Elsevier

Journal of Catalysis

Volume 381, January 2020, Pages 402-407
Journal of Catalysis

Tricopper-polyoxometalate catalysts for water oxidation: Redox-inertness of copper center

https://doi.org/10.1016/j.jcat.2019.11.025Get rights and content

Highlights

  • Theoretical insight into water oxidation catalysis by Cu-containing POM.

  • O-O bond formation via the single electron transfer-WNA mechanism.

  • Redox-inertness of Cu center.

  • Two-state reactivity.

Abstract

In this work, the mechanism of water oxidation catalyzed by a tricopper-containing polyoxometalate (POM) [(SbW9O33)2CuII3(H2O)3]12− ([POM-CuII2CuIIsingle bondOH2]12−), including (i) the deprotonation oxidation and (ii) Osingle bondO bond formation steps, was theoretically investigated for the first time. Calculations suggest that the Cu center is redox-inert and remains bivalence throughout the catalytic reaction. This is significantly different from the Ru-, Co- and Mn-containing POM-based water oxidation catalysts in which the metal centers (Ru, Co, Mn) are sequentially oxidized and reduced during catalytic reaction. In deprotonation oxidation step, two hydrogen atoms are removed from one water ligand. Therefore, an unusual metal-oxyl-radical species [POM-CuII2CuIIsingle bondOradical dotradical dot]12− is obtained, in which two unpaired electrons are assigned to the active O center. A monocopper-containing keggin POM [CuII(Oradical dotradical dot)SbW11O39]5− ([POM-CuIIsingle bondOradical dotradical dot]5−) was selected as the model to explore the Osingle bondO bond formation mediated by [POM-CuII2CuIIsingle bondOradical dotradical dot]12−. The Osingle bondO bond formation was proposed to proceed via the single electron transfer-water nucleophilic attack (SET-WNA) mechanism. Two electrons from the incoming water molecule is step by step transferred to the oxyl-radical ligand of [POM-CuIIsingle bondOradical dotradical dot]5−. Furthermore, the SET-WNA process is characterized by two-state reactivity and the observed spin inversions between two potential energy surfaces effectively reduce the activation barrier. Therefore, the present work would provide the promising information for further designing water oxidation catalysts (WOCs).

Introduction

Water splitting reaction driven by sunlight to generate oxygen (O2) and hydrogen (H2) molecules is considered as one of the most promising strategy for generating clean energy in a sustainable manner [1]. In this manner, solar energy is stored into chemical bonds, which is similar manner as done by photosystem II of green plants and algae [2]. Based on electrochemical perspective, the water splitting reaction includes two half-reactions: (i) water oxidation 2H2O → O2 + 4H+ + 4e; (ii) proton reduction (4H+ + 4e → 2H2). Particularly the water oxidation half-reaction is not only energy demanding, but also high overpotentials needed because it involves four Hsingle bondO bonds ruptures, removal of four electrons and four protons, and an Osingle bondO bond formation. Thus the water oxidation half-reaction is generally regarded as the bottleneck of efficient water splitting reaction [3]. One of the potential strategies for overcoming high activation energies is to use transition metals (TMs)-containing complexes as catalysts. To date, a considerable number of complexes containing Ru, Co, Ni, Fe, Cu, Mn have been reported as water oxidation catalysts (WOCs) [4]. Among them, Cu-based complexes have been regarded as promising WOCs due to their excellent activity, great stability and relatively low light-absorptivity [5].

However, under harsh reaction conditions, the organic ligands of organometallic-based WOCs are thermodynamically unstable, which ultimately result in inactivation of catalysts. Therefore, the design and synthesis of efficient catalysts with high stability are imperative for efficient water oxidation reaction. In this circumstance, polyoxometalate (POM)-based WOCs are distinct and attractive since they are oxidatively resistant and hydrolytically stable in certain pH ranges [4](c), [6]. In 2018, Ding group reported a tricopper-containing POM-based WOC [(SbW9O33)2CuII3(H2O)3]12− ([POM-CuII2CuIIsingle bondOH2]12−, 1, Fig. 1) [7], which is the first Cu-containing POM-based catalyst for O2 production via electrocatalytic water oxidation at neutral pH condition. Each CuII center (Cu1, Cu2, Cu3) in 1 coordinates with four O atoms from two POM ligands and one water molecule, which results in a sandwich-type assembly. That is, three CuII centers are equally reactive sites and locate in the identical chemical environment.

