Two Tungstates Containing Platinum Nanoparticles Prepared by Air-Calcining Keggin-Type Polyoxotungstate-Coordinated Diplatinum(II) Complexes: Effect on Sintering-Resistance and Photocatalysis

Two tungstates containing platinum nanoparticles (Pt Npts) were obtained by air-calcining α-Keggin-type diplatinum(II)-coordinated polyoxotungstates, Cs3[α-PW11O39{cis-Pt(NH3)2}2]⋅8H2O (Cs-P-Pt) and Cs4[α-SiW11O39{cis-Pt(NH3)2}2]⋅11H2O (Cs-Si-Pt), at 700–900 °C for 5 h. The polyoxotungstate Cs-P-Pt was transformed to a mixture of Pt Npts and Cs3PW12O40 upon calcination, while the Cs-Si-Pt structures were transformed to Pt Npts and Cs4W11O35. The Pt Npts generated by air-calcining Cs-P-Pt at 700 °C for 5 h were uniform with an average particle size of 3.6 ± 1.1 nm, which was much smaller than that of the Pt Npts obtained by calcining Cs-Si-Pt (19.9 ± 9.9 nm) under identical conditions. This demonstrated the significant inhibitory effect of Cs-P-Pt on aggregation during high-temperature air-calcination at a high platinum content (10.6 wt.%) and in the absence of a support. During calcination at 700–900 °C, Cs-P-Pt exhibited higher activities than Cs-Si-Pt with respect to hydrogen evolution from aqueous triethanolamine solutions under visible light irradiation in the presence of Eosin Y, α-Keggin-type mono-aluminum-substituted polyoxotungstate, and titanium dioxide. When Cs-P-Pt was calcined at 800 °C for 100 h, no decrease in activity was observed in comparison with that upon calcination for 5 h.


Introduction
Platinum nanoparticle catalysts are extensively used in various industrial processes, including hydrogenation, naphtha reforming, oxidation, automotive exhaust catalysis, and fuel generation (fuel cells) [1,2]. However, supported platinum nanoparticle catalysts frequently agglomerate into large particles at high reaction temperatures, losing their catalytically active surface areas [3]. For the suppression of such thermally induced deactivation, complex nanostructures, including core-shell nanostructures [4] and well-designed pore structures [5] have been proposed. Although these methods have been successful in maintaining the platinum nanostructures, aggregation still occurs at higher platinum contents and under thermal treatment at higher temperatures.
For the preparation of supported platinum catalysts, various platinum compounds, such as H 2 PtCl 6 , Pt(NH 3 ) 4 Cl 2 , Pt(NH 3 ) 4 (OH) 2 , Pt(NH 3 ) 4 (NO 3 ) 2 , H 2 Pt(OH) 6 , and Pt(C 5 H 7 O 2 ) 2 are used as precursors [2]. These compounds are heat-treated in the presence of various gases (air, oxygen, hydrogen, H 2 O, N 2 , and Ar) or under vacuum. In each case, the ligand was eliminated during the heat treatment, and reduction of the oxidized platinum sites (Pt 2+ and Pt 4+ ) to Pt 0 was observed. Further high-temperature treatment induced aggregation.
In the process of improving the activities of the platinum sites, it was found that air calcination of a cesium salt of α-Keggin diplatinum(II)-coordinated silicotungstate, Cs 4 [α-SiW 11 [17].

Instrumentation and Analytical Procedures
The infrared spectra were recorded on a PerkinElmer Spec-trum100 FT-IR spectrometer in KBr discs at approximately 25 °C in air. The services of Eurofins EAG Materials Science (USA) were enlisted for performing XPS analyses. A monochromated Al K α radiation (1486.6 eV) was used as the X-ray source. The binding energies are referenced to the C 1s binding energy at 284.8 eV. Powder X-ray diffraction (PXRD) measurements were performed on an X-ray powder diffractometer (SmartLab, Rigaku, Corp., Japan) using Cu K α radiation (Kα = 1.54 Å). Transmission electron microscopy (TEM) images were recorded using a JEOL-JEM 2100F electron microscope, and the elemental compositions of the samples were studied using energy-dispersive X-ray spectroscopy (EDS, JED-2300 T (JEOL, Japan)).

