Self-assembled nickel cubanes as oxygen evolution catalysts†

Ana C. García-Álvarez ^a, Stefani Gamboa-Ramírez ^b, Diego Martínez-Otero ^c, Maylis Orio *^b and Ivan Castillo *^a
aInstituto de Química, Universidad Nacional Autónoma de México, Circuito Exterior, CU, Ciudad de México, 04510, Mexico. E-mail: joseivan@unam.mx
^bAix Marseille Université, CNRS, Centrale Marseille, iSm2, 13397, Marseille, France. E-mail: maylis.orio@univ-amu.fr
^cCentro Conjunto de Investigación en Química Sustentable UAEM-UNAM, Carretera Toluca-Atlacomulco km. 14.5, Toluca, 50200, Estado de México, Mexico

First published on 2nd August 2021

---

---

Catalytic water oxidation is a global challenge representing the bottleneck for water splitting, which could provide abundant fuels from renewable sources. Molecular compounds and materials that catalyse the 4-electron oxidation of water to dioxygen have benefitted from the insight gained on the structure of the Mn₄O₅Ca oxygen-evolving complex of photosystem II (PSII, Chart 1a).¹ Nature accumulates four oxidising equivalents in a multimetallic-oxo cluster, and this has emerged as an appealing strategy to obtain compounds featuring metal-oxo subunits, with emphasis on complexes and materials that feature the “cubane”-type M₄O₄ motif (Chart 1b). Such clusters appear as superior alternatives to monometallic complexes, since the former may offer the synergistic action of four metal centres, without the problems inherent to the identification of well-defined active sites in heterogeneous systems that catalyse the oxygen evolution reaction (OER).

M₄O₄ structures that can be easily obtained from readily available starting materials and earth-abundant metals represent a synthetic challenge for efficient OER catalysts as renewable and sustainable energy sources,² with artificial photosynthesis as the ultimate goal.³ The need for predictable synthetic methods has resorted to computational approaches,⁴ including machine learning methods.⁵ Although breakthroughs for the synthesis of M₄O₄ cubanes have been reported,^2a,6 much remains to be established in the field.

In this context, the use of nickel has resulted in Ni₄O₄ structures, but benzimidazole-derived cubanes have not been tested in a rational and systematic fashion as OER electrocatalysts. Ni₄O₄ cubanes have been obtained with sterically encumbering,⁷ and intricate polydentate ligands.⁸ In contrast, 2-benzimidazolylmethanol formed in situ allowed the crystallisation of a Ni₄O₄ complex, featuring alkoxide moieties that serve as μ₃-edges (Chart 1c).⁹ The original report focused exclusively on structural and magnetic aspects of the serendipitously obtained cubane. Rational synthesis with equimolar amounts of LⁿOH and NiCl₂·6H₂O affords complexes with the empirical formulae [(LⁿO)NiCl(S)] (L¹OH = 1-H-2-benzimidazolylmethanol, L²OH = 1-methyl-2-benzimidazolylmethanol; S = MeOH, H₂O, Scheme 1), in 85% [(L¹O)NiCl(MeOH)] and 63% [(L²O)NiCl(H₂O)] yield.

Fast Atom Bombardment (FAB-MS) and/or Electrospray Mass Spectrometry (ESI-MS) revealed the presence of tetrameric aggregates in the form of [(LⁿO)₄Ni₄]⁺ (presumably by reduction of the Ni^II centres in the ionisation chamber) or [(LⁿO)₄Ni₄Cl₃]⁺ in methanolic solution, see (ESI†) Fig. S1–S3. ¹H NMR spectroscopy in methanol-d₄ indicates that the cubanes are paramagnetic,^8a,9 based on the broad signals observed from 30 to −15 ppm. Both complexes are ESR silent at X-band frequency in perpendicular mode, as expected for Ni^II systems with S > 1/2. Their optical spectra feature ligand-centred bands around 210 (ε ∼ 18–20000 M⁻¹ cm⁻¹) and 274 (18000), as well as d–d transitions at 675 nm (25) in methanol. UV-Vis titration of the deprotonated ligands with NiCl₂ in DMF shows the same bands at all ligand/metal ratios, consistent with a 1:1 stoichiometry for the species in solution (Fig. S2c, ESI†).

