Structurally-New Hexadecanuclear Ni-Containing Silicotungstate with Catalytic Hydrogen Generation Activity

A structurally-new, carbon-free hexadecanuclear Ni-containing silicotungstate, [Ni16(H2O)15(OH)9(PO4)4(SiW9O34)3]19-, has been facilely synthesized using a one-pot, solution-based synthetic method systematically characterized by single-crystal X-ray diffraction and several other techniques. The resulting complex works as a noble-metal-free catalyst for visible-light-driven catalytic generation of hydrogen, by coupling with a [Ir(coumarin)2(dtbbpy)][PF6] photosensitizer and a triethanolamine (TEOA) sacrificial electron donor. Under minimally optimized conditions, a turnover number (TON) of 842 was achieved for TBA-Ni16P4(SiW9)3-catalyzed hydrogen evolution system. The structural stability of TBA-Ni16P4(SiW9)3 catalyst under photocatalytic conditions was evaluated by the mercury-poisoning test, FT-IR, and DLS measurements. The photocatalytic mechanism was elucidated by both time-solved luminescence decay and static emission quenching measurements.


Compound
Ni 16  In addition, the structure of polyoxoanion Ni 16 P 4 (SiW 9 ) 3 ( Figure 2a) can also be simplified into a ball-and-stick model, therein the {SiW 9 O 34 } moieties and the central {Ni 16 } unit are regarded as external pendulum and central node (Figure 2b), respectively. Three external pendulums and one central node can be assembled into a triangular geometry ( Figure 2c). Interestingly, it is noted that the molecular structural units of polyoxoanion Ni 16 P 4 (SiW 9 ) 3 can form a zigzag one-dimensional (1-D) chain connected by Na + counter cations along a axis ( Figure S1).

Characterization of Ni 16 P 4 (SiW 9 ) 3
The FT-IR spectrum of Na-Ni 16 P 4 (SiW 9 ) 3 was collected in a 2 wt% KBr pellet in the region of 4000 to 400 cm −1 ( Figure S2, black curve). The signal at 987 cm −1 are attributed to vibrations of W-O t . The W-O-W vibrations peaks are located at 889, 861, 810, and 684 cm −1 , while the absorption peak at 937 cm −1 is consistent with Si-O a vibrations. All the characteristic bands of the Na-Ni 16 P 4 (SiW 9 ) 3 structure were observed in the FT-IR spectrum, which is similar to that of the lacunary [A-α-SiW 9 O 34 ] 10− POM ligand ( Figure S2, blue curve). The replacement of Na + cations with tetrabutylammonium (TBA + ) retains the molecular skeleton of Ni 16 P 4 (SiW 9 ) 3 ( Figure S2, red curve), the corresponding vibrational signals of TBA + cation is well observed [67,68]. The UV-vis spectrum of polyoxoanion Na-Ni 16 P 4 (SiW 9 ) 3 exhibits a strong absorption peak in the UV region, which can be assigned as the oxygen-to-metal charge-transfer that are typically observed in the POM structures ( Figure S3). Thermogravimetric analysis (TGA) showed a weight loss of 10.5%, which was calculated to be 57 crystallization H 2 O molecules in one formula unit ( Figure S4). Due to incomplete substitution of Na + in the crystal, the amount of TBA + was determined by the TGA. The TGA of TBA-Ni 16 P 4 (SiW 9 ) 3 shows~2.97% weight loss~at 20-100°C and 27.5% weight loss at 100-750°C, roughly corresponding to about 15 H 2 O molecules and 14 TBA + cations to replace the initial Na + cations, respectively. The chemical compositions of polyoxoanion Na-Ni 16 P 4 (SiW 9 ) 3 were characterized by ICP-AES tests (see Experimental section). Then, XPS data was further collected to characterize the existence and oxidation states of Ni ( Figure S5a Figure S5d) elements in complex Na-Ni 16 P 4 (SiW 9 ) 3 ( Figure S5). For instance, the binding energies of the Ni 2p 3/2 and 2p 1/2 (with corresponding satellite peaks at 862.5 eV and 880.3 eV, Figure S5a) peaks were located at 856.0 and 874.0 eV ( Figure S5), respectively, indicating the +2 oxidation state of the Ni centers in the cluster. The XPS results are in good consistence with the BVS calculations. In addition, SEM/EDX results also revealed the microscopic morphology of the Na-Ni 16 P 4 (SiW 9 ) 3 crystal and the existence of Si, Ni, P, and W elements ( Figure S6). The calculated atomic ratio of Ni/W (1:1.50) from the EDX results is in good agreement with the theoretical value (1:1.69) ( Figure S7). In addition, the PXRD pattern of Na-Ni 16 P 4 (SiW 9 ) 3 matched well with the simulated diffraction pattern, also indicating the phase purity of the title compound ( Figure S8).

