Mechanism of a one-photon two-electron process in photocatalytic hydrogen evolution from ascorbic acid with a cobalt chlorin complex

Abstract

A one-photon two-electron process was made possible in photocatalytic H₂ evolution from ascorbic acid with a cobalt(II) chlorin complex [Co^II(Ch)] via electron transfer from ascorbate to the excited state of [Ru(bpy)₃]²⁺ followed by electron transfer from [Ru(bpy)₃]⁺ to Co^II(Ch) with proton to give the hydride complex, which reacts with proton to produce H₂. [Co^III(Ch)]⁺ was reduced by ascorbate to reproduce Co^II(Ch).

---

Photocatalytic production of hydrogen (H₂) has attracted increasing attention as a clean energy source because of the ever-increasing demand for energy and climate change on our planet.¹ A number of highly efficient hydrogen evolving systems have been developed including homogeneous and heterogeneous photocatalytic systems.^2–13 Two electrons are required to produce H₂ from protons, although one photon generates normally only one electron. A mechanism of photocatalytic production of H₂ was reported to clarify how photoinduced electron transfer of a photosensitiser (a one-electron process) leads to H₂ production (a two-electron process).^14–16 Disproportionation of one-electron reduced species of metal complexes resulted in formation of the two-electron reduced species from which H₂ is formed.¹⁷ Bimolecular reactions of metal(III)–hydride complexes also generate H₂ accompanied by regeneration of metal(II) complexes.¹⁸ In each case, the maximum quantum yield of H₂ production per photon is 50%, because two photons are required to produce two electrons. Thus there has so far been no example for one photon to generate one H₂ molecule.

We report herein photocatalytic H₂ evolution from ascorbic acid (AscH₂) with a cobalt(II) chlorin complex [Co^II(Ch)] (a chemical structure shown in Scheme 1)¹⁹ in an aqueous acetonitrile solution (H₂O/MeCN), which proceeds via a one-photon two-electron process. The photocatalytic mechanism is clarified by nanosecond laser transient absorption spectra and by examining each step in the catalytic cycle independently.

Visible light irradiation of a deaerated (Ar-saturated) H₂O/MeCN solution (1 : 1 v/v) of [Ru(bpy)₃]²⁺ (bpy = 2,2′-bipyridine) containing ascorbic acid (AscH₂) and ascorbate (AscH⁻) (E_ox = 0.43 V vs. SCE) as an electron donor and Co^II(Ch) (E_red = −0.96 V vs. SCE) (Fig. S1 in the ESI†) as a catalyst resulted in H₂ evolution (Fig. 1, black line). When the ratio of AscH⁻ to AscH₂ was changed as fixed total concentrations of AscH₂ and AscH⁻ ([AscH₂] + [AscH⁻] = 1.1 M), the largest H₂ evolution activity was attained with AscH⁻ (0.30 M) and AscH₂ (0.80 M) (Fig. S2 in ESI†). The smaller concentration of AscH⁻ results in less efficient reductive quenching of the [Ru(bpy)₃]²⁺* emission (* denotes the excited state). The quenching efficiency of [Ru(bpy)₃]²⁺* (E_red = 0.77 V vs. SCE in MeCN)²⁰ by AscH⁻ (0.30 M) with AscH₂ (0.80 M) was determined to be 95% (Fig. S3 in ESI†). On the other hand, the smaller concentration of AscH₂ may retard H₂ production due to decreasing the acidity. When H₂O was replaced by D₂O, D₂ and HD were produced without formation of H₂. Thus, hydrogen was produced from water and ascorbic acid as electron and proton sources. The observed deuterium kinetic isotope effect (KIE) in Fig. 1 (k_H/k_D = 1.8 in the initial stage) suggests that the Co–H bond cleavage of a cobalt hydride intermediate ([Co^III(H)(Ch)]) by proton may be the rate-determining step for the photocatalytic H₂ evolution (vide infra).

Fig. 1 Time courses of H₂ evolution in the photocatalytic reduction of proton in an Ar-saturated H₂O/MeCN (black) and D₂O/MeCN (red) mixed solution (1 : 1 v/v) containing [Ru^II(bpy)₃]²⁺ (2.0 mM), AscH₂ (0.80 M), AscHNa (0.30 M) and Co^II(Ch) (25 μM) under irradiation of visible light (λ > 420 nm) at 298 K.

