A Hexadecanuclear Cobalt-Added Tungstogermanate Containing Counter Cobalt Hydrates: Synthesis, Structure and Photocatalytic Properties

The synthesis and exploration of the properties of structurally-new polyoxometalates (POMs) has been attracting considerable research interest. In this work, a hexadecanuclear cobalt-added tungstogermanate, H31(NH4)5Na16{CoⅢ(H2O)6}4{[CoⅡ4(μ3-OH)3(PO4)]4(A-α-GeW9O34)4}2·23-H2O (1), was synthesized under hydrothermal conditions and characterized by various techniques. Compound 1 can effectively drive the heterogeneous photocatalytic hydrogen evolution reaction in the presence of [Ir(ppy)2(dtbbpy)][PF6] as the photosensitizer, with triethanolamine (TEOA) and N-Hydroxy succinimide (NHS) used as the dual sacrificial reagents. Control experiments revealed the important role of NHS in enhancing the hydrogen-evolution activities. Under optimal catalytic conditions, a hydrogen yield of 54.21 μmol was achieved after 10-h photocatalysis, corresponding to a hydrogen evolution rate of 1807.07 μmol·g−1·h−1. Stability studies demonstrated that catalyst 1 can be isolated and reused for three successive photocatalytic cycles with negligible decline of the H2 yield, indicating the stability and recycling robustness of catalyst 1.

Among various TMAPs, the synthesis and application of high-nuclearity, cobaltadded polyoxometalates (CoAPs) remain one of the most attractive research directions for global researchers. By controlling reaction conditions, lacunary tungsten-oxo clusters can be made to undergo diverse configuration changes. Different lacunary tungstenoxo clusters possess contrasting coordination abilities, which can therefore induce cobalt The chemicals utilized in this work are commercially obtained and used without further purification. The tri-vacant POMs K 8 Na 2 [A-α-GeW 9 O 34 ]·25H 2 O is prepared according to the previous report [46,47]. The basic characterization instruments and methods are as follows: (1) A Bruker D8 Advance X-ray diffractometer (Bruker, Karlsruhe, Germany) equipped with Cu Kα radiation (λ = 1.54056 Å) scanning from 5 to 50 • was used to determine the powder X-ray diffraction (PXRD) patterns of 1. (2) The Nicolet iS10 FT-IR spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) was used for IR spectrum (400-4000 cm −1 ) measurements with KBr pellets. (3) The optical gap of 1 was measured and estimated from UV-Vis diffuse reflectance spectra (200-800 nm) using a Shimadzu UV3600 spectrometer (Shimadzu, Kyoto, Japan). (4) Thermal stability was confirmed by thermogravimetric analyses using a Mettler Toledo TGA/DSC 1100 analyzer (Mettler Toledo, Zurich, Switzerland). The testing temperature ranges from 25 to 1000 • C in air at a heating rate of 10 • C·h -1 . (5) Inductively coupled plasma mass spectrometry was measured on an ICP-OES (iCAP 7400, Thermo, Waltham, MA, USA) to complete the elemental analysis and post-catalytic characterization. was introduced into 5 mL of deionized water. After stirring for 30 min, the pH of the solution was regulated to 9.50 by adding 4M of NaOH solution. The resulting purple-red turbid mixture was transferred into a 25 mL, Teflon-lined, stainless-steel autoclave and kept at 100 • C for 3 days. Pink spindle crystals were obtained after cooling the solution to room temperature. The crystals were further purified by washing and sonicating with deionized water. (Yield: 21%, based on K 8

X-ray Crystallography
A translucent pink crystal of 1 was sealed into a glass tube with Vaseline to collect the diffraction data by a Gemini A Ultra CCD diffractometer equipped with graphite monochromated Mo Kα (λ = 0.71073 Å) and radiation of 296(2) K. The structure was solved using the direct method and refined using full-matrix least-squares, which were fit using the F 2 method in Olex2 [48] and the SHELX-2008 program package [49,50]. The contribution of the disordered water molecules to the diffraction data of 1 was examined using the SQUEEZE method in PLATON [51]. The crystalline water molecules cannot be fully identified by refinement ( Table 1). The number of water molecules was quantified based on the TGA data shown in Figure S1, where 23 lattice water molecules exist in the structure. To obey the charge balance motif, 31 protons were added. The elemental compositions of compound 1 were further characterized by ICP-OES tests, which correlate with the single-crystal measurement results. The CCDC number of 1 is deposited as 2262758. Further data can be obtained from http://www.ccdc.cam.ac.uk/, accessed on 10 June 2023.