Atomistic level understanding of catalytic process is indispensable for improving the activities of catalysts and designing more excellent catalysts. Considering the bulkiness and structural complexities of POM-based WOCs, revealing their catalytic mechanisms remain challenging from experimental insight. Therefore, the computational strategy has been highly expected. In 2013, the catalytic mechanism on tetraruthenium-containing POM-based WOC [Ru4O4(OH)2(H2O)4(γ-SiW10O36)2]10− was theoretically explored [8]. The RuVI-oxo species is obtained after stepwise oxidation of tetraruthenate core, which triggers the water nucleophilic attack (WNA) and O2 release. Meanwhile, our group investigated the reaction mechanism of the monoruthenium-containing POM-based WOCs [RuIII(H2O)SiW11O39]5− and [RuIII(H2O)GeW11O39]5− using density functional theory (DFT) calculations [9]. The Osingle bondO bond formation was proposed to occur via WNA mechanism from the RuV-oxo species. In 2017, Poblet group theoretically elucidated the catalytic process of the tetracobalt-containing POM-based WOC [Co4(H2O)2(PW9O34)2]10− [10], in which the CoIII-oxyl-radical species was proposed to be responsible for Osingle bondO bond formation via WNA mechanism. Recently, the water oxidation catalyzed by a polyoxometalate-based complex [Mn3(H2O)3(SbW9O33)2]12− was theoretically investigated by our group and the MnV-oxo species is computed to be responsible for Osingle bondO bond formation via WNA mechanism [11]. However, the study on catalytic mechanism mediated by Cu-containing POM-based WOC is absent.

In present work, we detailedly investigated the water oxidation catalyzed by 1, including (i) the deprotonation oxidation and (ii) Osingle bondO bond formation steps, using DFT methods.

Section snippets

Computational details

All optimizations were performed by DFT approach with (U)B3LYP hybrid functional [12] in Gaussian09 program package [13]. The solvent effect was included with the conductor-like polarizable continuum model (CPCM) and water solvent [14]. The LANL2TZ(f) [15], LANL2DZ [16] and 6-31G(d, p) [17] were applied for the Cu, W/Sb and H/O, respectively. The intrinsic reaction coordinate (IRC) calculations were performed to verify the correct connection between transition state (TS) and corresponding

Deprotonation oxidation

The ground state of 1 is quartet as the spin population on each CuII center is 0.7 |e|. The average distance among three Cu centers is as long as 4.90 Å, indicating the weak interaction among three CuII centers. Therefore, to simplify the calculations, a single-site pathway in which only one Cu center (Cu1) as well as its water ligand is involved and other two Cu centers (Cu2, Cu3) are kept in high-spin d9 electron configurations, was proposed [8], [10], [21]. The catalytic cycle of water

Conclusion

In present work, the mechanistic details of water oxidation catalyzed by a tricopper-POM [POM-CuII2CuIIsingle bondOH2]12−, especially deprotonation oxidation and Osingle bondO bond formation steps, have been theoretically investigated. During deprotonation oxidation step, the two hydrogen atoms are removed from the water ligand of [POM-CuII2CuIIsingle bondOH2]12−, which result in an unusual metal-oxyl-radical species [POM-CuII2CuIIsingle bondOradical dotradical dot]12−. The monocopper-containing keggin POM [POM-CuIIsingle bondOradical dotradical dot]5− was selected as suitable model to

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors gratefully acknowledge financial support by NSFC (grant no. 21403033 and 21571031). We acknowledge the LvLiang Cloud Computing Center of China, and the calculations were performed on TianHe-2.

References (25)

  • L. Sun et al.

    Chem. Soc. Rev.

    (2001)
    J.H. Alstrum-Acevedo et al.

    Inorg. Chem.

    (2005)
    D.G. Nocera

    Acc. Chem. Res.

    (2012)
    W.J. Youngblood et al.

    J. Am. Chem. Soc.

    (2009)
    H. Dau et al.

    ChemCatChem

    (2010)
  • P.J. Hay et al.

    J. Chem. Phys.

    (1985)
  • A. Melis

    Energy Environ. Sci.

    (2012)
    D. Gust et al.

    Acc. Chem. Res.

    (2009)
  • J. Nowotny et al.

    Int. J. Hydrogen Energy

    (2005)
    N. Cox et al.

    Acc. Chem. Res.

    (2013)
  • J.D. Blakemore et al.

    Chem. Rev.

    (2015)
    T.J. Meyer et al.

    Chem. Soc. Rev.

    (2017)
    Q. Han et al.

    Dalton Trans.

    (2018)
  • T. Zhang et al.

    J. Am. Chem. Soc.

    (2014)
    R. Terao et al.

    ChemPlusChem

    (2016)
    X.J. Su et al.

    Angew. Chem. Int. Ed.

    (2015)
    Y.J. Liu et al.

    Chem. Sci.

    (2019)
    S.J. Koepke et al.

    J. Am. Chem. Soc.

    (2017)
    K.J. Fisher et al.

    ACS Catal.

    (2017)
    F.F. Chen et al.

    Inorg. Chem.

    (2017)
  • A.R. Howells et al.

    J. Am. Chem. Soc.

    (2004)
    Y.V. Geletii et al.

    Angew. Chem., Int. Ed.

    (2008)
    A. Sartorel et al.

    J. Am. Chem. Soc.

    (2008)
    Q. Yin et al.

    Science

    (2010)
    M. Martin-Sabi et al.

    Nat. Catal.

    (2018)
    H.J. Lv et al.

    Chem. Soc. Rev.

    (2012)
  • L. Yu et al.

    Chem. Commun.

    (2018)
  • S. Piccinina et al.

    PNAS

    (2013)
  • Z.L. Lang et al.

    Dalton Trans.

    (2013)
  • J. Soriano-López et al.

    J. Catal.

    (2017)
  • X.F. Su et al.

    J. Catal.

    (2019)
  • Cited by (0)

    View full text