HOMO and LUMO Investigation
Electrochemical measurements and UV-vis absorption spectrum were conducted to calculate the HOMO and LUMO energy levels of the catalyst. Estimated from the UV-vis absorption spectrum, the energy bandgap of the crystal was obtained, thus the HOMO was calculated to be the sum of the energy bandgap and LUMO ( Figure S9). The LUMO orbital energy for the TBA-Ni 16 P 4 (SiW 9 ) 3 was measured as −3.31 eV. According to the UV-vis-NIR spectra of K-M function vs. energy (eV), the energy gap was 2.40 eV. Therefore, the HOMO orbital energy was −5.71 eV. In the previous research of our group, the orbital energy of photosensitizer [Ir(coumarin) 2 (dtbbpy)] + was calculated. Its LUMO and HOMO energies were −3.28 and −5.42 eV, respectively [69]. Therefore, the LUMO electron of excited state [Ir(coumarin) 2 (dtbbpy)] + * can be transferred to the LUMO orbitals of the catalyst, providing the possibility for establishing a photocatalytic hydrogen production system.

Photocatalytic Hydrogen Production and Evaluation of Catalyst Stability
Considering the vital challenges of energy shortage and environmental problems faced by modern mankind, the development of clean and renewable energy alternatives has been attracting tremendous research attention. Photocatalytic hydrogen production driven by solar energy represents a promising way to produce clean secondary energy carriers. Herein, the visible light-driven H 2 production activity of TBA-Ni 16 P 4 (SiW 9 ) 3 was investigated in a well-established three-component system by using [Ir(coumarin) 2 (dtbbpy)] + [69] as the photosensitizer, TBA-Ni 16 P 4 (SiW 9 ) 3 as the WRC, and TEOA as electron donor in a mixed CH 3 CN/DMF (v/v = 1/3) solvent. The reaction solution was exposed to 400 nm visible-light irradiation at room temperature. All turnover numbers (TONs) were calculated with respect to TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst. The effect of each component on photocatalytic activity was evaluated by different control experiments. As shown in Figure 3a, the catalytic system in the absence of photosensitizer, sacrificial reagent, or catalyst produces negligible H 2 production under otherwise identical conditions. In addition, the replacement of the catalyst TBA-Ni 16 P 4 (SiW 9 ) 3 with the lacunary {SiW 9 } POM stabilizing ligand causes very low amounts of H 2 production, proving the vital role of Ni sites. Moreover, the catalytic system using stoichiometric equivalents of NiCl 2 (320 µM) as that of 20 µM TBA-Ni 16 P 4 (SiW 9 ) 3 shows a remarkable decrease in hydrogen production. These results demonstrate that photosensitizer, sacrificial agent, and catalyst are all necessary component in the photocatalytic process. More importantly, the structural skeleton of TBA-Ni 16 P 4 (SiW 9 ) 3 is essential for efficient catalysis because the unique molecular structure of TBA-Ni 16 P 4 (SiW 9 ) 3 polyoxoanion can work as electron reservoir to effectively store electrons in the electron-deficient POM ligands and, in the meantime, supply electrons to the catalytically active Ni centers, thereby leading to the high catalytic efficiency of TBA-Ni 16 P 4 (SiW 9 ) 3 . The different concentrations of each component also significantly affect H 2 production (Figure 3b-d). Increasing the concentration of [Ir(coumarin) 2 (dtbbpy)] + photosensitizer from 0.1 to 0.3 mM enhances the H 2 yield from~3.6 to~100 µmol, corresponding to a TON change from~30 to~842 (Figure 3b). The catalytic performance of this TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst is comparable to that of some known Ni-containing POMs under homogeneous catalytic systems using Ir/Ru-based photosensitizers (Table  S3). While adjusting the concentration of TEOA from 0.05 M to 0.25 M, the H 2 production increases from~20 to~100 µmol. In addition, the H 2 yield was enhanced from~0.9 tõ 164 µmol as the concentration of TBA-Ni 16 P 4 (SiW 9 ) 3 changed from 5 to 20 µM. Based on the above experimental results, the optimal combination of the [Ir(coumarin) 2 (dtbbpy)] + photosensitizer, TEOA electron donor, and TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst are vital for highly efficient photocatalytic H 2 evolution.
The stability of catalyst agent has been a general concern in molecular photocatalytic systems. In this paper, the stability of TBA-Ni 16 P 4 (SiW 9 ) 3 was assessed using a range of optical methods and experimental evaluation. To investigate whether the TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst was decomposed into the Ni nanoparticles, we have carried out a mercury-poisoning test by the addition of 20 mg Hg to the photocatalytic solution. The addition of Hg does not significantly affect the hydrogen production, implying the integrity of TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst during photocatalysis (Figure 3a). Moreover, to further characterize the stability of TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst, the post-reaction catalyst was isolated in the form of [Ru(bpy) 3 ] x -Ni 16 P 4 (SiW 9 ) 3 adducts after photocatalysis by adding cationic [Ru(bpy) 3 ] 2+ species. FT-IR spectra of isolated [Ru(bpy) 3 ] x -Ni 16 P 4 (SiW 9 ) 3 adducts reveal almost no changes before and after photocatalysis for 6 h ( Figure S10), implying the decent molecular stability of the TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst. The DLS measurement illustrated a signal centered at~1.7 nm for the TBA-Ni 16 P 4 (SiW 9 ) 3 (20 µM) system after 6 h of catalysis ( Figure S11), which is consistent with the size of TBA-Ni 16 P 4 (SiW 9 ) 3 , about 18.98 Å, implying the integrity of the TBA-Ni 16 P 4 (SiW 9 ) 3 polyoxoanion.

