NA reached more than 260 [95].

Fig. 46.11
scheme 11

One-photon two-electron processes started from photoinduced electron transfer from the NA moiety to the singlet excited state of QuPh+-moiety in QuPh+–NA as an organic photoredox catalyst and oxalate as a source of two electrons. (Reprinted with permission from Ref. [95]. Copyright 2012, Royal Society of Chemistry)

Full size image
Fig. 46.12
scheme 12

The overall catalytic cycle for the photocatalytic hydrogen evolution with oxalate and QuPh+–NA and metal nanoparticles (MNPs, M = Ru, Ni and Pt). (Reprinted with permission from Ref. [95]. Copyright 2012, Royal Society of Chemistry)

Full size image
$$ {{\left(\mathrm{COO}\right)}_2}^{2\hbox{--}}\to {2\mathrm{CO}}_2+2{e}^{\hbox{--} } $$
(46.2)

A similar one-photon two-electron process occurs for photocatalytic oxidation of water using persulfate (S2O82−) as a sacrificial electron acceptor [96]. Electron transfer reduction of S2O82− results in S–S bond cleavage to produce SO42− and SO4•– that act as a much stronger electron acceptor than S2O82−, when S2O82− acts as a two-electron oxidant (Eq. 46.3) [96]. Thus, the maximum quantum efficiency for photocatalytic oxidation of water using S2O82− as a sacrificial electron acceptor is 200% via a one photon-two electron process. In fact, incorporation of a small amount of Ca2+ ions into a polymeric cobalt cyanide complex to form Cax[CoII(H2O)2]1.5-x[CoIII(CN)6] resulted in a significant enhancement of activity for photocatalytic water oxidation in a buffer solution (pH 7.0) containing [RuII(bpy)3]2+ as a photocatalyst and Na2S2O8 as an electron acceptor to achieve a quantum efficiency of 200% [97].

$$ {\mathrm{S}}_2{{\mathrm{O}}_8}^{2\hbox{--}}\to {{2\mathrm{SO}}_4}^{2\hbox{--}}\hbox{--} 2{e}^{\hbox{--} } $$
(46.3)

6 Photoinduced Electron Transfer Followed by Bond Formation

A one photon-two electron process also occurs via photoinduced electron transfer followed by bond formation. For example, photoinduced electron transfer from benzene to the singlet excited state of a photocatalyst (e.g., 3-cyano-1-methylquinolinium ion; QuCN+) occurs to produce benzene radical cation and QuCN•, followed by the addition of a nucleophile (X− = OH− and NHR−) to benzene radical cation to give the adduct (C6H6X•) as shown in Fig. 46.13 [98]. In the presence of CoII(dmgBF2)2 (dmg = dimethylglyoximate), electron transfer from C6H6X• to CoII(dmgBF2)2 occurs to produce C6H5X after deprotonation and CoI(dmgBF2)2 that reacts with two protons to produce H2 to generate [CoIII(dmgBF2)2]+ [98]. Electron transfer from QuCN• to [CoIII(dmgBF2)2]+ occurs to regenerate QuCN+ and CoII(dmgBF2)2 to complete the photocatalytic cycle in Fig. 46.13 [98]. Photoinduced electron transfer from benzene to 1QuCN+* is energetically feasible, because the one-electron reduction potential of 1QuCN+* (Ered vs. SCE = 2.72 V) [99] is more positive than the one-electron oxidation potential of benzene (Eox vs. SCE = 2.32 V) [100]. The occurrence of electron transfer from benzene to 1QuCN+* was confirmed by laser-induced transient absorption spectra as shown in Fig. 46.14a, where the transient absorption bands due to quinolinyl radical (QuCN•) and benzene dimer radical cation were observed at 520 nm and in the near-IR region, respectively. It is known that benzene radical cation is converted to benzene dimer radical cation by π-π association of benzene radical cation with large excess of benzene with the formation constant of 12 M−1 [101]. The rate constant of electron transfer from benzene to 1QuCN+* to produce benzene dimer radical cation was determined from the linear plot in Fig. 46.14b to be 2.1 × 1010 M−1 s−1, which is close to be the diffusion limited value in MeCN as expected from the exergonic electron transfer [102].