Photocatalytic Reaction Experiments
Typical photocatalytic reactions were performed at 25 °C. EY and K 5 [α-SiW 11 {Al(OH 2 )}O 39 ]·7H 2 O (2.5 μmol) were dissolved in 10 mL of 100 mM aqueous TEOA solution at pH 7.0. Subsequently, the calcined samples (containing 0.2 μmol Pt) and titanium dioxide (anatase:rutile = 80:20; 50 mg) were suspended in this solution. The suspension was placed in a glass reaction vessel, which was connected to a Pyrex conventional closed gas circulation system (245.5 cm 3 ). The photoreaction was initiated by light irradiation using a 300 W Xe lamp equipped with a cut-off filter (λ ≥ 440 nm). The evolution of hydrogen, oxygen, carbon monoxide, and methane was analyzed using a gas chromatography (GC) instrument equipped with a thermal conductivity detector (TCD), 5 Å molecular sieves, and stainlesssteel columns. The samples were assigned after comparison with standard samples analyzed under identical conditions. The turnover number (TON) was calculated as 2[H 2 evolved (mol)]/[Pt atoms (mol)].

Results and Discussion
When Cs-P-Pt was calcined at temperatures from 25 °C to 800 °C at a heating rate of 40 °C min -1 , followed by maintaining at 800 °C for 5 h in air (without flow), its color changed from yellow to black (the calcined sample was denoted as Cs-P-Pt-800-5), and the obtained powder was insoluble in water.
The PXRD pattern of Cs-P-Pt-800-5 (Fig. 2a) showed the same peaks as that of Cs 3 PW 12 O 40 (ICDD: 00-050-1857), and similar peaks were observed for Cs-P-Pt-700-5 and Cs-P-Pt-900-5, as shown in Fig. S3. These results prove that the {PW 11 O 39 } unit in Cs-P-Pt was transformed to {PW 12 O 40 } under the reported thermal treatment conditions. This was validated by the FT-IR spectra. In contrast, the PXRD pattern of Cs-Si-Pt-800-5 (Fig. 2b) exhibited the same peaks as that of Cs 4 W 11 O 35 (ICDD: 00-051-1891). A similar structural transformation of the {SiW 11 O 39 } unit to {W 11 O 35 } was observed for Cs-Si-Pt-700-5 and Cs-Si-Pt-900-5, as shown in Fig. S4. As previously reported, the {SiW 11 O 39 } unit in Cs-Si-Pt was transformed to {SiW 12 O 40 } by air calcination in the lower temperature range of 250-500 °C [17]. However, at higher temperatures in the range of 700 °C to 900 °C, the Keggin structure was transformed to {W 11 O 35 }.
A broad peak corresponding to Pt(111) (2θ(°) = 40, ICDD: 00-04-0802) was observed. This indicates that the di-platinum(II) sites in Cs-P-Pt and Cs-Si-Pt were reduced to crystalline Pt(0) by the applied thermal treatment. As shown in Figs. 2, S3 and S4, the peak intensity increased as the calcination temperature was increased. These results suggest that the crystallite size of platinum increases with the calcination temperature.
The Pt(4f) XPS spectra of Cs-P-Pt-800-5 and Cs-Si-Pt-800-5 are shown in Fig. 3. In both the spectra, the platinum sites were primarily composed of Pt 0 , with low levels of Pt 2+ and Pt 4+ . The occurrence of Pt 4+ needs to be  Table 1. Surprisingly, Cs-P-Pt-700-5 and Cs-P-Pt-800-5 exhibited average particle sizes of 3.6 ± 1.1 mm and 5.3 ± 2 nm, respectively, at a high platinum content (10.6 wt.%), in the absence of a support. The average particle size of Cs-P-Pt-900-5 was 9.3 ± 2.8 nm, which was approximately twice that of Cs-P-Pt-800-5, indicating slight aggregation at 900 °C (Fig.  S7). However, the average particle size reached upon calcination at 800 °C for 100 h was 5.4 ± 1.9 nm, and negligible aggregation of Pt Npts was observed after long-term heat treatment, as shown in Fig. S8.