Photocatalytic Mechanistic Studies
It is known that the photoexcited photosensitizer can work as both oxidizing and reducing species in the typical photocatalytic systems. Therefore, to reveal the photocatalytic mechanism, the quenching experiments of [Ir(coumarin) 2 (dtbbpy)] + by TEOA and TBA-Ni 16 P 4 (SiW 9 ) 3 has been performed in CH 3 CN/DMF using both steady-state emission quenching and time-resolved luminescence decay spectroscopy. As shown in Figure 4, a strong emission band in the region of 500−750 nm was observed upon excitation of [Ir(coumarin) 2 (dtbbpy)] + (λ e = 460 nm), and the emission intensity of [Ir(coumarin) 2 (dtbbpy)] + was progressively quenched with the addition of TBA-Ni 16 P 4 (SiW 9 ) 3 (0-60 µM) and TEOA (0-0.25 M). Luminescence quenching rate constants can be derived by the Stern-Volmer plot using a linear function ( Figure S12). The quenching rate constant (k rq ) for the reductive pathway by TEOA was calculated as 2.55 × 10 6 M −1 ·s −1 , while the oxidative quenching rate constant (k oq ) by TBA-Ni 16 P 4 (SiW 9 ) 3 was 6.27 × 10 9 M −1 ·s −1 . It is clear that the k oq value is three orders of magnitude higher than the k rq value, which can be attributed to the strong electrostatic interaction between positively-charged [Ir(coumarin) 2 (dtbbpy)] + photosensitizer and negatively-charged TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst. However, in the typical photocatalytic hydrogen evolution experiments, the concentrations of TEOA and TBA-Ni 16 P 4 (SiW 9 ) 3 were 0.25 M and 20 µM, respectively. Therefore, the corresponding quenching rates can be calculated by multiplying the values of quenching rate constants by the concentrations of quenchers, leading to the values of 0.6375 × 10 6 s −1 by TEOA and 1.254 × 10 5 s −1 by TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst. Such relatively higher quenching rates by TEOA revealed that that the reductive pathway was still the dominant one during photocatalysis ( Figure S13). By using the time-resolved fluorescence spectroscopy, the decay kinetics of the excited state [Ir(coumarin) 2 (dtbbpy)] + * was also investigated. The experimental phenomenon that both TBA-Ni 16 P 4 (SiW 9 ) 3 and TEOA can accelerate the decay of [Ir(coumarin) 2 (dtbbpy)] + * luminescence was obvious ( Figure 5). The single-exponential fitting of decay kinetics of [Ir(coumarin) 2 (dtbbpy)] + * yielded a lifetime of~1181.38 ns, which was further decreased to~906.22 and~967.85 ns in the presence of TEOA and TBA-Ni 16 P 4 (SiW 9 ) 3 , respectively. These results clearly revealed that both reductive and oxidative quenching processes existed during photocatalysis and the reductive quenching process is the dominant one, thus agreeing with the steady-state emission quenching results. According to the above mechanistic analyses, the possible photocatalytic process was proposed as follows. Under the light irradiation, the photons were absorbed by the [Ir(coumarin) 2 (dtbbpy)] + photosensitizer, generating the excited state [Ir(coumarin) 2 (dtbbpy)] + *. In addition to the oxidative quenching of [Ir(coumarin) 2 (dtbbpy)] + * by TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst, the photoexcited states can also be reductively quenched by TEOA to form oneelectron-reduced [Ir(coumarin) 2 (dtbbpy)] species. The TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst can be reduced by accepting electrons from this reduced [Ir(coumarin) 2 (dtbbpy)] species. During photocatalysis, the lacunary {SiW 9 } POM building blocks and transition metals act as electron storage mediator and catalytic active sites, respectively. The lacunary {SiW 9 } POM ligands can be reduced by reversibly storing multiple electrons and protons, then the electrons could be continuously utilized by the Ni active centers to effectively catalyze hydrogen evolution.