Photocatalytic Water Reduction Tests
The whole photocatalytic reaction system was comprised of acetonitrile/N,N-dimethyl formamide (v/v = 1:3) as the mixed solvent, [Ir(ppy) 2 (dtbbpy)][PF 6 ] as the photosensitizer, TEOA and NHS as the dual electron sacrificial reagents, 1 as the catalyst, and distilled water as the proton donor. The above components were transferred and sealed into the photoreactor. The reactor was deaerated by CH 4 /Ar (v/v = 4:1, CH 4 as the internal standard substance) in a mixed atmosphere for ten minutes, and then transferred to a PCX-50C multi-Nanomaterials 2023, 13,2009 4 of 12 channel photocatalytic reactor equipped with a white LED light plate (400-800 nm, electric power: 10 W) for photocatalytic reaction. During the photocatalysis, the reaction solution was continuously stirred at 400 rpm. The reaction temperature was kept at 25 • C by a water-circulating system. At an interval of two hours, 125 µL mix gas was extracted and qualitatively and quantitatively analyzed by a GC-979011 gas chromatograph analyzer. The whole photocatalytic reaction system was comprised of acetonitrile/N,N-dimethylformamide (v/v = 1:3) as the mixed solvent, [Ir(ppy)2(dtbbpy)][PF6] as the photosensitizer, TEOA and NHS as the dual electron sacrificial reagents, 1 as the catalyst, and distilled water as the proton donor. The above components were transferred and sealed into the photoreactor. The reactor was deaerated by CH4/Ar (v/v = 4:1, CH4 as the internal standard substance) in a mixed atmosphere for ten minutes, and then transferred to a PCX-50C multi-channel photocatalytic reactor equipped with a white LED light plate (400-800 nm, electric power: 10 W) for photocatalytic reaction. During the photocatalysis, the reaction solution was continuously stirred at 400 rpm. The reaction temperature was kept at 25 °C by a water-circulating system. At an interval of two hours, 125 µL mix gas was extracted and qualitatively and quantitatively analyzed by a GC-979011 gas chromatograph analyzer.   4 } is connected to the second layer that is a tri-phosphate-modified clover-like nonanuclear cobalt-oxo cluster {Co 9 (µ 3 -OH) 9 O 19 (PO 4 ) 3 } (Figure 2d,e). Three hexacoordinated Co 2+ are linked together by edgesharing modes, creating one of the three "leaves" of the "clover". The three "leaves" are held together by one µ 4 -O atom. Three {PO 4 } linkers are located in the same plane and fixed to the gap between the three "leaves" of the {Co 9 (µ 3 -OH) 9  links to {PO4} via three µ2-O atoms. The residual µ4-O of {PO4} is connected to the second layer that is a tri-phosphate-modified clover-like nonanuclear cobalt-oxo cluster {Co9(µ3-OH)9O19(PO4)3} (Figure 2d,e). Three hexacoordinated Co 2+ are linked together by edgesharing modes, creating one of the three "leaves" of the "clover". The three "leaves" are held together by one µ4-O atom. Three {PO4} linkers are located in the same plane and fixed to the gap between the three "leaves" of the {Co9(µ3-OH)9O19(PO4)3}, connected by P-µ2-O-Co, P-µ3-O-Co, and P-µ4-O-Co linkages. Each of the three {PO4}s in the second layer can share a µ2-O to connect with the pedestal position Co of the third layer. The third layer is {Co4(µ3-OH)3O14} (Figure 2f,g), which is the traditional defective cubane-like configuration tetranuclear cobalt-oxo cluster. The vertex position Co of the third layer was linked to the center of the second layer through an edge-sharing mode.    (Figure 4d). The packing mode of 1 in the bc plane is shown in Figure S3.  (Figure 4d). The packing mode of 1 in the bc plane is shown in Figure S3.