The concentration of Co^II(Ch) was optimised to be 50 μM for the efficient photocatalytic H₂ evolution. The absorption of [Ru(bpy)₃]²⁺ is blocked by the larger concentration of Co^II(Ch) (Fig. S4 in ESI†).²¹ The quantum yield of the photocatalytic H₂ evolution was determined to be 12% using a ferric oxalate actinometer (see the Experimental section in ESI†). This value is similar to the highest value reported for photocatalytic H₂ evolution using a cobalt terpyridine complex (Φ = 0.13).²²

Nanosecond transient absorption spectra of an H₂O/MeCN solution of [Ru(bpy)₃]²⁺ with AscH₂ and AscH⁻ are shown in Fig. 2, where appearance of the absorption band at 500 nm due to [Ru(bpy)₃]⁺ is observed upon the nanosecond laser excitation. Thus, electron transfer from AscH⁻ to [Ru(bpy)₃]²⁺* occurred to produce AscH˙ and [Ru(bpy)₃]⁺. The rate constant of electron transfer from AscH⁻ to [Ru(bpy)₃]²⁺* (k_et) was determined to be 8.0 × 10⁸ M⁻¹ s⁻¹ from a slope of Stern–Volmer plot (K_SV = 3.5 × 10² M⁻¹) and the lifetime of [Ru(bpy)₃]²⁺* (0.44 μs in water/MeCN 1 : 1 v/v) (Fig. S5 in ESI†).²³ The decay rate of absorbance at 500 nm due to [Ru(bpy)₃]⁺ obeyed the second-order kinetics of bimolecular back electron transfer from [Ru(bpy)₃]⁺ to AscH˙. In the presence of Co^II(Ch), the decay of absorbance became much faster because of electron transfer from [Ru(bpy)₃]⁺ to Co^II(Ch) as shown in Fig. 2b. The decay rate constant linearly increased with increasing the concentration of [Co^II(Ch)] (Fig. S6 in ESI†). The rate constant of electron transfer from [Ru(bpy)₃]⁺ to Co^II(Ch) was determined to be 2.5 × 10⁹ M⁻¹ s⁻¹ from the slope of dependence of the first-order decay rate constant on concentration of Co^II(Ch) (Fig. S6b in ESI†).

Fig. 2 (a) Transient absorption spectra after laser excitation (λ = 450 nm) of [Ru^II(bpy)₃]²⁺ (80 μM) in the presence of AscH₂ (0.80 M) and AscHNa (0.30 M) in a deaerated H₂O/MeCN mixed solution (1 : 1 v/v) at 298 K. (b) Time profiles of absorbance at 500 nm due to decay of [Ru(bpy)₃]⁺ in the presence of various concentrations of Co^II(Ch) (0–20 μM) in deaerated H₂O/MeCN mixed solutions (1 : 1 v/v) containing [Ru^II(bpy)₃]²⁺ (80 μM), AscH₂ (0.80 M), AscHNa (0.30 M).

To examine the reaction of [Co^I(Ch)]⁻ that is produced by electron transfer from [Ru(bpy)₃]⁺ to Co^II(Ch), [Co^I(Ch)]⁻ was prepared by the one-electron reduction of Co^II(Ch) by decamethylcobaltocene [Co(Cp*)₂] in MeCN (Fig. S7 in ESI†). The UV-vis absorption band of [Co^I(Ch)]⁻ (red line in Fig. 3a; λ_max = 510 nm) decreased with increasing absorption band at 660 nm (black line) at 66 ms after addition of acetic acid (CH₃COOH) (0.30 M). Then, this absorption band was finally blue shifted to λ_max = 652 nm, which is due to [Co^III(Ch)]⁺.^24,25 Thus, [Co^I(Ch)]⁻ may react with CH₃COOH to form the hydride complex ([Co^III(H)(Ch)(CH₃COO)]⁻: λ_max = 660 nm), from which H₂ was evolved by the reaction with CH₃COOH to produce [Co^III(Ch)]⁺. The reaction of [Co^I(Ch)]⁻ with CH₃COOH was monitored by the absorption change at 652 nm due to [Co^III(Ch)]⁺ as shown in Fig. 3, where the rate of the formation of [Co^III(Ch)]⁺ obeyed first-order kinetics (Fig. S8 in ESI†). The first-order rate constant increased with increasing concentration of CH₃COOH to approach a constant value (Fig. 3b). Such a saturation behaviour indicates that CH₃COOH is not involved in the rate-determining step and that the reaction of [Co^I(Ch)]⁻ with CH₃COOH proceeds via formation of the hydride complex ([Co^III(H)(Ch)(CH₃COO)]⁻), followed by the rate-determining heterolytic cleavage of the Co^III–H bond. The subsequent reaction of the released hydride ion with CH₃COOH to produce H₂ and [Co^III(Ch)]⁺ may be fast as compared with the back reaction of the Co^III–H bond cleavage (Scheme 1). The kinetic equation for the formation of [Co^III(Ch)]⁺ is given by eqn (1),