Methods and Materials
All chemicals were used as received without further purification, unless otherwise specified. Trivacant lacunary POM Na 10 [A-α-SiW 9 O 34 ]·18H 2 O was synthesized according to the literature method [70]. Single-crystal X-ray crystallography was performed on a Bruker APEXII DUO (Bruker, Karlsruhe, Germany) diffractometer CCD detector operated at 40 kV and 40 mA with Mo Kα radiation (λ = 0.71073 Å). Fourier transform infrared (FT-IR) were recorded on a Bruker TENSOR II spectrometer (Bruker, Karlsruhe, Germany) with~2 wt% KBr pellets. Ultraviolet-visible (UV-Vis) absorption spectra were measured by using a Techcomp UV 2600 (Techcomp, Shanghai, China) spectrophotometer. Scanning electron microscopy (SEM) associated with energy-dispersive X-ray spectroscopy (EDX) data were collected on a JSM-7500F (JEOL, Tokyo, Japan) instrument. ICP-AES was conducted on an Agilent ICP-AES 5110 (Agilent, Santa Clara, CA, USA) to analyze the elemental composition of the resulting complex, which contains Ni, Si, P, W, and Na. Thermogravimetric data (TGA) were collected on a HITACHI TG/DTA7300 (HITACHI High-Technologies, Yamaguchi, Japan) instrument from 20 to 800 • C under N 2 atmosphere. X-ray photoelectron spectroscopy (XPS) measurements were performed on a PHI 5000 Versaprobe III (Ulvac-Phi, Osaka, Japan) instrument. The X-ray powder diffraction pattern was collected on a Shimadzu XRD-6000 instrument (Shimadzu, Kyoto, Japan).
The TBA + salt of Ni 16 P 4 (SiW 9 ) 3 (TBA-Ni 16 P 4 (SiW 9 ) 3 ) was synthesized using the following procedure: the crystalline Na-Ni 16 P 4 (SiW 9 ) 3 (0.2 g, 17.86 µmol) sample was dissolved in 5 mL of H 2 O, to which a solution of TBA bromide (5 g, 15.6 mmol) in 5 mL of 0.5 M sodium acetate buffer (pH 4.8) was added and then stirred vigorously for 0.5 h. The light green precipitate was formed, separated by centrifugation, washed with ice water to remove excess TBA bromide, and finally dried under vacuum. The final TBA-Ni 16 P 4 (SiW 9 ) 3 product was collected and characterized by FT-IR and TGA. The empirical molecular formula of the TBA + salt of Ni 16 P 4 (SiW 9 ) 3 was calculated as TBA 14

Single-Crystal X-Ray Crystallography
An appropriate high-quality crystal (0.19 × 0.20 × 0.21 mm 3 ) was selected and encapsulated in a single crystal tube with Vaseline on both ends for data collection at 298 K. The data of Na-Ni 16 P 4 (SiW 9 ) 3 crystal was collected on a Bruker APEXII diffractometer. The APEX 3 software (APEX3 v2016.1-0, Bruker, Karlsruhe, Germany) was installed on the diffractometer for data collection, indexing, and initial cell refinements [71]. Optimal reflections have been collected for high-quality frame integration and final cell refinements using SAINT software (APEX3 v2016.1-0, Bruker, Karlsruhe, Germany) [72]. The Olex 2 software (Olex2 v1.3, England) equipped with Superflip structure solution program was used to solve the crystal structures, which were further refined by least squares using ShelXL [73][74][75]. All these non-H atoms were refined with anisotropic thermal parameters. The hydrogen atoms were located by bond valence sum (BVS) calculations. Details of the crystallographic data and analyses for the compound Na-Ni 16 P 4 (SiW 9 ) 3 are given in Table 1, and important bond lengths as well as corresponding bond valence sum (BVS) calculations are summarized in Tables S1 and S2. The Cambridge Crystallographic Data Centre (CCDC) number of Na-Ni 16 P 4 (SiW 9 ) 3 is deposited as 2223896. Further details on the crystal structure investigations may be obtained from http://www.ccdc.cam.ac.uk/deposit (accessed on 12 February 2022) on quoting the depository number as mentioned.