Results and Discussion
From another view, {[Co4(µ3-OH)3(PO4)]4(A-α-GeW9O34)4} can also be presented as a tri-nested structure by simplifying {A-α-GeW9O34} as orange balls, cobalt ions as pink balls, and {PO4}s as yellow balls (Figure 4a). The four {A-α-GeW9O34} are located in the vertices of a tetrahedron, and the distance between each {A-α-GeW9O34} varies from 11.020 to 11.218 Å (Figure 4b). The arrangement of the four {PO4} linkers also shows a tetrahedron configuration. The distances between each {PO4} are in the range of 5.341-5.392 Å ( Figure  4c). In the simplified {Co16(µ3-OH)12O36(PO4)4}, vertices of the four external {Co4} tetrahedrons can form an internal {Co4} tetrahedron. The distances between each Co Ⅱ are in the range of 3.017-3.811 Å (Figure 4d). The packing mode of 1 in the bc plane is shown in Figure S3.

FT-IR Measurement
As illustrated in Figure S4, the FT-IR spectrum of 1 involves four major characteristic vibrational groups. First, the wide peaks in 3423-3203 cm −1 and 1633 cm −1 can be attributed to the existence of lattice and coordination water in 1. The distinct vibrational peak observed at ~1400 cm −1 could be attributed to the existence of NH4 + counter cations. In the fingerprint region, the absorption peak centered at 1079 cm -1 is ascribed to the ν(P-O) stretching vibration mode. The typical vibration peaks of the tri-vacant {A-α-GeW9O34} precursor (e.g., ν(Ge-O) ν(W-Ot), ν(W-Ob), and ν(W-Oc)) are observed at 933 (vs), 869 (vs), 806 (vs), and 688 cm −1 , respectively.

Powder X-ray Diffraction Patterns
To confirm the phase purity of 1, it is noted that there is agreement between the experimental PXRD pattern and the simulated pattern from the single-crystal X-ray diffraction data ( Figure S5). The modest shift in peak intensity may be attributed to a change in the favored orientation of the powder sample during the PXRD data-collecting process.

FT-IR Measurement
As illustrated in Figure S4, the FT-IR spectrum of 1 involves four major characteristic vibrational groups. First, the wide peaks in 3423-3203 cm −1 and 1633 cm −1 can be attributed to the existence of lattice and coordination water in 1. The distinct vibrational peak observed at~1400 cm −1 could be attributed to the existence of NH 4 + counter cations. In the fingerprint region, the absorption peak centered at 1079 cm -1 is ascribed to the ν(P-O) stretching vibration mode. The typical vibration peaks of the tri-vacant {A-α-GeW 9 O 34 } precursor (e.g., ν(Ge-O) ν(W-Ot), ν(W-Ob), and ν(W-Oc)) are observed at 933 (vs), 869 (vs), 806 (vs), and 688 cm −1 , respectively.

Powder X-ray Diffraction Patterns
To confirm the phase purity of 1, it is noted that there is agreement between the experimental PXRD pattern and the simulated pattern from the single-crystal X-ray diffraction data ( Figure S5). The modest shift in peak intensity may be attributed to a change in the favored orientation of the powder sample during the PXRD data-collecting process.

UV-Vis Diffuse Reflectance Test
Compound 1 exhibited strong absorption in the range of 200-800 nm. The wide absorption peak at 200-400 nm corresponds to the O-to-W charge transfer, while the absorption of 500-600 nm is attributed to the d-d transitions of Co centers. The obtained absorbance-wavelength data is further processed according to the Kubelka-Munk formula α/S = (1 − R) 2 /(2R) (R = reflection coefficient, α = absorption factor, S = scattering factor) to make the α/S-Energy plot, and the tangent plot is made in the near-linear range of the obtained curve; the intersection point with α/S = 0 is the optical band gap value. As shown in Figure S6, the optical gap of compound 1 is estimated as 2.65 eV.

