Porous silica materials derived from cage-siloxane: structural confinement and stabilisation of dispersed Au(0) nanoparticles

Abstract

Highly dispersed gold(Au) species exhibit unique catalytic properties that differ considerably from those of bulk Au or larger particles. However, achieving such dispersion while preserving their intrinsic characteristics remains a challenge. This difficulty often arises because conventional methods can alter the characteristics of the Au itself. This study addresses this issue using silica, a support material known for its weak interaction with active metal species. Specifically, we developed a novel nanoporous silica support functionalised with hydrosilyl groups using a cage-siloxane as a building block. These groups enable on-site reduction of Au³⁺ ions, allowing highly dispersed Au loading without introducing strongly interacting ligands. Importantly, the dispersion was retained even after calcination, indicating a confinement effect within the porous structure. This approach offers a promising strategy for stabilising Au species on inert supports, potentially advancing the design of robust and efficient Au-silica catalysts for heterogeneous catalysis.

---

Introduction

Gold (Au) has attracted considerable attention as a catalytic material since the pioneering discoveries reported by Haruta¹ and Hutchings,² whose work revealed that Au nanoparticles (NPs) exhibit unexpectedly high catalytic activity. In such systems, the particle size and dispersion of Au NPs are key determinants of catalytic performance.^1,3,4 However, controlling both parameters on conventional supports remains as a formidable challenge. Reducible oxides (e.g., TiO₂, ZrO₂, CeO₂) can often facilitate the deposition of small Au NPs mainly through deposition–precipitation methods,⁴ whereas “inert” silica supports,⁵ which are valued for their thermal and chemical robustness,^6,7 interact only weakly with Au,^7,8 typically leading to poor dispersion and facile aggregation of metallic NPs.^4–7

Various approaches reportedly improve the dispersion of Au NPs on silica: the introduction of heteroatom dopants or partial modification with other oxides,⁹ use of gold ammine complexes,¹⁰ and surface functionalisation with amine or thiol groups.^11–13 However, these interactions often alter the electronic state of the Au particles to a considerable degree.^14,15 Consequently, examples of highly zero-valent, finely dispersed Au being applied as a catalyst are rare. Nonetheless, several studies have demonstrated that even zero-valent Au can exhibit catalytic activity for processes such as glucose oxidation.^16,17 Moreover, in partially oxidised catalysts, metallic Au(0) is known to play a pivotally important role as a catalyst for the activation of reactive oxygen species.^4,18

Therefore, the development of silica-based catalysts that support highly dispersed, zero-valent Au particles is important as a foundational platform for designing catalysts that utilise the Au(0) state effectively, or which allow tunability of the Au charge state.

To address the challenge of achieving high dispersion and stabilisation of Au(0) nanoparticles, a promising approach involves molecular-level structural control using siloxane-based building blocks. This bottom-up synthesis enables the uniform introduction of functional groups, facilitating the design of advanced silica materials with tailored cage-type oligosiloxanes DnR (n = 4, 6 and 8) offering thermal and chemical stability, with tunable vertices for diverse functionalization.^19–23 Additionally, their use as precursors for synthesising porous materials has been reported.^23–25 They are promising building blocks for silica-based porous materials. Particularly, D4R, a secondary structural unit of LTA zeolites,²⁶ holds potential for the synthesis of robust porous supports.

This report describes the synthesis of a novel silica-based porous support using cage siloxanes as building blocks, designed to achieve high dispersion and confinement of Au nanoparticles, thereby suppressing agglomeration.

We crosslinked D4R-type cage siloxane, Q8H8, with triethoxysilane (TES) (Scheme 1). The hydrosilyl groups present in triethoxysilane can reduce noble metal salts, as shown in eqn (1). The groups reportedly facilitate on-site reduction and deposition of metallic species on monolithic structures.²⁷

Scheme 1 The synthesis of TES-Q8H8 and deposition of Au NPs.

Additionally, silanol groups generated from the oxidation of hydrosilyl groups can undergo condensation to form siloxane bonds,²⁸ potentially enhancing aggregation resistance through structural stabilisation and producing the expected higher selectivity for several reactions.^29,30

Herein, we describe the synthesis of silica-based porous materials featuring hydrosilyl-functionalized D4R frameworks and the use of these materials to achieve on-site reduction, dispersion, and confinement of Au(0) nanoparticles, along with their aggregation-suppression effects.