The cubane-type architecture of [(μ₃-L¹O)NiCl(MeOH)]₄ (1) and [(μ₃-L²O)NiCl(H₂O)]₄ (2) was confirmed in the solid-state by X-ray crystallography (Table S1 in ESI†). Although 1 and 2 have similar structures, differences among bond lengths appear to be related to the better σ-donor properties of N-methylated L²OH, with shorter average Ni–N distance of 2.028 Å in 2vs. 2.058 Å in 1. Conversely, the average Ni–O distance in the Ni₄O₄ framework of 1 is shorter at 2.077 vs. 2.081 Å in 2. The average Ni–O distances to the exogenous MeOH or H₂O donors are also shorter in 1 (2.092 vs. 2.114 Å in 2). Distortion of the cubanes is reflected in the Ni–O lengths within the Ni₄O₄ framework, varying from 2.034(2) to 2.117(2) Å in 1, and 2.040(5) to 2.123(4) Å in 2. The O–Ni–O angles range from 79.36(7) to 82.3(1)° in 1, and 79.0(2) to 82.5(2)° in 2. All Ni–O–Ni¹⁰ angles are around 99° in both cases, with similar 2–2.5° dispersion (Fig. 1). Comparison with the analogous cubane featuring L¹OH reveals similar metric parameters.⁹ In contrast, a related cubane featuring 2-(hydroxymethyl)pyridine ligands is less distorted,¹¹ as reflected in the smaller dispersion of bond lengths and angles around the Ni₄O₄ core. A significantly longer average Ni–N distance to the pyridine donors at 2.070 Å was determined.

All Ni^II centres in 1 and 2 have one solvent molecule and one chloride as monodentate and potentially labile ligands (see computational studies), necessary for the coordination of water molecules as substrate. Moreover, H-bonding facilitates proton transfer processes such as OER. Methanol acts as H-donor towards the chlorides on the same face of the cubane in 1, with O5 through O8 at distances of 3.006–3.092 Å to Cl1–Cl4; the corresponding O–H⋯Cl angles range from 163 to 172° (Table S2 and Fig. S5, ESI†). In 2, H₂O plays the role of H-donor towards the adjacent chlorides at distances ranging from 3.017 to 3.083 Å, and O–H⋯Cl angles from 150 to 168° (Table S3 and Fig. S6, ESI†).

Cyclic voltammetry (CV) of 1 and 2 in DMF reveals irreversible oxidation waves at anodic peak potentials (E_ap) of 1.15 and 1.09 V vs. Ag/AgCl (Figs. S7 and S13, ESI†). Comparison of the currents observed for 1 and 2vs. that of ferrocene as internal standard at the same concentration of the cubanes indicates multi-electron processes (Fig. S8 and S14, ESI†). In aqueous phosphate buffer (K-Pi) at pH 7, irreversible peaks are also detected (Fig. 2). The high currents are indicative of a catalytic process, with onset potentials determined at 0.15 mM as shown in Figs. S11 and S17 (ESI†), and overpotentials η = 950 and 900 mV vs. RHE at ∼0.7 mA cm⁻² (η = 210 and 160 mV vs. NHE) for 1 and 2, respectively. These values are lower than the one example of nickel cubanes previously reported as WOC,¹² while a related Ni₄O₄ was reported as having no activity.¹³ A linear dependence of the current relative to the concentration of 1 and 2 in the range 0.05–0.30 mM is consistent with a pseudo-first-order rate constant, with κ_obs = 2.05 and 1.18 s⁻¹ respectively, corresponding to the turnover frequencies (TOF) at single-site catalysts.¹⁴

Reuse of the glassy carbon working electrode in fresh buffer after electrolysing solutions of 1 or 2 does not show electrocatalytic response (Fig. 2, blue trace). This indicates that Ni-containing deposited materials are not responsible for heterogeneous catalysis. Further support for the integrity of the molecular complexes was provided by FAB⁺ MS, showing the same fragmentation pattern of 1 before (Fig. S1a, ESI†) and after (Fig. S1b, ESI†) electrolysis. Scanning electron microscopy (SEM) on a planar graphite working electrode before and after controlled potential electrolysis of 0.15 mM 1 and 2 in K-Pi evidenced no deposition after 30 min (Fig. S18–S20, ESI†).^8d UV-Vis absorption spectra of 0.15 mM solutions of 1 and 2 in 0.1 M K-Pi were measured before and after electrolysis, with no significant changes (Fig. S21 and S22, ESI†). Lastly, successive CV scans of 0.15 mM of 1 and 2 in 0.1 M K-Pi at pH 7 showed no increase in catalytic current that may be ascribed to deposition of a heterogenous catalyst (Fig. S23 and S24, ESI†). All these experiments are consistent with a homogeneous catalytic process by the robust 1 and 2, which is not affected by phosphate at different concentrations (Fig. S25a and S26, ESI†); the electrochemical response is identical in borate buffer at pH 8.5 (Fig. S25b, ESI†).