d[[Co^III(Ch)]⁺]/dt = k[[Co^III(H)(Ch)(CH₃COO)]⁻] (1)

where k is the rate constant of the hydrogen evolution. From the equilibrium constant (K), the concentration of a complex between [Co^I(Ch)]⁻ and CH₃COOH is given by eqn (2), where

[[Co^III(H)(Ch)(CH₃COO)]⁻] = K[CH₃COOH]([[Co^I(Ch)]⁻]₀ − [[Co^III(Ch)]⁺])/(1 + K[CH₃COOH]) (2)

[[Co^I(Ch)]⁻]₀ is the initial concentration. Eqn (1) is rewritten by eqn (3).

d[[Co^III(Ch)]⁺]/dt = kK[CH₃COOH]([[Co^I(Ch)]⁻]₀ − [[Co^III(Ch)]⁺])/(1 + K[CH₃COOH]) (3)

Under the conditions, the concentration of CH₃COOH is much higher than that of [Co^I(Ch)]⁻, the k_obs value is given by eqn (4). To determine the k value, eqn (4) is rewritten by eqn (5), which predicts

k_obs = kK[CH₃COOH]/(1 + K[CH₃COOH]) (4)

k_obs⁻¹ = 1/kK·[CH₃COOH]⁻¹ + 1/k (5)

a linear correlation between k_obs⁻¹ and [CH₃COOH]⁻¹ (Fig. S9 in ESI†). The k and K values were determined from the intercept and slope of the linear plot of k_obs⁻¹vs. [CH₃COOH]⁻¹ to be 5.9 s⁻¹ and 7.1 M⁻¹.

Fig. 3 (a) UV-vis absorption spectral changes of [Co^I(Ch)]⁻ (30 μM) upon addition of CH₃COOH (0.30 M) in dearated MeCN at 298 K. The black and blue lines show the spectra taken at 66 ms and 1 s after mixing, respectively. The red line shows UV-vis absorption spectrum of [Co^I(Ch)]⁻ (15 μM) formed by the electron-transfer reduction of Co^II(Ch) (15 μM) with CoCp₂* (300 μM) in dearated MeCN at 298 K. (b) Plot of k_obs for the rate of formation of [Co^III(Ch)]⁺vs. [CH₃COOH].

Scheme 1 Mechanism of hydrogen formation by the reaction of [Co^I(Ch)]⁻ with CH₃COOH.

When CH₃COOH was replaced by CH₃COOD, the deuterium kinetic isotope effect (KIE) was observed (Fig. S10 in ESI†),²⁶ indicating that the cleavage of the Co–H bond of [Co^III(H)(Ch)(CH₃COO)]⁻ or O–H bond of CH₃COOH is involved in the rate-determining step of the reaction of [Co^I(Ch)]⁻ with CH₃COOH. Because CH₃COOH is not involved in the rate-determining step (vide infra), the cleavage of the Co–H bond of [Co^III(H)(Ch)(CH₃COO)]⁻ is the rate-determining step of the reaction of [Co^I(Ch)]⁻ with CH₃COOH. The KIE value was 1.7 which is virtually the same as observed for the photocatalytic H₂ evolution (KIE = 1.8, Fig. 1), indicating that the heterolytic Co–H bond cleavage of [Co^III(H)(Ch)(CH₃COO)]⁻ is also the rate-determining step in the photocatalytic H₂ evolution.