Experimental

Materials

The following chemicals were used without further purification: tetraethoxysilane (TEOS, ≥95.0%; Fujifilm Wako Pure Chemical Corp.), triethoxysilane (TES; Enamine Ltd) tetramethylammonium hydroxide pentahydrate (TMAOH-5H₂O, ≥95.0%; Aldrich Chemical Co. Inc.), ethanol (EtOH, ≥99.5%; Fujifilm Wako Pure Chemical Corp.), tetrahydrofuran (THF, ≥99.5%, Stabilizer Free; Fujifilm Wako Pure Chemical Corp.), magnesium sulphate (MgSO₄, ≥98.0%; Fujifilm Wako Pure Chemical Corp.), nitric acid (1.42) (HNO₃; Fujifilm Wako Pure Chemical Corp.), N,N-dimethylacetamide (DMAc, ≥99.0%; Hayashi Pure Chemical Ind. Ltd), hydrochloric acid (HCl; Fujifilm Wako Pure Chemical Corp.), hydrogen tetrachloroaurate(III) tetrahydrate (HAuCl₄-4H₂O, >99.9%; Kanto Chemical Co. Inc.).

Preparation of TES-Q8H8

The building block, Q8H8, Si₈O₁₂(OH)₈, dispersed in a mixture of THF and DMAc was obtained after acid treatment and subsequent dehydration by MgSO₄ of TMA₈(Si₈O₂₀) after using a method reported earlier.³¹ Then, Q8H8 was reacted with HSi(OEt)₃ to obtain a porous material. Specifically, 135 mL of Q8H8 solution in THF and 5.212 ml of HSi(OEt)₃ were mixed at a molar ratio of 3 : 8, ensuring stoichiometric amounts of silanol groups and ethoxy groups. The mixture was refluxed at 353 K for 1 day under a N₂ atmosphere. After a white solid was precipitated in the solution, it was filtered with THF and dried under reduced pressure to obtain a slightly lumpy white powder. The powder was stirred in EtOH for 1 h to remove residual DMAc. Then it was filtered and dried under reduced pressure, yielding a smooth white powder designated as TES-Q8H8.

Deposition of Au particles on TES-Q8H8 and its calcination

Au particles were deposited onto/into TES-Q8H8 through the redox reaction of hydrosilyl groups and Au³⁺ ions. A small amount of water was used to disperse TES-Q8H8, with subsequent addition of a tetrachloroauric(III) acid solution equivalent to 5 wt%. The mixture was stirred for 1 day under light shielding. After filtration with water and drying under reduced pressure, a dark purple powder was obtained, denoted as Au/TES-Q8H8. The sample was then calcined in air at 773 K for 1 h to yield a dark purple powder, designated as Au/TES-Q8H8-calc.

Acid treatment of Au-supported TES-Q8H8

Au/TES-Q8H8 and Au/TES-Q8H8-calc. were stirred in a mixture of 15 mL nitric acid and 5 mL hydrochloric acid for 5 min. The solution was then diluted with water immediately while cooling in an ice bath. Subsequently, the process was repeated several times, involving centrifugation and decantation. The resulting precipitate was dried under reduced pressure, yielding white to light purple powders, designated respectively as Au/TES-Q8H8-acid and Au/TES-Q8H8-calc.-acid.