As a control experiment, Pt Npt-supported TiO 2 (Pt content: 11 wt.%) was prepared as follows: cisdiamminedichloroplatinum(II) (cisplatin) (0.169 g; 0.56 mmol) was dissolved in 50 mL water. TiO 2 (0.83 g) was added to the solution, and the mixture was stirred for 2 h at approximately 25 °C. After evaporation to dryness at 100 °C, the obtained solid was calcined at 700 °C for 5 h in air. The calcined product was denoted as cisplatin/TiO 2 -700-5. The average particle size of cisplatin/TiO 2 -700-5 was 31.4 ± 19.7 nm, which was larger than that of Cs-P-Pt-700-5 and Cs-Si-Pt-700-5, as shown in Fig. S11. When the platinum salt of α-Keggin silicotungstate [Pt(NH 3 ) 4 ] 2 [α-SiW 12 O 40 ]⋅4H 2 O (platinum content: 11.2 wt.%) was calcined at 800 °C for 5 h in air (the calcined product was denoted as Pt-SiW12-800-5), particles with non-uniform sizes were formed with some particles growing above 50 nm  S12a). It can be seen from the FT-IR spectrum of Pt-SiW12-800-5 (Fig. S12b)  3could not be obtained using the same method as that used to obtain SiW 12 O 40 4-. It was confirmed from the FT-IR spectrum (Fig. S13b) that the Keggin structure was decomposed when TMA-P-Pt (platinum content: 11 wt.%) was calcined at 800 °C for 5 h in air (the obtained sample was denoted as TMA-P-Pt-800-5). The particle size was non-uniform, and some particles grew to several tens of nanometers, as shown in Fig. S13a. According to these results, Cs-P-Pt, which retained the α-Keggin polyoxotungstate and Cs 3 PW 12 O 40 after calcination, exhibited an excellent aggregation inhibitory effect at high platinum content and in the absence of a support. Such inhibition of platinum aggregation is not observed in platinum compounds from which the ligands disappear upon calcination [2]. It is critical to coordinate the platinum species to the mono-vacant sites in α-Keggin-type mono-lacunary phosphotungstates and isolate them as cesium salts for the expression of such an effect.
The activities of the Pt Npts in the calcined samples as co-catalysts were studied by using Cs-P-Pt-800-5 and  Cs-P-Pt-700-5 10.6 a 3.6 ± 1.1 Cs-P-Pt-800- 5 5.3 ± 2 Cs-P-Pt-900- 5 9.3 ± 2.8 Cs-P-Pt-800-100 5.4 ± 1.9 Cs-Si-Pt-700- 5 10.1 a 19.9 ± 9.9 Cs-Si-Pt-800-5 42.8 ± 21.  [14,17,24,25]. The precursors, Cs-P-Pt and Cs-Si-Pt were weighted while ensuring that the amount of platinum was 0.2 μmol (the contents of Pt in the catalysts containing TiO 2 and the platinum samples were 0.084 wt.% and 0.080 wt.%, respectively). Subsequently, they were calcined at 800 °C for 5 h in air. TEOA was used as the sacrificial reagent. EY and K-Si-Al were used as the photosensitizer and EY stabilizer, respectively. Although TiO 2 was used to promote charge separation, the tungstates containing Pt Npts were not supported on the surface of TiO 2 . In order to compare the activities with the samples before calcination, TiO 2 was simply dispersed in the solutions. Hydrogen was formed with 100% selectivity, and oxygen, carbon dioxide, carbon monoxide, and methane were not detected under these reaction conditions. It has already been confirmed that no reaction was observed in the absence of platinum catalysts or Eosin Y under the reported conditions. After 6 h of light irradiation, the amounts of hydrogen evolved in presence of Cs-P-Pt-800-5 and Cs-Si-Pt-800-5 were 180 μmol and 156 μmol, respectively, The TONs (= 2[hydrogen evolved (mol)]/[Pt atoms (mol)]) observed for Cs-P-Pt-800-5 and Cs-Si-Pt-800-5 were 1801 and 1556, respectively. These values were higher than those of Cs-P-Pt and Cs-Si-Pt (TONs observed after 6 h were 703 and 409, respectively. These results suggest that the photocatalytic activities of Cs-P-Pt and Cs-Si-Pt were improved upon calcination. The turnover frequencies (TOF = TON/ reaction time (h)) of Cs-P-Pt-800-5 and Cs-Si-Pt-800-5 after 1 h were 685 and 336 h -1 , respectively. These values are higher than those of similar photocatalytic systems containing platinum co-catalysts prepared via the photoreduction of H 2 PtCl 6 , EY, and TiO 2 . For example, the TOFs of Pt/EY/nitrogen-doped TiO 2 [26], Pt/EY/modified TiO 2 with phosphate [27], Pt/EY/Fe 3+ /TiO 2 [28], and Pt/EY/ SiW 11 O 39 8-/TiO 2 [25] are less than 150 h -1 with respect to hydrogen evolution from aqueous TEOA solutions under visible-light irradiation. When platinum black (3.0 μmol Pt) was used as a co-catalyst, the TOF after 1 h was 11.7 h -1 , which was significantly lower than those of Cs-P-Pt-800-5 and Cs-Si-Pt-800-5. It is difficult to discuss the differences in