Photocatalytic Hydrogen Evolution Tests
Photocatalytic hydrogen production was conducted in degassed CH 3 CN-DMF (v/v = 1/3) solution containing triethanolamine (TEOA) electron donor, H 2 O proton source, iridium complexes ([Ir(coumarin) 2 (dtbbpy)][PF 6 ]) photosensitizers [69], and Ni-substituted polyoxometalate TBA-Ni 16 P 4 (SiW 9 ) 3 as the catalyst. The reaction solution was deaerated with Ar/CH 4 (v/v = 4/1), the internal standard of CH 4 was used for better quantification. The degassed mixture was illuminated using a 300 W Xe-lamp (PerfectLight, Beijing, China) equipped with a 400 nm cutoff filter at room temperature with constant stirring. Gas chromatograph (Thermo GC7900, thermal conductivity detector (TCD), Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the hydrogen in the reaction headspace using a 5 Å molecular sieve capillary column. All turnover numbers (TONs) were calculated based on catalyst TBA-Ni 16 P 4 (SiW 9 ) 3 . Control experiments were conducted under similar experimental conditions by removing one component one at a time. Additional control experiments were conducted by replacing the TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst with NiCl 2 or trivacant {SiW 9 } POM under otherwise identical conditions.

Electrochemical Measurements
The cyclic voltammetry was conducted on a CHI660E (Chinstruments, Shanghai, China) instrument using 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF 6 ) as electrolyte, glassy carbon as working electrode, Pt wire as counter electrode, and nonaqueous Ag/Ag + as reference electrode. The scan rate of CV was 50 mV/s. The E ox or E red was measured by using the internal standard substance ferrocene (E ox (Fc/Fc+) = 0 V vs. Ag/Ag+) [76]. The working electrode was treated by grinding with 0.3 µm and 0.05 µm alumina for about 4 min, flushing with deionized water, sonicating with acetone for 2 min and drying with nitrogen gas flow. The solution was degassed to remove oxygen. At the time of measurement, the degassing device is placed above the liquid level to avoid large disturbance. The HOMO and LUMO energy levels was calculated by the following equations:

Steady-State and Time-Resolved Fluorescence Decay Measurements
The Edinburgh Instruments FS5 (Edinburgh Instruments, Livingston, UK) spectrofluorometer was used to test the steady-state luminescence quenching spectra and time-resolved luminescence decay kinetics. The solvent for photoluminescence decay measurements was mixing CH 3 CN/DMF with a volume ratio of 1:3. The different concentrations of TBA-Ni 16 P 4 (SiW 9 ) 3 or TEOA were degassed with Ar for 10 min to avoid the influence of oxygen before experiment. An intense emission band of [Ir(coumarin) 2 (dtbbpy)] + at 450-750 nm (λ excitation = 450 nm) was recorded and an EPL-450 picosecond pulsed diode laser system (pulse output 450 nm) was used to measure the emission lifetime at the emission maximum.

Conclusions
In summary, we reported the successful synthesis of a structurally-new, carbon-free hexadecanuclear Ni-containing silicotungstate, [Ni 16 6 ] photosensitizer and a triethanolamine (TEOA) sacrificial electron donor, the title TBA-Ni 16 P 4 (SiW 9 ) 3 polyoxoanion works as a noble-metal-free catalyst for hydrogen generation under visible light irradiation, achieving a turnover number (TON) of 842 under minimally optimized conditions. The mercury-poisoning test, FT-IR spectra of the isolated [Ru(bpy) 3 ] x -Ni 16 P 4 (SiW 9 ) 3 adducts, and the DLS measurements revealed the structural stability of TBA-Ni 16 P 4 (SiW 9 ) 3 catalyst under photocatalytic conditions. More importantly, both time-solved luminescence decay and static emission quenching measurements elucidated the photocatalytic mechanism, confirming the existence of both reductive and oxidative quenching pathways with the reductive quenching pathway as the dominant one. This work presents another good example of using Ni-substituted POMs as efficient hydrogen evolution catalyst, which might provide insights for future design of additional high-nuclearity metal-containing POMs as catalysts for solar energy conversion.

Conflicts of Interest:
The authors declare no conflict of interest.
Sample Availability: Samples of the compounds are available from the authors.