Characterization

Liquid-state ¹H, ¹³C and ²⁹Si nuclear magnetic resonance (NMR) spectra were recorded using a spectrometer (JNM ECZ500R; JEOL) with resonance frequencies of 500.16, 125.77, and 99.37 MHz, respectively, at room temperature using 5 mm glass tubes. Acetone-d₆ was used to obtain lock signals. Tetramethylsilane (δ = 0 ppm) was used as the internal reference. Tetraethylsilane (δ = 5.75 ppm) was used to quantify the Si concentration in solutions of the samples for liquid-state ²⁹Si NMR spectroscopy. A small amount of Cr(acac)₃ (acac = acetylacetonate) was used as a relaxation agent for the ²⁹Si nuclei. Liquid-state ²⁹Si NMR spectra were measured using a 45° pulse and recycle delay of 10 s. Solid-state ²⁹Si magic angle spinning (MAS) NMR spectra were recorded on a spectrometer (JNM-ECA400; JEOL) at a resonance frequency of 79.43 MHz with a 90° pulse and pulse delay of 250 s at ambient temperature. Solid-state ¹³C cross-polarization (CP)/MAS NMR spectra were recorded on the same spectrometer at a resonance frequency of 100.53 MHz with a contact time of 5 ms and pulse delay of 5 s. The samples were put in a 5 mm zirconia rotor and were spun at 8 kHz. Hexamethylbenzene was used as an external reference (δ = 17.4 ppm) for solid-state ¹³C NMR spectroscopy. Poly(dimethylsilane) was used as an external reference (δ = −33.8 ppm) for solid-state ²⁹Si NMR spectroscopy. Solid-state two-dimensional ¹H−²⁹Si heteronuclear correlation solid-state (HETCOR) NMR spectra were recorded using the same spectrometer at resonance frequencies of 400.0 MHz for ¹H and 79.4 MHz for ²⁹Si with contact time of 5 ms. Fourier transform infrared (FT-IR) spectra were obtained using a spectrometer (FT/IR-6100; Jasco Corp.) with the KBr method. Powder X-ray diffraction (XRD) patterns were obtained using a diffractometer (Smart Lab; Rigaku Corp.) employing CuKα radiation (40 kV, 40 mA). Also, N₂ adsorption–desorption measurements were performed with a at 77 K (Autosorb-iQ; Quantachrome Instruments). As pretreatments, the samples were heated at 393 K for 4.5 h under reduced pressure. Brunauer–Emmett–Teller (BET)³² areas were calculated using the Rouquerol method.³³ Transmission electron microscopy (TEM) images were taken using a microscope (JEM-2100F; JEOL) operating at 200 kV. Energy dispersive X-ray spectra (EDXS) mapping was done (JED-2300T; JEOL). X-ray photoelectron spectroscopy (XPS) measurements were taken (VersaProbe 4; ULVAC-PHI) under vacuum conditions. The depth profiles of samples were obtained using an argon gas cluster ion beam with energy of 4 kV. Au L3-edge X-ray absorption fine structure (XAFS) analyses were conducted at the beamline BL14B2 of SPring-8 at the Japan Synchrotron Radiation Research Institute. The X-rays were monochromatized using a Si(311) double-crystal monochromator. The photon energy was calibrated at the inflection point of the Au L3-edge X-ray absorption near-edge structure (XANES) spectrum (pre-edge peak) of the Au metal foil to 11 921.7 eV.³⁴ The powdered samples were diluted with boron nitride (BN), pressed into pellets in a metal holder, and used for X-ray absorption spectroscopy (XAS) measurements. Inductively coupled plasma (ICP, 5100 ICP-OES; Agilent Technologies Inc.) analysis was conducted. For ICP analysis, the samples were dissolved in a mixture of hydrofluoric acid and aqua regia.

Results and discussion

Synthesis of TES-Q8H8

FT-IR spectra of TES-Q8H8 and TMA₈(Si₈O₂₀) are presented in Fig. 1(a). TES-Q8H8 showed distinctive bands at 1080 cm⁻¹ and a shoulder peak at 1150 cm⁻¹ (Fig. 1(b)), assignable to (Si–O)₄ ring-symmetry/asymmetry mode respectively.³⁵ The bands indicate the D4R structures remain even after cross-linking. Bands observed at 881 and 2264 cm⁻¹ are presumably derived respectively from Si–H bending and stretching vibrations. Then they represent the existence of hydrosilyl groups.^35,36 These results support that TES-Q8H8 has properties of both TES and Q8H8. The broad band of TES-Q8H8 at 3000–3700 cm⁻¹ can be assigned to the SiO-H vibration of silanol groups.³⁷ Although the HO-H weak band observed at 1634 cm⁻¹ indicates the existence of the adsorbed water molecules,^37–39 the presence of a shoulder peak around 3700 cm⁻¹, attributed to isolated silanol groups, provides evidence for the existence of silanol.^36,37

Fig. 1 (a) FT-IR spectra of TMA₈(Si₈O₂₀) (blue) and TES-Q8H8 (red). (b) Enlarged image of FT-IR spectra near 1000–1300 cm⁻¹ for TES-Q8H8. Orange and blue dashed lines show the fitted peaks for the asymmetric and symmetric modes of (Si–O)₄ rings, respectively. The black dashed line in (b) represents the combined fitting curve.

