Water splitting into hydrogen and oxygen by non-traditional redox inactive zinc selenolate electrocatalyst

The development of alternative energy sources is the utmost priority of the developing society. Unlike many prior homogenous electrocatalysts that rely on the change in an oxidation state of metal center and or electrochemically active ligand, the synthesized novel bimetallic zinc selenolate complex consisting of redox inactive zinc metal ion and catalytically inactive ligand catalyzes the electrochemical oxygen evolution from the water with rate constant 7.28 s-1 at onset potential 1.028 V vs. NHE. On the other hand, the hydrogen evolution reaction proceeds with 47.32 s-1 observed rate constant and -0.256 V onset potential vs. NHE. DFT computations and control experiments suggest that the redox chemistry at selenium center, Lewis acidity, and cooperativity effect of two zinc atoms facilitate the electrochemical oxidation and reduction of water into oxygen and hydrogen, respectively.


Introduction
With the exponential increase in the consumption of current energy sources across the world, the biggest challenge that lies ahead is the development of alternative energy sources that are competitive with the currently available energy resources. 1 Over the past decades, substantial progress has been made in energy production electrochemically, mainly those involving water, hydrogen, and oxygen, wherein water oxidizes at the anode, known as oxygen evolution reaction (OER) and the reduction of the proton to the hydrogen, termed as hydrogen evolution reaction (HER) occurs at the cathode. [2][3][4] Generally, heavy transition metal (TM) catalytic system augment OER and HER. [5][6][7][8][9][10][11][12] In these catalysts, the metal ion performs the electron transfer and also interacts with the substrate during bond-forming and breaking events. [13][14] Consequently, a metal center is associated with the formation of metal hydride and metal oxide intermediates in HER and OER, respectively. Moreover, earth-abundant 3d-transition metals such as nickel, copper, and cobalt complexes (Scheme 1) have been prepared with the appropriate choice of organosulfur and organoselenium ligands having selenium or sulfur at the active site to interact with the proton and to assist the formation of a metal-hydride bond by the proton shuttling process, which is the crucial step in the reduction of proton in hydrogen evolution reaction. [15][16][17][18][19][20] Besides the organotransition metal selenolate electrocatalysts, the selenium and transition metals (Fe, Co, Ni, Cu) derived materials also provide an alternate for the electrocatalytic water oxidation and water reduction reactions (OER and HER). 21 Further, for oxygen evolution reaction (OER) from water, redox active ligands have been employed to stabilize the high oxidation state of central 3d-transition metal in the catalytic intermediates which enhances the reactivity of the oxidized metal oxide intermediates. [22][23][24][25][26][27] Thus, the reactivity of metalhydride and metal oxide bonds is crucial for optimum catalytic e ciency. Also, the traditional metal hydrides and metal oxides require open coordination sites and able to accommodate multiple redox two and four-electron processes. [28][29] In this context, it is desirable to explore new pathways in which metal hydrides are not necessary to involve in the catalytic cycle. Further, economic and sustainable alternatives are highly desirable to ease out the dependency on the transition metals (TMs, vide infra). In 2016, Graupperhaus and co-workers reported the electrocatalytic proton reduction of acetic acid and oxidation of hydrogen gas by redoxactive thiosemicarbazone ligand along with its zinc complex by avoiding the traditional metal hydride approach (Scheme 1). [30][31] Nonetheless, zinc complexes have not been reported, which could electrocatalyze water splitting. Sun et al. have unsuccessfully attempted the electrocatalytic water reduction by using zinc(II)pentapyridine complex. However, the cobalt(II) complex of the same ligand was successfully able to reduce water electrocatalytically. 32 From the past decade, our group has been active in the synthesis and catalytic activity of organoselenium compounds. [33][34][35] Recently our group has reported a diorgano diselenide derived from the o-aminodiselenide ligand, which could activate aerial oxygen towards the oxidation of organothiols. [36][37] Inspired by nature 38 and by the non-transition metal electrocatalysts; zinc thiosemicarbazone and aluminium-bis aminopyridine (Scheme 1) for hydrogen evolution reactions (HER) from water, [30][31][39][40] herein, we report the synthesis and structural characterization of the novel bimetallic zinc selenolate electrocatalyst 1 that catalyzes the water-splitting reaction by ligand assisted pathways in both directions; oxygen evolution and hydrogen evolution reactions without adding acid or base. Mechanistic insights into the electrocatalytic activity of bimetallic zinc selenolate complex have also been gained by synthesizing mercury selenolate catalyst.