Photocatalytic Hydrogen Generation
Given the poor solubility of 1 in water, N,N-dimethylformamide, or acetonitrile, 1 was employed as a heterogeneous catalyst in the visible-light-driven hydrogen generation reaction. Various control experiments confirmed that the existence of the catalyst, photosensitizer, and sacrificial agent was indispensable for efficient photocatalysis (Figure 5a, Table S1). It is noted that the very small amount of hydrogen (0.42 µmol) was produced from the control experiment containing Ir photosensitizer and TEOA sacrificial donor, which is commonly observed in the literature [42][43][44]54]. Commonly, one type of electron donor is used in photocatalytic HER systems. Herein, when TEOA was used as the only sacrificial agent, only 6.81 µmol of H 2 evolved. In 2019, Mialane's group reported a tetradecanuclear nickel-added tungsten-oxo cluster [(A-α-SiW 9 O 34 )Ni 14 (AleH) 5 (Ale) 2 (H 2 O) 11 -(OH) 7 ] 12− (Ale = alendronate), which can act as a heterogeneous catalyst to drive the photocatalytic hydrogen evolution reaction in the presence of TEOA, 1-benzyl-1,4-dihydronicotinamide (BNAH), and [Ir(ppy) 2 (dtbbpy)][PF 6 ]. Therein, BNAH served as the sacrificial electron donor in the system, and TEOA serves as a proton donor in the anhydrous system. The catalytic system obtained 18 mmol H 2 after 7-h illumination [55]. Clearly, an additive sacrificial electron donor may improve the catalytic performance. Therefore, NHS was selected as an additive electron donor for the photocatalytic reaction in the present study. With the addition of NHS, other components showed similar control effects as that of the NHS-free system. To verify the potential proton donor in this dual sacrificial reagent system, an anhydrous system under otherwise identical conditions was tested and showed no H 2 evolution (Figure 5a 6 ]. After 10-h photocatalysis, the H 2 yield of 54.21 µmol was obtained, corresponding to the highest hydrogen evolution rate of 1807.07 µmol·g −1 ·h −1 . A nine-fold increase in hydrogen yield was achieved by introducing NHS into the catalytic system compared to that of NHS-free catalytic system (Figure 5a). To better illustrate the strong catalytic performance of compound 1, we have summarized the photocatalytic results of reported systems using cobalt-added tungsten-oxo clusters. Compared to the reported studies (Table S2), compound 1 exhibited one of the best photocatalytic performances in the absence of a noble metal co-catalyst (e.g., Pt). Also, to our knowledge, this is the first time that cobalt-added tungstogermanate has been investigated for visible-light-driven hydrogen production in the dual sacrificial reagent system. Considering that 1 contains Co 2+ , {GeW 9 O 34 } and {PO 4 }, additional control experiments using molar equivalents of Co(NO 3 ) 2 ·6H 2 O, K 8 Na 2 [A-α-GeW 9 O 34 ], and Na 2 HPO 4 as 3 mg of 1 have been further conducted to better understand the photocatalytic performance. As shown in Figure 5b In addition, it is worth mentioning that the color of the reaction system rapidly changed from bright yellow to dark blue upon light irradiation, simultaneously coupling the linear hydrogen evolution with the irradiation time. In the meanwhile, catalyst 1 was also transformed into the heteropoly blue intermediates, indicating the reduction of 1 during photocatalysis, which is commonly observed in the literature reports [42,54]. When the reaction system was moved to a dark environment, hydrogen production stopped immediately. The same phenomenon appeared in the light ON-OFF experiments (Figure 5c). After exposure to air, the heteropoly blue intermediates will return to their original pink color. The rapid formation and disappearance of heteropoly blue intermediates show the strong and reversible multielectron storage abilities of POMs. As discussed in previous reports, heteropoly blue intermediates act as an active species by receiving multiple electrons in the vacant W 5d orbitals, which promote the hydrogen evolution process [42,54]. The catalytic activity is also related to the amount of catalyst used. When the catalyst dosage was 1, 3, and 6 mg, the hydrogen yield reached 24.07, 54.21, and 51.44 µmol in 10 h, corresponding to the hydrogen production rate of 2407.62, 1807.07, and 857.46 µmol·g −1 ·h −1 , respectively (Figures 5d and S7). In addition, the concentration of photosensitizers is also a significant factor to the enhancement of hydrogen production. When the concentration of [Ir(ppy) 2 (dtbbpy)][PF 6 ] was changed from 0.10 to 0.25 mM, the hydrogen evolution exhibited linear growth with reaction time, yielding 11.04 to 71.84 µmol H 2 after 10-h. The hydrogen evolution rates ranged from 924.47 to 2394.91 