The existence of hydrosilyl groups in TES-Q8H8 was confirmed further by TGA measurements (Fig. S1). The slight increase in weight observed over 673 K for TES-Q8H8 provides a reasonable explanation for the oxidation of SiH groups.^36,40

The solid-state ²⁹Si MAS NMR spectra of TES-Q8H8 in Fig. 2(a) showed signals attributed to T²(SiHOEt(OSi)₂ and/or SiHOH(OSi)₂), T³(SiH(OSi)₃), Q³(SiOH(OSi)₃), and Q⁴(Si(OSi)₄) sites, respectively at −74.09, −83.61, −99.40, and −108.22 ppm. The integrated intensity ratio of (Q³ + Q⁴)/(T² + T³) was 2.78, which was similar to the theoretical ratio (Q/T = 3), implying that the stoichiometric reaction proceeded, whereas T² and Q³ respectively illustrate the existence of uncross-linked parts of TES and Q8H8. Taking this supposition into consideration, we can calculate the ratio of cross-linked parts of triethoxysilane with Q⁴/(3T³ + 2T²) (theoretical ratio is 1) and the ratio of cross-linked parts of Q8H8 with Q⁴/(Q⁴ + Q³). Experimentally obtained results show the ratio of Q⁴/(3T³ + 2T²) as 0.603 and that of Q⁴/(Q⁴ + Q³) as 0.632: the difference between them was estimated as the result of condensation between Q8H8. In addition, when comparing the area ratios of the T² and Q³ signals corresponding to the uncross-linked parts, the ratio of Q³/T² was 12.0. These results imply that cross-linking occurred between triethoxysilanes, whereas almost no cross-linking between Q8H8 was observed.

Fig. 2 (a) Solid-state ²⁹Si MAS NMR spectra of TES-Q8H8 (red) and Au/TES-Q8H8 (green). Pale blue, blue, yellow and orange dotted lines respectively represent fitting data of TES-Q8H8 attributable to T², T³, Q³, and Q⁴ environments. (b) Si environment in the assumed structure of TES-Q8H8.

The N₂ adsorption–desorption isotherm of TES-Q8H8 (Fig. 3(a)) appeared to be combined Type I and Type II of the IUPAC classification with the Brunauer–Emmett–Teller (BET)³³ area of 510 m² g⁻¹ and total pore volume of 0.52 cm³ g⁻¹. However, when Q8H8 was refluxed without TES and was subjected to the same treatment, a white powder denoted as Q8H8-only was produced, with much smaller BET area (76 m² g⁻¹) and total pore volume (0.038 cm³ g⁻¹) than those of TES-Q8H8. Furthermore, when TES without Q8H8 was refluxed at 353 K, a much smaller amount of solid was obtained. These results indicate that cross-linking of Q8H8 using the appropriate linker is necessary to achieve high porosity. Furthermore, pore-size distributions calculated using the Barrett–Joyner–Halenda (BJH) method show that TES-Q8H8 mainly has mesopores with diameters less than 10 nm (Fig. 3(b)),³² and the Saito–Foley (SF)⁴¹ method shows micro-pore size distribution. The surface area and pore volume of TES-Q8H8-based materials are summarized in Table S1.

Fig. 3 (a) N₂ adsorption–desorption isotherms of TES-Q8H8 (red), Q8H8-only (blue), and TES-Q8H8-calc. (green). (b) Pore size distribution, as determined from BJH analysis of Adsorption data for TES-Q8H8. (c) Micro pore size distribution determined from SF analysis of TES-Q8H8.

In situ reduction of Au³⁺ ion in TES-Q8H8

Fig. 2(a) shows ²⁹Si MAS NMR spectra of Au/TES-Q8H8. The Si environment of Au/TES-Q8H8 differs from TES-Q8H8 in that T signals have diminished. These results indicate that the oxidation of hydrosilyl groups across Au/TES-Q8H8 proceeded. The Q² signal in Au/TES-Q8H8 can be attributed to the partial cleavage of D4R structures and/or triethoxysilane-derived cross-linkers.