Electrochemical Studies
The cyclic voltammogram (CV) of catalyst 1 at 1 mM in propylene carbonate obtained using a glassy carbon electrode displayed a pronounced catalytic wave. Propylene carbonate solvent was chosen for the electrocatalysis because of it has wide potential window with an oxidative limit of >2.0 (vs. NHE), and weak coordinating ability in comparison with water. 7,50-51 The onset of the catalytic wave was observed at 1.13 V versus normal hydrogen electrode (NHE, the potentials presented in this study are referenced to it). The cyclic voltammogram of catalyst 1 revealed a one-electron oxidation wave at 1.14 V vs. NHE, corresponding to the oxidation of Se -2 to Se -1 , 52 while a weak reverse peak was also observed (Figure 2A).
A solution of bimetallic zinc catalyst 1 (1 mM) in propylene carbonate solution containing 1% water shows an increase in anodic current of 19 µA in comparison to without water ( Figure 2A).
Further, the anodic current increases up to 31 µA with increasing water concentration in the electrochemical cell, which is indicative of an electrocatalytic OER process. The catalytic current remained unaltered beyond 4% of water, suggesting saturation of OER with a signi cant increase in the anodic current (ΔI = 16.42 µA) with k obs = 7.28 s -1 ( Figure S4, SI). However, the rst oxidation wave remains unaltered during the electrocatalysis, implying the stability of the ligand during water oxidation by catalyst 1. Further, a decrease in the catalytic current was observed with an increase in the scan rate ( Figure S5, SI), which indicates that the current is associated with the catalytic process.
The addition of water to the solution (1 mM) of diselenide 3, which is a ligand for zinc selenolate catalyst 1, increases the current by only ΔI = 0.12 µA ( Figure S28, SI). Moreover, the addition of water does not enhance the current, which implies the inability of diselenide ligand 3 to catalyze water oxidation. Similarly, ZnCl 2 solution (1 mM) does not catalyze the water oxidation as no change in the anodic current was observed ( Figure S29, SI), and instead, precipitation was observed in the electrochemical cell during the rst cycle of electrocatalysis.
Next, we sought to study the hydrogen evolution reaction (HER) from water in propylene carbonate solution by using glassy carbon (GC) working electrode. For this, the electrochemical experiment was performed in the presence of 1 mM of catalyst 1 and 0.25% of water under cathodic (negative) potential.
A successive increase of water in the cell results in an increase in the cathodic current ( Figure 2C). The maximum cathodic current 12 µA (ΔI = 9 µA) was elevated at 1.5% water concentration. However, the onset potential of -0.256 V vs. NHE was observed along with rate constant (k obs ) of 47.32 s -1 ( Figure S10, SI). Notably, the hydrogen evolution reaction (HER) from water in the presence of catalyst 1 under acidic conditions (using aqueous acetic acid and strong tri uoracetic acid, TFA) led to low cathodic current 14.8 µA (ΔI = 5.1 µA) and 3.2 µA (ΔI = 0.8 µA), ( Figure S14 and S16, SI) contrary to the earlier reported catalysts in which addition an acid increases the cathodic current. It seems that the hydrogen evolution reaction (HER) catalyzed by zinc selenolate catalyst 1 proceeds preferentially by the deprotonation from water and may not from the acid, and signi cantly low rate constant in a strong tri uoroacetic acid could be due to the poor stability of the catalyst 1 in TFA.
To gain a deeper insight into the catalytic activity, the kinetic activities have been studied for OER and HER. A plot of i cat /i p vs. [H 2 O] 1/2 was found to be linear, indicating bimolecular rst-order reaction kinetics ( Figure S4 and S10, SI). Similarly, a linear plot for i cat vs. [catalyst 1] was observed upon varying the concentration of zinc selenolate catalyst 1, which suggests rst-order reaction kinetics with respect to the catalyst 1 ( Figures 2B and 2D).