Fig. 46.13
scheme 13

The overall catalytic cycle for the photocatalytic hydrogen evolution with benzene , H-X (X− = OH− and NHR−), QuCN+, and CoII(dmgBF2)2 . (Reprinted with permission from Ref. [98]. Copyright 2016, American Chemical Society)

Full size image
Fig. 46.14
figure 14

(a) Transient absorption spectra of QuCN+ in the absence (red dots) and presence (1.5 M, blue dots) of benzene in deaerated MeCN taken at 200 ps after femtosecond laser excitation (λex = 355 nm). (b) Plot of the observed rate constant (kobs) of formation of benzene dimer radical cation vs. [C6H6]. (Reprinted with permission from Ref. [102]. Copyright 2001, American Chemical Society)

Full size image

Acr+-Mes in Fig. 46.5 can also be applied to photocatalytic generation of H2, accompanied by dehydrogenative oxygenation of an alkene with CoII(dmgH)2py [103]. The photocatalytic reaction is started by photoexcitation of Acr+–Mes to produce the ET state, Acr•–Mes•+, followed by electron transfer from alkenes to the Mes•+ moiety of Acr•–Mes•+ to produce the alkene radical cation and Acr•–Mes. The alkene radical cation reacts with H2O to produce the OH adduct radical after the deprotonation. The OH adduct radical is oxidized by CoIII(dmgH)2pyCl (or Acr•–Mes•+) to produce the carbonyl compound after the deprotonation. Acr•–Mes can reduce CoII(dmgH)2py to [CoI(dmgH)2py]− that reacts with H+ to produce the Co(III)-hydride species, followed by the reaction with H+ to produce H2 and CoIII(dmgH)2pyCl [103]. Based on the catalytic cycle in Fig. 46.15, a number of alkenes are oxygenated by water (Eq. 46.4) in the presence of Acr+–Mes and CoII(dmgH)2py as a photocatalyst and a hydrogen evolution catalyst, respectively [103]. This dual catalytic system possesses the single anti-Markovnikov selectivity due to the property of the visible light-induced alkene radical cation intermediate [103].

$$ {\mathrm{R}}_1{\mathrm{R}}_2\mathrm{C}=\mathrm{C}\left(\mathrm{H}\right){\mathrm{R}}_3+{\mathrm{H}}_2\mathrm{O}\underset{{\mathrm{Acr}}^{+}\hbox{--} \mathrm{Mes},{\mathrm{Co}}^{\mathrm{II}}{\left(\mathrm{dmgH}\right)}_2\mathrm{py}}{\overset{hv}{\to }}{\mathrm{R}}_1{\mathrm{R}}_2\mathrm{C}\left(\mathrm{CO}\right){\mathrm{R}}_3+{\mathrm{H}}_2 $$
(46.4)
Fig. 46.15
scheme 15

The overall catalytic cycle for the photocatalytic hydrogen evolution coupled by oxygenation of alkenes with H2O in the presence of Acr+–Mes and CoII(dmgH)2py. (Reprinted with permission from Ref. [103]. Copyright 2016, American Chemical Society)

Full size image

In Fig. 46.15, the alkene radical cation reacts with a nucleophile (H2O) to undergo C-O bond formation following photoinduced electron transfer from the alkene to Acr•–Mes•+. In the case of styrene, styrene radical cation undergoes the C-C bond formation by the dimerization with neutral syrene to produce the dimer radical cation following photoinduced electron transfer from the alkene to Acr•–Mes•+ as shown in Fig. 46.16 [104]. The syrene dimer radical cation undergoes deprotonation to produce the neutral radical that is oxidized by the Co(II) complex to yield the dimerized product after deprotonation [104]. The maximum yield of 1-phenyl-1,2-dihydronaphthalene was 72% in the presence of NaH2PO4 that acts as a good proton acceptor. Other photocatalysts such as rose bengal, rhodamine B, [Ir(dtbbpy)(ppy)2][PF6], eosin Y, and [Ru(bpy)3]Cl2 did not afford the desired product at all. In the control experiments, no desired product was observed without the organic redox catalyst, Acr+–Mes, cobalt catalyst, or visible light, indicating that Acr+–Mes and cobalt catalysts are essential for this transformation [104].