The N₂ adsorption–desorption isotherm shows that Au/TES-Q8H8 has a smaller BET area (409 m² g⁻¹) and pore volume (0.33 cm³ g⁻¹) than those of TES-Q8H8 (Fig. S2). This difference is probably attributable to the incorporation of Au particles and/or possible partial rearrangement of the porous structure.¹³

The amount of Au for Au/TES-Q8H8 obtained by ICP-OES was 2.11 wt%. Showing the existence of Au after immersing TES-Q8H8 with the solution of HAuCl₄.

X-ray photoelectron spectroscopy (XPS) measurement of Au/TES-Q8H8 detected Au4f_7/2 and Au4f_5/2 photoelectron signals respectively at around 84 and 88 eV (Fig. 4(a)),⁴² also showing the existence of Au in Au/TES-Q8H8. We note that the surface charging effect could not be excluded for this material,^10,43 which might have led to deviations in the measured binding energies. Therefore, the electric state of Au was not quantified with XPS. The percentage of Au considering Au, O, and Si in Au/TES-Q8H8 was 0.17 mol% for the first ten cycles on average, which remained almost constant (0.21 mol%) for the last ten cycles on average after 250 argon sputter cycles at 30 s per cycle (Fig. 4(b)).

Fig. 4 (a) XPS spectra of Au/TES-Q8H8 before argon sputtering. (b) Atomic ratio of Au in Au/TES-Q8H8 (red) and Au/TES-Q8H8-calc.-acid (blue) with argon sputtering.

X-ray absorption fine structure (XAFS) measurements were taken to investigate the electronic states of Au particles in Au/TES-Q8H8. Au L3-edge X-ray absorption near edge structure (XANES) spectra (Fig. S3(a)) of Au/TES-Q8H8 show much less intense pre-edge peaks for Au/TES-Q8H8 and correspond well with those of Au foil. These spectra indicated that Au in Au/TES-Q8H8 was mostly Au(0). In this regard, Au³⁺ was surely reduced in the inner part of Au/TES-Q8H8.

Scanning transmission electron microscope (STEM) images of Au/TES-Q8H8 in Fig. 5(a) and energy dispersive X-ray spectrometry (EDS) mapping in Fig. 5(b) reveal the presence of supported Au nanoparticles across the entire support. However, some particles larger than 10 nm in diameter were observed. Fig. 5(c) shows that such larger Au particles are distributed mainly on the Au/TES-Q8H8 surface: a phenomenon often seen in Au deposition on porous materials.¹¹ This finding is consistent with the pore size distribution of TES-Q8H8.

Fig. 5 (a) STEM image of Au/TES-Q8H8, (b) STEM-EDS mapping of Au/TES-Q8H8, and (c) TEM images of Au/TES-Q8H8.

The XRD patterns of TES-Q8H8 and Au/TES-Q8H8 are presented in Fig. S4. The appearance of diffraction peaks derived from Au(JCPDS#04-0784) was observed after immersion of TES-Q8H8 with HAuCl₄ solution, and the average Au particle radius calculated from Scherrer's equation was 25.4 nm. This is even larger than the radii of externally supported Au particles observed in TEM images. This discrepancy is presumably due to the pre-treatment of TEM measurement. The TEM samples were prepared using the supernatant obtained after dispersing the powders in ethanol, suggesting that the larger externally supported particles may have sedimented during stirring. The corresponding XRD results of the sample after removing these externally supported particles are discussed later.

Therefore, removing surface-located Au particles is crucially important for not only achieving a more uniform distribution of Au nanoparticles throughout the material but also the analysis of the properties for confined Au particles.

Information on the Au loading and average particle size of Au/TES-Q8H8-based materials is summarized in Table S2.

Calcination of Au-supported TES-Q8H8

To investigate resistance to agglomeration, the calcination process was applied to Au/TES-Q8H8 and was designated as Au/TES-Q8H8-calc. Usually, metal species agglomerate as heating proceeds according to the Ostwald ripening model mechanism.^44,45 However, Au particle distributions from TEM images show that the average radius of Au particles was almost identical after calcination, although the number of Au particles smaller than 5 nm decreased (Fig. 6(a) and (b)). These results suggest that some agglomerated Au particles might have been redispersed during the calcination process, thereby contributing to the maintenance of average particle size.⁴⁵ Qian et al. reported that redispersion of Au particles occurs when the Au/SiO₂ is calcined.⁴⁶ In their research, the particle size distribution, which was uniform before calcination, became broader after calcination, leading to highly dispersed smaller Au particles. Simultaneously, large Au particles were also produced. By contrast, the absence of larger Au particles after calcination in this study is likely attributable to the suppression of agglomeration by the confinement effect of Au particles in the porous silica network.