The stability of zinc selenolate catalyst 1 under oxygen and hydrogen evolution reaction conditions is con rmed by constant potential electrolysis (CPE) at an applied potential of 1.34 V (vs. NHE for OER) and -0.26 V (vs. NHE for HER) for 2h at GC electrode surface in which a signi cant change in current was not observed ( Figure 3A and 3B). Further, the stability of catalyst 1 under electrocatalysis was also con rmed by Electron Dispersive X-Ray Spectroscopy (EDXS) in which decomposition of catalyst 1 was not realized as a residue for Zn or zinc oxide/ zinc selenide was not observed in the spectra ( Figure 3C, EDXS, blue part), and also scanning electron microscopy (SEM) study of Pt-electrode does not show any change in the surface morphology ( Figures 3C). The UV-Visible study was also performed on the solution from electrolysis cell after 2 h of the bulk electrolysis and before the electrolysis. The characteristic absorbance at 380 nm of zinc selenolate catalyst 1 was nearly identical (red and blue lines in Figure 3D) to the one before electrolysis (black line) con rms that the structure of the catalyst 1 remains unchanged after the electrocatalysis. These results demonstrate that the bimetallic zinc selenolate complex 1 serves as a robust catalyst for water splitting in a homogeneous system. Also, the evolved oxygen from the water during electrocatalysis was then determined by CPE, reveals a Faradic e ciency of 79% for oxygen evolution ( Figure S22, SI).
Heavier analog mercury selenolate 4 also catalyzed oxygen evolution reaction (Figure, S23, SI), albeit low cathodic current 14 µA, and rate constant (0.038 s -1 ) ( Figure S24, SI), presumably attributed to monometallic nature, was observed in comparison to the bimetallic zinc selenolate catalyst 1. Further, mercury selenolate 4 failed to catalyze the hydrogen evolution reaction (HER) from water ( Figure S27). Mercury selenolate 4 was found unstable under negative potential as deposition was observed on the surface of the working electrode, presumably due to reduced mercury and could be due to feasible standard reduction potential ( 0.74V vs. SHE) of mercury ion to mercury than that of hydrogen ion to H 2.

Mechanistic Studies
Mechanistic insights on bimetallic zinc selenolate 1 catalyzed OER and HER from water were gained from the theoretical assessment of the likely reaction pathways supported by the mass analysis and control experiments. The free energies of intermediates along with possible HER and OER pathways, depicted in Scheme 3, were calculated at DFT/B3LYP/def2-TZVP level of theory.
Both reactions proceed via adsorption of two hydrogen-bonded water molecules, with one coordinating to Zn(a) and the other forming a hydrogen bond to bridging oxygen yielding 1a (ΔG = 54.91 kcal mol -1 ), the formation of which also con rmed by mass spectrometry (Figure S1, SI). In OER, diaqua species 1a undergoes proton-coupled electron transfer (PCET) to form 1b (con rmed by mass spectrometry Figure   S2, SI) with Zn(a)-OH and Zn(b)-OH 2 centers. Spin-density and NBO analysis ( Figure S30, SI) indicates that the electron is lost primarily from Se(a), consistent with the Se -2 /Se -1 oxidation peak seen in the CV. A second PCET, accompanied by intramolecular rearrangements, leads to selenenic acid 1c. Here selenium plays a crucial role in stabilizing the -OH bridging it to Zn(a), whereas the second -OH migrates to bridge the two Zn centers and bridging µ-phenolic oxygen become terminal. The next two successive PCET steps provide selenoxide species 1d and the intramolecular ZnO---H(N) hydrogen-bonded intermediate 1e, respectively. Poor stability of selenoxide (Se=O) bond attributed to weaker π-overlap of selenium with oxygen 53 and better leaving group tendency of selenium in the ligand 37,53 facilitate 1e to undergo an intramolecular rearrangement to 1f containing divalent selenium center and a peroxo linkage (r O-O = 1.48Å) between Zn centers and a phenolic ring. Subsequently, better ligation ability of water than oxygen to zinc would lead to oxygen evolution from 1f with the concomitant release zinc selenolate catalyst 1a.
The HER mechanism, also initiated by the formation of 1a, proceeds with the abstraction of a proton from water by Se(a) to form 1g under applied negative potential. Subsequently, H from -Se(a)H and -NH groups combine to evolve as H 2 yielding selone 1h. 37 An added electron then reduces 1h to the anion radical intermediate 1i where the electronic charge on Se (0.24 e -) and radical on N (0.33 e -) are in conjugation through the phenyl ring (0.39 e -) as con rmed using spin density (see inset of Scheme 3) and natural bond order (NBO) analysis. Next, the addition of water molecule and removal of OHanion effectively adds a proton to the system resulting in the radical intermediate 1m stabilized by conjugation with the phenyl ring. Subsequently, the sequential addition of electron and proton regenerates the water adsorbed state 1a (Scheme S3, SI).