[Co^III(Ch)]⁺ produced by the reaction of [Co^III(H)(Ch)–(CH₃COO)]⁻ with CH₃COOH is reduced by AscH⁻ to form Co^II(Ch) as shown by stopped-flow measurements in Fig. 4.²⁷ The rate constant of electron transfer from AscH⁻ to [Co^III(Ch)]⁺ that was prepared by the one-electron oxidation of Co^II(Ch) with (p-BrC₆H₄)₃N˙⁺SbCl₆⁻ in H₂O/MeCN was determined to be 1.5 × 10³ M⁻¹ s⁻¹ from the linear dependence of the first-order rate constant on concentration of AscH⁻ (Fig. S11 in ESI†).

Fig. 4 (a) UV-vis absorption spectral changes in the electron-transfer reduction of [Co^III(Ch)]⁺ (15 μM) with AscHNa (50 mM) in air-saturated H₂O/MeCN mixed solutions (1 : 1 v/v) at 298 K taken at 70 ms and 118 ms after mixing. (b) Decay time profiles of absorbance at 652 nm due to [Co^III(Ch)]⁺ in the presence of various concentrations of AscHNa in air-saturated H₂O/MeCN mixed solutions (1 : 1 v/v) at 298 K.

The photocatalytic cycle is summarized in Scheme 2. Photoexcitation of [Ru(bpy)₃]²⁺ resulted in electron transfer from AscH⁻ to [Ru(bpy)₃]²⁺* to produce [Ru(bpy)₃]⁺, followed by electron transfer from [Ru(bpy)₃]⁺ to Co^II(Ch) to produce [Co^I(Ch)]⁻, which reacts with AscH₂ to produce [Co^III(H)(Ch)(AscH)]⁻. Hydrogen is generated by the reaction of [Co^III(H)(Ch)(AscH)]⁻ with AscH₂via the Co–H bond heterolysis to produce [Co^III(Ch)]⁺,^28,29 which is reduced by AscH⁻ to regenerate Co^II(Ch). In such a case, a one-photon two-electron process is made possible, because one photon is required to produce [Co^I(Ch)]⁻ for H₂ evolution and another electron is provided thermally by AscH⁻.

Scheme 2 Mechanism of photocatalytic hydrogen evolution from AscH⁻ and AscH₂ with [Ru(bpy)₃]²⁺ and Co^II(Ch).

In conclusion, Co^II(Ch) acts as an efficient catalyst for photocatalytic H₂ evolution from ascorbic acid with [Ru(bpy)₃]²⁺ as a photocatalyst to attain the high quantum yield via a one-photon two-electron process in which the second electron is provided thermally from ascorbic acid.

This work was supported by Grants-in-Aid (no. 26620154 and 26288037 to K.O.) and JSPS fellowship (No. 25-727 to K.M.) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT); ALCA and SENTAN projects from JST, Japan (to S.F.).