Fig. 6 (a) TEM images of Au/TES-Q8H8-calc. and (b) Au particle distribution of Au/TES-Q8H8 (blue), Au/TES-Q8H8-calc. (green) and Au/TESQ8H8-calc.-acid (red).

The 2D ¹H-²⁹Si CP/MAS NMR HETCOR spectra of Au/TES-Q8H8 and Au/TES-Q8H8-calc. show calcination effects as presented in Fig. S5(a) and (b). One typical change was diminishment of the Q² signal. Because this signal is apparent in short contact time spectra of Au/TES-Q8H8 (Fig. S5(c)), silanol groups in proximity likely underwent condensation, forming Si–O–Si linkage.⁴⁷ The indistinguishable broadened Q³ and Q⁴ signals represent various Si–O–Si and Si-OH environments.

To confirm that reduction of Au³⁺ occurs by the hydrosilyl groups in TES-Q8H8, TES-Q8H8 heated in air for one hour before impregnation is hereinafter designated as TES-Q8H8-calc. This TES-Q8H8-calc. was found to retain its microporosity (Fig. 3(a)). The impregnation of TES-Q8H8-calc. with HAuCl₄ solution for 1 day yields a white powder designated as TES-Q8H8-calc.-Au. TEM images confirm the near lack of Au particles in TES-Q8H8-calc.-Au. The loading amount of Au obtained by ICP-OES was 0.032 wt%. These results indicate that deposition of Au inside TES-Q8H8 was driven by the reduction of hydrosilyl groups.

Acid treatment of Au/TES-Q8H8 and Au/TES-Q8H8-calc.

Au/TES-Q8H8 and Au/TES-Q8H8-calc. were treated with acid for the removal of externally supported Au particles. The Au loading on Au/TES-Q8H8-acid was 0.0029 wt%, indicating that Au particles across Au/TES-Q8H8 were leached out after the treatment. By contrast, not all Au was leached out for the calcined samples Au/TES-Q8H8-calc.-acid, with loading of 0.22 wt%. The Au/TES-Q8H8-calc.-acid depth profile presented in Fig. 4(b) shows almost no Au deposition on the Au/TES-Q8H8-calc.-acid surface. However, as the sputter cycles progressed, more Au was detected in Au/TES-Q8H8-calc.-acid. Furthermore, the XRD analysis of Au/TES-Q8H8-calc.-acid in Fig. S2 revealed that the diffraction peak corresponding to the Au(111) plane became weaker and broader than Au/TES-Q8H8. Based on the Scherrer equation, the average particle size of Au was estimated to be 7.1 nm. Taking into account that this equation reflects a volume-weighted average of particle sizes,⁴⁸ the result is consistent with the particle size distribution in Fig. 6(b).

Considering the dehydration–condensation occurring between silanol groups during the calcination process, one can reasonably infer that the Au particles were confined within the support.

Conclusions

We have synthesised a new porous silica (TES-Q8H8) using cage siloxane as a building block. Hydrosilyl groups introduced throughout the TES-Q8H8 structure enabled the support of Au particles by immersion with chloroauric acid (Au/TES-Q8H8). The Au particles in a metallic state were uniformly supported on the whole material. Some particles were deposited as large particles on the surface. However, they were removable by acid treatment. Even after calcination of Au/TES-Q8H8 at 773 K for 1 h, the average Au particle size of Au/TES-Q8H8 changed only slightly. An effect of confinement was observed. In addition, Au/TES-Q8H8 after calcination showed improved resistance to acid treatment, indicating that the Au particle confinement had improved. This new material is expected to be applied as a newly porous silica-based catalyst.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included in the manuscript itself and in the SI. Supplementary information is available. See DOI: https://doi.org/10.1039/d5dt01824b.