Electrochemical measurements in D₂O K-Pi buffer allowed determination of a kinetic isotope effect (KIE), affording values of 0.7 and 1.4 for 1 and 2, respectively (Fig. S27 and S28, ESI†). This is consistent with observations involving two proximal metal-oxygen moieties during O–O bond formation in the rate-limiting step (I2M mechanism).¹⁵ A relatively small KIE suggests that primary H/D isotope effects that would directly involve O–H bonds is not likely.^3c,16–19 Secondary effects, such as protons involved within hydrogen bond networks,^20–23 could contribute to the extent measured. For comparative purposes, the monometallic complex [bis(2-(1-methylbenzimidazol-2-yl)ethyl)amine]NiCl₂ (3) featuring a tridentate benzimidazole-based ligand was employed in analogous electrochemical measurements. CV in DMF reveals E_ap at 1.10 vs. Ag/AgCl, while in K-Pi at a concentration of 0.28 mM to reach comparable current densities to those observed for 1 and 2, an overpotential η = 1.0 V vs. RHE and KIE = 1.7 (Fig. S29–S33, ESI†) was determined. This points to a cooperative effect in 1 and 2 that is absent in monometallic 3. The overall behaviour of the cubanes is analogous to that of related homogeneous monometallic Ni systems previously reported as water oxidation electrocatalysts, and differs from Ni(NO₃)₂ that serves as precursor for NiO as heterogeneous catalyst.²⁴ Confirmation of dioxygen formation during the electrocatalytic processes mediated by the cubanes was obtained by controlled potential electrolysis at 1.5 V to ensure production of O₂ in K-Pi by Clark electrode measurements (Fig. S34 and S35, ESI†).

To gain insight on the mechanism of water oxidation, DFT calculations were carried out on 1 and 2 by gas phase optimisation of geometric parameters with the BP86²⁵ functional, and subsequently in water. Standard deviations are in the range of DFT precision, 0.035–0.043 Å and 1.5–1.6° for bond lengths and angles of 1 and 2, respectively (ESI† Fig. S38–S40, S54–S56 and Tables S7, S8, S22, S23). Single-point broken symmetry calculations confirmed the high-spin nature of the electronic ground state (Fig. S41, S42, S57, S58 and Tables S9, S24, ESI†), with weak ferromagntic coupling among the Ni^II ions, as expected given the average Ni–O–Ni angle of 99° (Fig. S44 and Tables S11, S12, S25, S26, ESI†).^9,26–28 Cooperativity among metal centres for O–O bond formation during OER has been invoked in related systems, and a reasonable sequence would require initial ligand exchange. First, substitution of MeOH by H₂O in 1 is calculated to be exothermic by −7.1 kcal mol⁻¹ (Fig. S45, S46 and Tables S13, S14, ESI†). Water molecules at adjacent Ni centres requires Cl⁻ exchange for H₂O, estimated at 2.8 and 1.6 kcal mol⁻¹ for 1 and 2 (Scheme 2, Fig. S47, S48, S59, S60 and Tables S15, S16, S27, S28, ESI†). Oxidation of one nickel ion per water-substituted cubane [(μ₃-LⁿO)₄Ni₄Cl₃(H₂O)₅]⁺ to formally Ni^III in [(μ₃-LⁿO)₄Ni₄Cl₃(H₂O)₅]²⁺ (Fig S49–S51, Tables S17, S18, ESI†) was calculated at E = 1.16 and 1.31 V (vs. RHE), which are reasonably close to the experimentally determined onset potentials for OER by 1 and 2 (Fig. S52, S61 and Tables S19, S20, S29, S30, ESI†). Next, exergonic deprotonation of a water molecule coordinated to the formally Ni^III centre would result in a Ni^III–OH moiety, calculated at ΔG = −31.5 (1) and −16.1 kcal mol⁻¹ (2) with HPO₄²⁻ as base at pH 7 (Fig. S53, S62 and Tables S21, S31, S32, ESI†). Sequential electron/proton transfer (ET/PT) steps would lead to oxo/hydroxo ligands at adjacent Ni centres, and ultimately to O₂ evolution as depicted in the proposed I2M mechanism in Scheme 2.

Experimental support for the initial oxidation was provided by low-temperature ESR spectroscopy: reaction of 1 with one equiv. of ceric ammonium nitrate in DMF at 77 K resulted in a weakly anisotropic ESR signal that was simulated considering a single paramagnetic species, leading to three different principal g values at g₁ = 2.045, g₂ = 2.053, g₃ = 2.061 and ¹⁴A_N = 25 MHz. The g values are consistent with a Ni^III-centred S = 1/2 species (Fig. 3)²⁹ and in good agreement with the DFT findings regarding the computed ESR parameters for 1 (Table S33, ESI†); oxidation of 2 afforded a weak signal similar to that observed for 1.