Notes and references

1. R. A. Kerr and R. F. Service, Science, 2005, 309, 101.
2. X. Song, H. Wen, C. Ma, H. Chen and C. Chen, New J. Chem., 2015, 39, 1734.
3. K. Kawano, K. Yamauchi and K. Sakai, Chem. Commun., 2014, 50, 9872.
4. X. Wang, S. Goeb, Z. Ji, N. A. Pogulaichenko and F. N. Castellano, Inorg. Chem., 2011, 50, 705.
5. L.-Z. Fu, L.-L. Zhou, L.-Z. Tang, Y.-X. Xhang and S.-Z. Zhan, J. Power Sources, 2015, 280, 453.
6. S. Fukuzumi, Curr. Opin. Chem. Biol., 2015, 25, 18.
7. D. Basu, S. Mazumder, X. Shi, H. Baydoun, J. Niklas, O. Poluektov, H. B. Schlegel and C. N. Verani, Angew. Chem., Int. Ed., 2015, 54, 2105.
8. A. Call, Z. Codola, F. Acuna-Pares and J. Lloret-Fillol, Chem. – Eur. J., 2014, 20, 6171.
9. H. Lv, W. Guo, K. Wu, Z. Chen, J. Bacsa, D. G. Musaev, Y. V. Geletii, S. M. Lauinger, T. Lian and C. L. Hill, J. Am. Chem. Soc., 2014, 136, 14015.
10. L. Chen, G. Chen, C.-F. Leung, S.-M. Yiu, C.-C. Ko, E. Anxolabéhère-Mallart, M. Robert and T.-C. Lau, ACS Catal., 2015, 5, 356.
11. A. Zarkadoulas, E. Koutsouri, C. Kefalidi and C. A. Mitsopoulou, Coord. Chem. Rev., 2014, 11, 6.
12. K. Maeda, M. Eguchi and T. Oshima, Angew. Chem., Int. Ed., 2014, 53, 13164.
13. J. Zhao, Y. Ding, J. Wei, X. Du, Y. Yu and R. Han, Int. J. Hydrogen Energy, 2014, 39, 18908.
14. E. Deponti, A. Luisa, M. Natali, E. lengo and F. Scandola, Dalton Trans., 2014, 43, 16345.
15. H. Ozawa and K. Sakai, Chem. Commun., 2011, 47, 2227.
16. A. Rodenberg, M. Orazietti, B. Probst, C. Bachmann, R. Alberto, K. K. Baldridge and P. Hamm, Inorg. Chem., 2015, 54, 646.
17. S. Fukuzumi, T. Kobayashi and T. Suenobu, Angew. Chem., Int. Ed., 2011, 50, 728.
18. J. L. Dempsey, B. S. Brunschwig, J. R. Winkler and H. B. Gray, Acc. Chem. Res., 2009, 42, 1995.
19. K. Mase, K. Ohkubo and S. Fukuzumi, J. Am. Chem. Soc., 2013, 135, 2800.
20. J. Yuasa and S. Fukuzumi, J. Am. Chem. Soc., 2006, 128, 14281.
21. The photocatalytic H₂ evolution in our optimized conditions, the absorption at λ = 450 nm of [Ru(bpy)₃]²⁺ (2.0 mM, ε_450nm = 1.4 × 10⁴ M⁻¹ cm⁻¹) is not significantly blocked by that of Co^II(Ch) (Abs_450nm = 0.38; ε_450nm = 1.8 × 10⁴ M⁻¹ cm⁻¹) under this experimental conditions.
22. C. V. Krishnan and N. Sutin, J. Am. Chem. Soc., 1981, 103, 2141.
23. The emission lifetime of [Ru(bpy)₃]²⁺* in water at 298 K was reported to be 0.58 μs; see: J. V. Houten and R. J. Watts, J. Am. Chem. Soc., 1976, 98, 4853.
24. The spectrum of [Co^III(Ch)]⁺ obtained by the reaction [Co^I(Ch)]⁻ with CH₃COOH was identical to that of [Co^III(Ch)]⁺ prepared by the electron-transfer oxidation of Co^II(Ch) by a one-electron oxidizing reagent of (p-BrC₆H₄)₃N˙⁺SbCl₆⁻ (E_red = 1.05 V vs. SCE).¹⁹.
25. [Co^III(Ch)]⁺ or Co^III(H)(Ch) species is not re-reduced by large excess of Co(Cp*)₂, under the present reaction conditions because Co^II(Ch) with 20 molar equiv. of Co(Cp*)₂ is necessary to quantitatively produce [Co^I(Ch)]⁻ as shown in ESI,† Fig. S7. Co(Cp*)₂ (E_1/2^+/0 = −1.47 V vs. SCE) is unstable even in carefully degassed and dehydrated MeCN.
26. The KIE value was determined from the k_obs values at [CH₃COOH] = [CH₃COOD] = 1.0 M.
27. Neither oxidation of [Co^III(Ch)]⁺ nor O₂ reduction was observed under the basic reaction conditions.
28. S. Mandal, S. Shikano, Y. Yamada, Y.-M. Lee, W. Nam, A. Llobet and S. Fukuzumi, J. Am. Chem. Soc., 2013, 135, 15294.
29. No Co(II)–H complex is involved in the heterolysis of the Co–H bond as reported in ref. 28.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and cyclic voltammograms (Fig. S1), time courses of H₂ evolution (Fig. S2), emission spectra (Fig. S3 and S5), UV-vis absorption spectra (Fig. S4 and S7) and kinetic data (Fig. S6–S11). See DOI: 10.1039/c5cc05064b

---