µmol·g −1 ·h −1 (Figure 5e). Moreover, because the experiments varied the concentration of TEOA from 0.03 to 0.06 M, the hydrogen production ranged from 35.72 to 54.21 µmol. The calculated hydrogen evolution rate ranged from 1190.69 to 1807.07 µmol·g −1 ·h −1 (Figure 5f). It is noted that varying the concentration of NHS from 0.075 mM to 0.300 mM barely changed the H 2 yield, which maintained a constant concentration of TEOA (0.060 M) ( Figure S8). In addition, it is worth mentioning that the color of the reaction system rapidly changed from bright yellow to dark blue upon light irradiation, simultaneously coupling the linear hydrogen evolution with the irradiation time. In the meanwhile, catalyst 1 was also transformed into the heteropoly blue intermediates, indicating the reduction of 1 during photocatalysis, which is commonly observed in the literature reports [42,54]. When the reaction system was moved to a dark environment, hydrogen production stopped immediately. The same phenomenon appeared in the light ON-OFF experiments (Figure 5c). After exposure to air, the heteropoly blue intermediates will return to their original pink color. The rapid formation and disappearance of heteropoly blue intermediates show the strong and reversible multielectron storage abilities of POMs. As discussed in previous reports, heteropoly blue intermediates act as an active species by receiving multiple electrons in the vacant W 5d orbitals, which promote the hydrogen evolution process [42,54]. The catalytic activity is also related to the amount of catalyst used. When the catalyst dosage was 1, 3, and 6 mg, the hydrogen yield reached 24.07, 54.21, and 51.44 µmol in 10 h, corresponding to the hydrogen production rate of 2407.62, 1807.07, and 857.46 µmol·g −1 ·h −1 , respectively (Figures 5d and S7). In addition, the concentration of photosensitizers is also a significant factor to the enhancement of hydrogen production. When the concentration of [Ir(ppy)2(dtbbpy)][PF6] was changed from 0.10 to 0.25 mM, the hydrogen evolution exhibited linear growth with reaction time, yielding 11.04 to 71.84 µmol H2 after 10-h. The hydrogen evolution rates ranged from 924.47 to 2394.91 µmol·g −1 ·h −1 (Figure 5e). Moreover, because the experiments varied the concentration of TEOA from 0.03 to 0.06 M, the hydrogen production ranged from 35.72 to 54.21 µmol. The calculated hydrogen evolution rate ranged from 1190.69 to 1807.07 µmol·g −1 ·h −1 (Figure 5f). It is noted that varying the concentration of NHS from 0.075 mM to 0.300 mM barely changed the H2 yield, which maintained a constant concentration of TEOA (0.060 M) ( Figure S8). In order to further assess the stability of 1 in the photocatalytic process, a long-term reaction lasting 24 h was carried out, and an amount of 85.40 µmol of hydrogen was cumulatively produced (Figure 6a). The hydrogen production rate gradually decreased over time due to the partial consumption of the sacrificial reagent and photosensitizer in the catalytic system. The catalysts after photocatalysis were isolated by centrifugation and washed with CH 3 CN. The isolated catalyst was further used for three successive photocatalytic cycles in the presence of fresh sacrificial reagents and the photosensitizer; the photocatalytic hydrogen production performance of 1 was maintained with negligible decline of the H 2 yield (Figure 6b), proving the good stability and recycling robustness of catalyst 1. The structural stability of catalyst 1 was further confirmed by the perfectly matched FT-IR spectra of the isolated catalyst with that of the fresh compound 1 (Figure 6c). In order to prove the heterogeneity of the reaction system, an experiment was carried out to analyze the leaching of Co in the post-catalysis solution using the ICP-OES test. The results showed there is little Co (0.007 µmol) dissolution after the photocatalytic reaction (Table S3), which cannot significantly contribute to the hydrogen production activity, as demonstrated by the control experiment (Figure 5b, red line). Overall, all above experimental results and analyses demonstrated that compound 1 can work as a promising heterogeneous catalyst for visible-light-driven hydrogen production with decent activity, stability, and recyclability. results showed there is little Co (0.007 µmol) dissolution after the photocatalytic reaction (Table S3), which cannot significantly contribute to the hydrogen production activity, as demonstrated by the control experiment (Figure 5b, red line). Overall, all above experimental results and analyses demonstrated that compound 1 can work as a promising heterogeneous catalyst for visible-light-driven hydrogen production with decent activity, stability, and recyclability.