Acknowledgements

The author expresses sincere gratitude to Dr Taiki Hayashi for his valuable advice and support in initiating this research. Appreciation is also extended to Ms Madoka Chino for her assistance with various aspects of the experiments, Mr Koki Saegusa for his invaluable help in establishing the experimental environment, and Dr Ayaka Shigemoto for her constructive suggestions related to the experimental work. The author also thanks Dr Takuma Higo and Mr Hiromu Akiyama for their cooperation with XPS and XAFS measurements, and Mr Kosuke Watanabe for his assistance in setting up the TGA system. This work benefited from research equipment shared in the MEXT Project for Promoting Public Utilization of Advanced Research Infrastructure (Program for Supporting Construction of Core Facilities, Grant Number JPMXS0440500023). A part of this work was supported by JSPS KAKENHI Grant Numbers 23H05404, 24H00487, and 23K20034.

References

1. M. Haruta, T. Kobayashi, H. Sano and N. Yamada, Chem. Lett., 1987, 16, 405–408.
2. G. J. Hutchings, J. Catal., 1985, 96, 292–295.
3. B. Hvolbæk, T. V. W. Janssens, B. S. Clausen, H. Falsig, C. H. Christensen and J. K. Nørskov, Nano Today, 2006, 2, 14–18.
4. T. Ishida, T. Murayama, A. Taketoshi and M. Haruta, Chem. Rev., 2020, 120, 464–525.
5. G. Kennedy, L. R. Baker and G. A. Somorjai, Angew. Chem., Int. Ed., 2014, 53, 3405–3408.
6. Y. Wei, W. Yang and Z. Yang, Int. J. Hydrogen Energy, 2022, 47, 9537–9565.
7. C. S. Ewing, G. Veser, J. J. McCarthy, D. S. Lambrecht and J. K. Johnson, Surf. Sci., 2016, 652, 278–285.
8. C. Pan, M. Tsai, W. Su, J. Rick, N. G. Akalework, A. K. Agegnehu, S. Cheng and B. Hwang, J. Taiwan Inst. Chem. Eng., 2017, 74, 154–186.
9. J. P. Gabaldon, M. Bore and A. K. Datye, Top. Catal., 2007, 44, 253–262.
10. F. Somodi, I. Borbáth, M. Hegedűs, A. Tompos, I. E. Sajó, Á. Szegedi, S. Rojas, J. L. G. Fierro and J. L. Margitfalvi, Appl. Catal., A, 2008, 347, 216–222.
11. P. R. Selvakannan, K. Mantri, J. Tardio and S. K. Bhargava, J. Colloid Interface Sci., 2013, 394, 475–484.
12. Y. Ma, G. Nagy, M. Siebenbürger, R. Kaur, K. M. Dooley and B. Bharti, J. Phys. Chem. C, 2022, 126, 2531–2541.
13. E. Eom, M. Song, J. C. Kim, D. Kwon, D. N. Rainer, K. Gołąbek, S. C. Nam, R. Ryoo, M. Mazur and C. Jo, JACS Au, 2022, 2, 2327.
14. R. Schürmann, E. Titov, K. Ebel, S. Kogikoski, A. Mostafa, P. Saalfrank and A. R. Milosavlijević, Nanoscale Adv., 2022, 4, 1599–1607.
15. Y. Zhang, J. Zhang, B. Zhang, R. Si, B. Han, F. Hong, Y. Niu, L. Sun, L. Li, B. Qiao, K. Sun, J. Huang and M. Haruta, Nat. Commun., 2020, 11, 558.
16. Y. Mikami, A. Dhakshinamoorthy, M. Alvaro and H. García, Catal. Sci. Technol., 2013, 3, 58–69.
17. M. Comotti, C. D. Prina, E. Falletta and M. Rossi, Adv. Synth. Catal., 2006, 348, 313–316.
18. G. Corro, S. Cebada, U. Pal and J. L. G. Fierro, J. Catal., 2017, 347, 148–153.
19. A. Shimojima and K. Kuroda, Molecules, 2020, 25, 524.