In summary, the robust nickel cubanes 1 and 2 can be easily prepared in good yields from readily available starting materials under ambient conditions. The tetranuclear complexes are active in electrocatalytic water oxidation at neutral pH. Extension of this self-assembly strategy for the preparation of other water oxidation catalysts with benzimidazole-based ligands and earth-abundant metals is currently underway.

A. C. García-Álvarez and S. Gamboa-Ramírez contributed in investigation, methodology, analysis, and writing; D. Martínez-Otero in crystallographic data collection and analysis; M. Orio and I. Castillo in conceptualisation, funding acquisition, project administration, supervision and writing.

The authors thank L. Velasco and J. Pérez for FAB MS, M. P. Orta for combustion analysis, V. Gómez-Vidales and S. Bertaina for EPR, R. Patiño for IR, Prof. Jorge Tiburcio and Geiser Cuéllar at Cinvestav for ESI MS. We gratefully acknowledge financial support from Conacyt (A1-S-8682, beca 336107), Conacyt-ECOS Nord 291247, DGAPA-PAPIIT (IN217020), and the French National Research Agency (CUBISM, grant no. ANR-18 CE092 0040 01).

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. Y. Umena, K. Kawakami, J. R. Shen and N. Kamiya, Nature, 2011, 473, 55 CrossRef CAS PubMed .
2. (a) G. C. Dismukes, R. Brimblecombe, G. A. N. Felton, R. S. Pryadun, J. E. Sheats, L. Spiccia and G. F. Swiegers, Acc. Chem. Res., 2009, 42, 1935 CrossRef CAS PubMed ; (b) R. Mattheu, P. Garrido-Barros, M. Gil-Sepulcre, M. Z. Ertem, X. Sala, C. Gimbert-Suriñach and A. Llobet, Nat. Rev. Chem., 2019, 3, 331 CrossRef ; (c) M. D. Kärkäs and B. Åkermark, Dalton Trans., 2016, 45, 14421 RSC .
3. (a) N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 20142 CAS ; (b) R. Eisenberg and H. B. Gray, Inorg. Chem., 2008, 47, 1697 CrossRef CAS PubMed ; (c) M. D. Kärkäs, O. Verho, E. V. Johnston and B. Åkermark, Chem. Rev., 2014, 114, 11863 CrossRef PubMed .
4. (a) E. Kim, K. Huang, S. Jegelka and E. Olivetti, npj Comput. Mater., 2017, 3, 1120 Search PubMed ; (b) M. G. Mavros, J. J. Shepherd, T. Tsuchimochi, A. R. McIsaac and T. Van Voorhis, J. Phys. Chem. C, 2017, 121, 15665 CrossRef CAS .
5. (a) C. W. Coley, W. H. Green and K. F. Jensen, Acc. Chem. Res., 2018, 51, 1281 CrossRef CAS PubMed ; (b) S. Back, K. Tran and Z. W. Ulissi, ACS Catal., 2019, 9, 7651 CrossRef CAS .
6. (a) M. D. Symes, Y. Surendranath, D. A. Lutterman and D. G. Nocera, J. Am. Chem. Soc., 2011, 133, 5174 CrossRef CAS PubMed ; (b) P. F. Smith, C. Kaplan, J. E. Sheats, D. M. Robinson, N. S. McCool, N. Mezle and G. C. Dismukes, Inorg. Chem., 2014, 53, 2113 CrossRef CAS PubMed ; (c) A. I. Nguyen, M. S. Ziegler, P. Oña-Burgos, M. Sturzbecher-Hohne, W. Kim, D. E. Bellone and T. D. Tilley, J. Am. Chem. Soc., 2015, 137, 12865 CrossRef CAS PubMed .
7. A. N. Ponomaryov, N. Kim, J. Hwang, H. Nojiri, J. Van Tol, A. Ozarowski, J. Park, Z. Jang, B. Suh, S. Yoon and K. Y. Choi, Chem. – Asian J., 2013, 8, 1152 CrossRef CAS .