Fig. 46.16
scheme 16

The overall catalytic cycle for the photocatalytic hydrogen evolution with dimerization of styrenes in the presence of Acr+–Mes and CoII(dmgH)2py. (Reprinted with permission from Ref. [104]. Copyright 2018, Elsevier)

Full size image

A selective C(sp2)-H amination of arenes (alkyl-substituted benzenes, biphenyl, and anisole derivatives) accompanied by hydrogen evolution by using heterocyclic azoles as nitrogen sources has also been made possible by using Acr+-Mes/cobalt catalyst as shown in Fig. 46.17 [105]. Methylarenes such as p-xylene can be oxidized by electron transfer to the Mes•+ moiety of Acr•–Mes•+ produced upon photoexcitation of Acr+–Mes to produce p-xylene radical cation, which is attacked by nucleophile pyrazole (Eox = 2.27 V vs. SCE), which is difficult to be oxidized by Acr•–Mes•+ (Ered = 2.06 V vs. SCE), to produce the radical adduct. Then, the radical adduct is oxidized by electron transfer to Co(II) catalyst, which quickly loses one proton and generates the amination product. At the same time, Co(I) could capture the present proton to produce Co(III)-H that reacts with H+ to release H2, accompanied by regeneration of the Co(III) complex.

Fig. 46.17
scheme 17

The overall catalytic cycle for the photocatalytic hydrogen evolution with cross-coupling between simple arenes and heterocyclic amines in the presence of Acr+–Mes and CoII(dmgH)2py. (Reprinted with permission from Ref. [105]. Copyright 2017, Springer Nature)

Full size image

7 One Photon-Two Electron Excitation

Photoexcitation of [Cp*IrIII(bpy)H]+ results in a one photon-two electron process to produce [Cp*IrI(bpy)] and H+ [106]. Before releasing H+, the bimolecular reaction of the excited state of [Cp*IrIII(bpy)H]+, which is equivalent to the [Cp*IrI(bpy)]/H+ pair, with [Cp*IrIII(bpy)H]+ occurs to produce H2, [Cp*IrIII(bpy)(CH3CN)]2+, and [Cp*IrI(bpy)] as shown in Fig. 46.18, where [Cp*IrI(bpy)] reacts with H+ to regenerate [Cp*IrIII(bpy)H]+ [107]. The quantum yield of photochemical H2 production from [Cp*IrIII(bpy)H]+ increased with an increase in concentration of [Cp*IrIII(bpy)H]+ to reach 100% [100]. [Cp*IrIII(bpy)H]+ may act as an efficient photoredox catalyst for H2 generation via a one photon-two electron process.

Fig. 46.18
scheme 18

Proposed mechanism of photochemical production of H2 from [Cp*IrIII(bpy)H]+. (Reprinted with permission from Ref. [107]. Copyright 2016, American Chemical Society)

Full size image

8 Conclusion, Challenge, and Future Perspective

In order to produce hydrogen from two protons, two photons are generally required by conversion of two photons to two electrons. However, as shown in this chapter, only one photon is enough to produce hydrogen via photoinduced electron transfer followed by disproportionation of a redox catalyst to produce two-electron reduced species. Photoinduced electron transfer can be combined with thermal electron transfer to provide two electrons from one photon, leading to hydrogen production from protons. Depending on a type of electron donor substrate, the substrate radical cation produced by photoinduced electron transfer to a photoredox catalyst undergoes bond cleavage or bond formation to produce a much stronger electron donor to provide the second electron for hydrogen production from protons. Photoredox catalysts normally provide only one electron via photoinduced electron transfer of the photoexcited state. In the case of [Cp*IrIII(bpy)H]+, however, the photoexcitation results in a one photon-two electron process to produce hydrogen via self-quenching with 100% quantum yield.

In the natural photosynthesis, NADP+ is reduced by plastoquinol to produce NADPH in photosystem I [108, 109]. It is desired to develop efficient photocatalytic systems for hydrogen evolution from plastoquinol analogs to mimic the function of photosystem I by optimizing the conditions where photons are used to produce hydrogen as discussed in this chapter. On the other hand, water is oxidized by plastoquinone to produce O2 and plastoquinol in photosystem II [110]. A molecular photocatalytic system to mimic the function of photosystem II has recently been reported to oxidize water by plastoquinone analogs to produce O2 and plastoquinol analogs [111]. Thus, combination of such a molecular photocatalytic system to mimic photosystem II with a molecular photocatalytic system to mimic photosystem I may enable to produce hydrogen from water more efficiently than the natural system in future. The biggest challenge to use a molecular photocatalyst for H2 evolution is the photostability because organic compounds are susceptible to the photoredox reactions, which cause degradation of the photocatalyst. Incorporation of a molecular photocatalyst into mesoporous silica alumina is reported to improve the photostability, because the immobilization of the catalyst prohibits the intermolecular reaction, which is the main reason of the photodegradation [112]. Construction of such immobilized molecular photocatalysts may provide more practical applications of molecular photocatalysts for hydrogen evolution.