Conclusions
Under hydrothermal conditions, a water-insoluble and stable CoAPs, H31(NH4)5-Na16{Co Ⅲ (H2O)6}4{[Co Ⅱ 4(µ3-OH)3(PO4)]4(A-α-GeW9O34)4}2·23H2O (1), was synthesized and characterized by various techniques. Distinguished from the reported hexadecanuclear cobalt-added POM, four additional free {Co(H2O)6} crystallized in 1. The introduction of phosphate linkers contributes to the aggregation of cobalt-oxo clusters. The hydrothermal synthetic method is beneficial for the synthesis of stable and structurally-new compound 1. While coupling with the photosensitizer and sacrificial reagent, 1 worked as a heterogeneous catalyst for visible-light-driven hydrogen evolution. Under optimal catalytic conditions, a hydrogen yield of 54.21 µmol was achieved after 10-h photocatalysis, corresponding to a hydrogen evolution rate of 1807.07 µmol·g −1 ·h −1 . Compared to the classic three-component HER system, NHS was firstly applied to construct a dual sacrificial reagent system, which can significantly enhance the hydrogen evolution activity. The photocatalytic results confirm the enhanced effects of such dual-electron-sacrificial-reagents

Conclusions
Under hydrothermal conditions, a water-insoluble and stable CoAPs, H 31 (NH 4 ) 5 -Na 16 6 } crystallized in 1. The introduction of phosphate linkers contributes to the aggregation of cobalt-oxo clusters. The hydrothermal synthetic method is beneficial for the synthesis of stable and structurally-new compound 1. While coupling with the photosensitizer and sacrificial reagent, 1 worked as a heterogeneous catalyst for visible-light-driven hydrogen evolution. Under optimal catalytic conditions, a hydrogen yield of 54.21 µmol was achieved after 10-h photocatalysis, corresponding to a hydrogen evolution rate of 1807.07 µmol·g −1 ·h −1 . Compared to the classic three-component HER system, NHS was firstly applied to construct a dual sacrificial reagent system, which can significantly enhance the hydrogen evolution activity. The photocatalytic results confirm the enhanced effects of such dual-electron-sacrificial-reagents systems. This study provides new guidance for designing highly stable cobalt-added POMs and expanding their catalytic applications in solar conversion.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nano13132009/s1, Figure S1. TGA curve of 1. Figure S2. Ball and stick model diagram of 1. Color codes: W atom, black; Ge atom, orange; Co atom, turquoise; P atom, yellow; O atom, red. Figure S3. The packing mode of 1 in the bc plane. Color codes: WO 6 , red; CoO 6 , turquoise; PO 4 , yellow; GeO 4 , orange. Figure S4. FT-IR spectrum of 1. Figure S5. Experimental and simulated PXRD patterns of 1. Figure S6. UV-Vis diffuse reflectance spectrum of 1 and the corresponding α/S-Energy curve (inset). Figure S7. Relationship between hydrogen yield and hydrogen evolution rate in different qualities of 1. Figure S8. Hydrogen production comparison of different concentrations of NHS. Table S1. Control experiments for photocatalytic water reduction reaction in NHS-free conditions. Table S2. Overview of different photocatalytic HER performance using cobalt-added tungsten-oxo clusters [39][40][41]56,57]. Table S3. ICP-OES analyses of the cobalt contents in compound 1 and the post-catalytic solution after 10-h photocatalysis.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.