8. (a) K. Isele, F. Gigon, A. F. Williams, G. Bernardinelli, P. Franz and S. Decurtins, J. Chem. Soc., Dalton Trans., 2006, 332 Search PubMed ; (b) A. Das, F. J. Klinke, S. Demeshko, S. Meyer, S. Dechert and F. Meyer, Inorg. Chem., 2012, 51, 8141 CrossRef CAS ; (c) F. Song, K. Al-Ameed, M. Schilling, T. Fox, S. Luber and G. R. Patzke, J. Am. Chem. Soc., 2019, 141, 8846 CrossRef CAS ; (d) G. Azadi, Z. Zand, Y. Mousazade, R. Bagheri, J. Cui, Z. Song, R. Bikas, K. Wozniak, S. I. Allakhverdiev and M. M. Najafpour, Int. J. Hydrogen Energy, 2019, 44, 2857 CrossRef CAS .
9. X.-Y. Song, Y.-H. Xu, L.-C. Li, D.-Z. Liao and Z.-H. Jiang, Inorg. Chim. Acta, 2007, 360, 2039 CrossRef CAS .
10. M. A. Halcrow, J.-S. Sun, J. C. Huffman and G. Christou, Inorg. Chem., 1995, 34, 4167 CrossRef CAS .
11. C. G. Efthymiou, C. Papatriantafyllopoulou, N. I. Alexopoulou, C. P. Raptopoulou, R. Boča, J. Mrozinski, E. G. Bakalbassis and S. P. Perlepes, Polyhedron, 2009, 28, 3373 CrossRef CAS .
12. J. Wang, X. Meng, W. Xie, X. Zhang, Y. Fan and M. Wang, J. Biol. Inorg. Chem., 2021, 26, 205 CrossRef CAS PubMed .
13. F. Song, R. Moré, M. Schilling, G. Smolentsev, N. Azzaroli, T. Fox, S. Luber and G. R. Patzke, J. Am. Chem. Soc., 2017, 139, 14198 CrossRef CAS PubMed .
14. W. S. Gao, J. M. Wang, N. N. Shi, C. N. Chen, Y. H. Fan and M. Wang, New J. Chem., 2019, 43, 4640 RSC .
15. M. D. Kärkäs and B. Åkermark, Dalton Trans., 2016, 45, 14421 RSC .
16. V. Fourmond, P. A. Jacques, M. Fontecave and V. Artero, Inorg. Chem., 2010, 49, 10338 CrossRef CAS PubMed .
17. V. Artero and J. M. Saveant, Energy Environ. Sci., 2014, 7, 3808 RSC .
18. A. M. Appel and M. L. Helm, ACS Catal., 2014, 4, 630 CrossRef CAS .
19. C. Costentin, S. Drouet, M. Robert and J. M. Savéant, J. Am. Chem. Soc., 2012, 134, 11235 CrossRef CAS .
20. R. S. Khnayzer, V. S. Thoi, M. Nippe, A. E. King, J. W. Jurss, K. A. El Roz, J. R. Long, C. J. Chang and F. N. Castellano, Energy Environ. Sci., 2014, 7, 1477 RSC .
21. 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 CrossRef CAS .
22. Z. Han, W. R. McNamara, M. S. Eum, P. L. Holland and R. Eisenberg, Angew. Chem., Int. Ed., 2012, 51, 1667 CrossRef CAS PubMed .
23. Y. Xu, X. Yin, Y. Huang, P. Du and B. Zhang, Chem. – Eur. J., 2015, 21, 4571 CrossRef CAS PubMed .
24. D. Wang and C. O. Brunauer, Inorg. Chem., 2017, 56, 13638 CrossRef CAS PubMed .
25. J. P. Perdew, Phys. Rev. B: Condens. Matter Mater. Phys., 1986, 33, 8822 CrossRef .
26. J. Mayans, A. A. Athanasopoulou, A. T. Pham, M. Font-Bardia, E. C. Mazarakioti, M. Pilkington, T. C. Stamatatos and A. Escuer, Dalton Trans., 2019, 10427 RSC .
27. H. A. Rudbari, F. Lloret, M. Khorshidifard, G. Bruno and M. Julve, RSC Adv., 2016, 6, 7189 RSC .
28. Q.-L. Zhang, Z.-L. Wu, H. Xu, B. Zhai, Y.-F. Wang, G.-W. Feng and Y.-L. Huang, Z. Anorg. Allg. Chem., 2016, 642, 414 CrossRef CAS .
29. (a) F. F. Pfaff, F. Heims, S. Kundu, S. Mebs and K. Ray, Chem. Commun., 2012, 48, 3730 RSC ; (b) M.-C. Kafentzi, M. Orio, M. Réglier, S. Yao, U. Kuhlmann, P. Hildebrandt, M. Driess, A. J. Simaan and K. Ray, Dalton Trans., 2016, 45, 15994 RSC .

---

Footnote

† Electronic supplementary information (ESI) available: Crystallographic, spectroscopic, cyclic voltammetry, and DFT data. CCDC 2034867 and 2034868. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/d1cc03227e

This journal is © The Royal Society of Chemistry 2021