activities based on the sizes of the Pt Npts alone because of the influences of Cs 3 PW 12 O 40 and Cs 4 W 11 O 35 contained in Cs-P-Pt-800-5 and Cs-Si-Pt-800-5, respectively. However, upon calcination at 700-900 °C, Cs-P-Pt showed higher activities than Cs-Si-Pt, as shown in Fig. 7. Upon using Cs-P-Pt-800-100, 192 μmol hydrogen was generated after 6 h (TON: 1919), and there was no reduction in catalytic activity upon extending the calcination time from 5 h to 100 h. In the cases of Cs-P-Pt-800-5 and Cs-Si-Pt-800-5, the amount of hydrogen generated gradually decreased with time, as shown in Figs. 6a and b. For the hydrophilic platinum-polyoxotungstate colloidal particles obtained by air-calcining Cs-Si-Pt at 300 °C for 5 h, the TON after 3 h exceeded 4600 under similar reaction conditions. However, rapid decomposition of the platinum sites was observed and the activity did not improve upon re-addition of EY [17]. To confirm that Cs-P-Pt-800-5 was not decomposed, we performed the following experiment: The aforementioned photocatalytic reaction was performed in the presence of Cs-P-Pt-800-5 (0.2 μmol Pt) under light irradiation for 20 h. The residual solid was collected using a membrane filter (JG 0.2 μm) and washed with water and ethanol. Thereafter, the solid containing Cs-P-Pt-800-5 and TiO 2 was dispersed again in an aqueous TEOA solution containing EY and K-Si-Al. After 6 h of light irradiation, approximately 200 μmol of hydrogen was generated. It was also confirmed that the photocatalytic activity of Cs-Si-Pt-800-5 was restored by the re-addition of EY (Fig.  S14). These results demonstrate that Cs-P-Pt-800-5 and Cs-Si-Pt-800-5 did not decompose under the applied reaction conditions and were recyclable.
Finally, the following experiment was conducted to investigate the sizes of the Pt Npts after light irradiation: Cs-P-Pt-800-5 (37.6 mg) was dispersed in aqueous TEOA containing dissolved EY, and subjected to light irradiation for 6 h. The residual solid was collected through a membrane filter (JG 0.2 μm) and washed with water and ethanol. The TEM images and the corresponding size distributions of the obtained samples are shown in Figs. 8a and b. After light irradiation, homogeneously dispersed Pt Npts were observed in the tungstate matrix. The average particle size was 5.1 ± 2.5 nm, which was nearly the same as that prior to light irradiation. Although it has been reported that the sizes of Pt Npts increase during hydrogen evolution under light irradiation [29], no aggregation of the Pt Npts comprising Cs-P-Pt-800-5 was observed under the applied reaction conditions. This reaction mechanism is similar to that of a reported Pt co-catalyst/EY/K-Al-Si/TiO 2 system [14]. However, since the water-insoluble calcined samples were used without being supported on the surface of titanium oxide, it is possible that the efficiency of electron transfer between titanium oxide and the platinum and/or tungstate sites was reduced. Sintering at high temperatures may also induce three-dimensional microstructural changes and inhibit proton transfer to the platinum sites. In order to overcome these problems, we are investigating methods to support the calcined samples on the surfaces of some semiconducting photocatalytic materials. The results will be reported in due course.

Conclusion
Two t u n g s t a t e s c o n t a i n i n g p l a t i n u m n a n op a r t i c l e s w e r e o b t a i n e d b y a i r-c a l c i n i n g Cs 3 [α-PW 11  conditions. The di-platinum(II) sites in Cs-P-Pt and Cs-Si-Pt transformed to Pt Npts. The average particle size of Pt Npts generated via air-calcination of Cs-P-Pt was much smaller than that observed when Cs-Si-Pt was calcined. Thus, Cs-P-Pt exhibited an excellent aggregation inhibitory effect at high platinum content and in the absence of a support.
The calcined samples were applied as co-catalysts for hydrogen evolution from aqueous TEOA solutions under visible light irradiation (λ ≥ 440 nm) in the presence of Eosin Y, K 5 [α-SiW 11 {Al(OH 2 )}O 39 ]·7H 2 O, and titanium dioxide. Upon calcination at 700-900 °C, Cs-P-Pt exhibited higher activities than Cs-Si-Pt. The activity of Cs-P-Pt did not diminish after calcination at 800 °C for 100 h, in comparison with its activity after 5 h of calcination. It was also confirmed that the Cs-P-Pt sample calcined at 800 °C for 5 h exhibited no change in the sizes of the Pt Npts when subjected to light irradiation. Platinum aggregation was thus suppressed under the applied photocatalytic reaction conditions.