Discussion
In summary, a novel bimetallic zinc selenolate has been synthesized and structurally characterized.
Further, bimetallic zinc selenolate electrocatalyzed both OER and HER from water in without adding acid or base and could complement TM catalysts and redox-active ligands. The synthesized complex catalyzes the water splitting via ligand assisted pathway, which bypasses the formation of metal hydrides during catalysis. However, ligand and ZnCl 2 alone are not able to electrocatalyze the water splitting; only by combining these two show the catalytic activity where the Lewis acidity of Zn(II) plays a vital role in the mechanism as it binds with a water molecule and then would help in the removal of a proton from ligated water and also tune the redox potential of catalyst. From the mechanism, it is clear that the presence of a selenium center is crucial for the electrocatalysis of water. The reaction mainly occurs at the selenium center for the four-electron and two-electron transfer in oxygen evolution reaction (OER) and hydrogen evolution reaction (HER), respectively, while the Lewis acidic Zn(II) center brings the water molecule near to the active selenium center. Furthermore, we noted that the oxygen and nitrogen heteroatoms present in 1 is not only crucial for the isolation of bimetallic zinc complex but also facilitate electron transfer in OER and HER as this not only serve as two electrons donor and acceptor due to the presence of the ortho-amino group but also selenium seems weak electrophile and avoid forming an undesirable stable selenium-oxygen bond. It is also evident from our DFT analysis that selenium is a crucial active site on bimetallic zinc selenolate electrocatalyst for both OER and HER. Further progress is being made to enhance the reactivity of ligand in the bimetallic selenolate complexes to synergistically activate the small molecules, as we have presented here.

Methods
Representative Procedure. To the stirred solution of schiff base diselenide 2 (248 mg, 0.45 mmol, 1 equiv.) in ethanol, we added sodium borohydride (76 mg, 2.0 mmol, 4 equiv) to generated in-situ selenol and stirred the solution up to 4 h at room temperature. Then we added zinc chloride (122 mg, 0.9 mmol, 2 equiv.) and stirred the solution for 2 h. After that, the solvent was removed by the rotatory evaporator, and the solid residue was washed with aqueous sodium bicarbonate solution several times to afford a light yellow colored novel bimetallic zinc selenolate complex 1 in (230 mg) 75% yield. Crystallization was done in DMSO water (2:1) mixture to afford yellow-colored crystals.
Electrochemistry. A potentiostat (SP-240 Bio-logic Instrument) was used for electrochemical measurements. The three-electrode electrochemical cell consisted of a Glassy carbon (3 mm Diameter) as the working electrode, a nonaqueous Ag/AgNO 3 (10 mM AgNO 3 ) as a reference electrode and a platinum wire as a counter electrode were used for the electrochemical measurements.
The electrochemical potential was converted relative to the normal hydrogen electrode (NHE; all potentials reported in this work are referenced to the NHE) following a literature protocol. Current  The chronoamperometry experiment has been performed in a stirring electrolyte solution in order to make the solution free from in-situ generated oxygen bubbles. A potential of 1.34 V vs. NHE (for OER) and − 0.26 V vs. NHE (for HER) has been chosen for the chronoamperometry experiment.
Computational Details. All the electronic structure calculations presented in this work were performed using density functional theory implemented in the TURBOMOLE 7.4 electronic structure package. Optimization of molecular geometry in the gas phase was carried out using B3LYP hybrid functional. The electronic con guration of the atoms was described with def2-TZVP basis set. For all the calculations, resolution of identity (RI) approximation with corresponding auxiliary basis set was used to speed up the calculations. The stability of the optimized geometries was con rmed by the absence of imaginary vibrational frequencies. Thermal corrections to Gibbs free energy, including zero-point energy, were obtained from vibrational frequency analysis at the same level of theory at 298.15 K and 1 atm within ideal-gas rigid-rotor harmonic-oscillator approximations.
Data availability. The authors declare that the data supporting the ndings of this study are available within the paper and the Supplementary Information, as well as from the authors upon request.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.