20. D. B. Cordes, P. D. Lickiss and F. Rataboul, Chem. Rev., 2010, 110, 2081–2173.
21. T. Hikino, K. Fujino, N. Sato, H. Wada, K. Kuroda and A. Shimojima, Chem. Lett., 2021, 50, 1643–1647.
22. S. Saito, H. Wada, A. Shimojima and K. Kuroda, Inorg. Chem., 2018, 57, 14686–14691.
23. T. Nishitoba, T. Matsumoto, F. Yagihashi, J. Satou, T. Kikuchi, K. Sato and M. Igarashi, Chem. Mater., 2024, 36, 10198–10204.
24. N. Sato, Y. Kuroda, H. Wada, A. Shimojima and K. Kuroda, Chem. – Eur. J., 2018, 24, 17033–17038.
25. T. Hayashi, T. Matsuno, K. Kuroda and A. Shimojima, Chem. Lett., 2024, 53, upae025.
26. C. S. Cundy and P. A. Cox, Chem. Rev., 2003, 103, 663–702.
27. N. Moitra, K. Kanamori, T. Shimada, K. Takeda, Y. H. Ikuhara, X. Gao and K. Nakanishi, Adv. Funct. Mater., 2013, 23, 2714–2722.
28. N. Sato, Y. Kuroda, T. Abe, H. Wada, A. Shimojima and K. Kuroda, Chem. Commun., 2015, 51, 11034–11037.
29. Y. Zhang, R. Takahashi, K. Kimura, H. Fujitsuka and T. Tago, Catal. Surv. Asia, 2023, 27, 56–66.
30. C. Wang, L. Wang, J. Zhang, H. Wang, J. P. Lewis and F. S. Xiao, J. Am. Chem. Soc., 2016, 138, 7880–7883.
31. M. Igarashi, T. Nozawa, T. Matsumoto, F. Yagihashi, T. Kikuchi and K. Sato, Nat. Commun., 2021, 12, 7025.
32. S. Brunauer, P. H. Emmett and E. Teller, J. Am. Chem. Soc., 1938, 60, 309–319.
33. J. Rouquerol, P. Llewellyn and F. Rouquerol, Stud. Surf. Sci. Catal., 2007, 160, 49–56.
34. H. Asakura, S. Yamazoe, T. Misumi, A. Fujita, T. Tsukuda and T. Tanaka, Radiat. Phys. Chem., 2020, 175, 108270.
35. E. S. Park, H. W. Ro, C. V. Nguyen, R. L. Jaffe and D. Y. Yoon, Chem. Mater., 2008, 20, 1548–1554.
36. M. Marrone, T. Montanari, G. Busca, L. Conzatti, G. Costa, M. Castellano and A. Turturro, J. Phys. Chem. B, 2004, 108, 3563–3572.
37. E. A. Paukshtis, M. A. Yaranova, I. S. Batueva and B. S. Bal'zhinimaev, Microporous Mesoporous Mater., 2019, 288, 109582.
38. S. P. Zhdanov, L. S. Koshelev and T. I. Titova, Langmuir, 1987, 3, 960–967.
39. A. Burneau and O. Barrès, Langmuir, 1990, 6, 1364–1372.
40. T. Shioda, T. Gunji, N. Abe and Y. Abe, Appl. Organomet. Chem., 2011, 25, 661–664.
41. A. Saito and H. C. Foley, AIChE J., 1991, 37, 429–437.
42. N. Kruse and S. Chenakin, Appl. Catal., A, 2011, 391, 367–376.
43. A. Pélisson-Schecker, H. J. Hug and J. Patscheider, Surf. Interface Anal., 2011, 44, 29–36.
44. P. W. Voorhees, J. Stat. Phys., 1985, 38, 231–252.
45. L. Liu, J. Lu, Y. Yang, W. Ruettinger, X. Gao, M. Wang, H. Lou, Z. Wang, Y. Liu, X. Tao, L. Li, Y. Wang, H. Li, H. Zhou, C. Wang, Q. Luo, H. Wu, K. Zhang, J. Ma, X. Cao, L. Wang and F. Xiao, Science, 2023, 383, 94–101.
46. K. Qian, Z. Jiang and W. Huang, J. Mol. Catal. A: Chem., 2007, 264, 26–32.
47. S. Smeets, Z. J. Berkson, D. Xie, S. I. Zones, W. Wan, X. Zou, M.-F. Hsieh, B. F. Chmelka, L. B. McCusker and C. Baerlocher, J. Am. Chem. Soc., 2017, 139, 16803–16812.
48. C. Weidenthaler, Nanoscale, 2011, 3, 792–810.

---