# Fileset

[s41467-024-45066-9.pdf](https://mdr.nims.go.jp/filesets/ea4d66a7-975e-42e9-92c9-5bd2f1e13cab/download)

## Creator

[Kang Xia](https://orcid.org/0000-0002-9376-8953), [Takafumi Yatabe](https://orcid.org/0000-0001-5504-4762), Kentaro Yonesato, Soichi Kikkawa, [Seiji Yamazoe](https://orcid.org/0000-0002-8382-8078), [Ayako Nakata](https://orcid.org/0000-0002-3311-6283), [Ryo Ishikawa](https://orcid.org/0000-0001-5801-0971), [Naoya Shibata](https://orcid.org/0000-0003-3548-5952), [Yuichi Ikuhara](https://orcid.org/0000-0003-3886-005X), [Kazuya Yamaguchi](https://orcid.org/0000-0002-7661-4936), [Kosuke Suzuki](https://orcid.org/0000-0002-8123-1462)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Ultra-stable and highly reactive colloidal gold nanoparticle catalysts protected using multi-dentate metal oxide nanoclusters](https://mdr.nims.go.jp/datasets/cc4548a1-7abb-41fa-9f59-ecdbfe56fb26)

## Fulltext

Ultra-stable and highly reactive colloidal gold nanoparticle catalysts protected using multi-dentate metal oxide nanoclustersArticle https://doi.org/10.1038/s41467-024-45066-9Ultra-stable and highly reactive colloidalgold nanoparticle catalysts protected usingmulti-dentate metal oxide nanoclustersKang Xia 1, Takafumi Yatabe 1, Kentaro Yonesato1, Soichi Kikkawa2,Seiji Yamazoe 2, Ayako Nakata 3, Ryo Ishikawa 4, Naoya Shibata 4,Yuichi Ikuhara 4, Kazuya Yamaguchi 1 & Kosuke Suzuki 1Owing to their remarkable properties, gold nanoparticles are applied indiverse fields, including catalysis, electronics, energy conversion and sensors.However, for catalytic applications of colloidal gold nanoparticles, the trade-off between their reactivity and stability is a significant concern. Here wereport a universal approach for preparing stable and reactive colloidal small(~3 nm) gold nanoparticles by using multi-dentate polyoxometalates as pro-tecting agents in non-polar solvents. These nanoparticles exhibit exceptionalstability even under conditions of high concentration, long-term storage,heating and addition of bases. Moreover, they display excellent catalytic per-formance in various oxidation reactions of organic substrates usingmolecularoxygen as the sole oxidant. Our findings highlight the ability of inorganicmulti-dentate ligands with structural stability and robust steric and electroniceffects to confer stability and reactivity upon gold nanoparticles. Thisapproach can be extended to prepare metal nanoparticles other than gold,enabling the design of novel nanomaterials with promising applications.Gold nanoparticles have been the subject of extensive investigation inrecent decades because of their exceptional reactivity and broadapplicability in various research fields, including catalysis, energyconversion, medicine, electronics, optics, magnetic materials andsensors1–8. Owing to their large surface area and abundant surfaceactive sites, small gold nanoparticles exhibit high reactivity, whichrenders them particularly attractive for catalytic applications1–6. In thiscontext, intense research attention has been devoted to the develop-ment of supported gold nanoparticles to address their inherentinstability in past decades; however, the complexity indetermining thereactive sites and the difficulty in realising fine-tuning catalysis arecritical issues that appear during their applications1–4. Hence, exploringgold nanoparticles without supports, which are often called colloidalgold nanoparticles, has become an important avenue towardsachieving distinct performance infine-tuning catalysis2,3,5,9. To stabilisereactive gold nanoparticles in solutions, various chemicals have beenemployed as protecting agents, such as citric acid, alkanethiols,amines, surfactants and organic polymers (Supplementary Table 1,Entries 1–12)9–17. In particular, the introduction of the use of alka-nethiols constituted an important breakthrough in the development ofsmall (1–3 nm) gold nanoparticles with long-term (months) stabilitybased on the strong affinity of alkanethiols to gold nanoparticles(Fig. 1a)10,11. However, the stability of these gold nanoparticles comes atthe expense of their reactivity due to reserved surface active sites andhigh ligand packing density, which hinder substrate access9,12,13.Meanwhile, although organic polymers such as polyvinylpyrrolidone(PVP) are used to prepare gold nanoparticles that perform in alcoholoxidation reactions, their applicability in catalytic reactions requiresfurther validation (Fig. 1a)17. Another concern is that organic protectingagents can react under catalytic conditions, undergoing structuralReceived: 19 June 2023Accepted: 11 January 2024Check for updates1Department of Applied Chemistry, School of Engineering, The University of Tokyo, Tokyo, Japan. 2Department of Chemistry, Graduate School of Science,Tokyo Metropolitan University, Tokyo, Japan. 3Research Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS),Ibaraki, Japan. 4Institute of Engineering Innovation, The University of Tokyo, Tokyo, Japan. e-mail: ksuzuki@appchem.t.u-tokyo.ac.jpNature Communications |          (2024) 15:851 11234567890():,;1234567890():,;http://orcid.org/0000-0002-9376-8953http://orcid.org/0000-0002-9376-8953http://orcid.org/0000-0002-9376-8953http://orcid.org/0000-0002-9376-8953http://orcid.org/0000-0002-9376-8953http://orcid.org/0000-0001-5504-4762http://orcid.org/0000-0001-5504-4762http://orcid.org/0000-0001-5504-4762http://orcid.org/0000-0001-5504-4762http://orcid.org/0000-0001-5504-4762http://orcid.org/0000-0002-8382-8078http://orcid.org/0000-0002-8382-8078http://orcid.org/0000-0002-8382-8078http://orcid.org/0000-0002-8382-8078http://orcid.org/0000-0002-8382-8078http://orcid.org/0000-0002-3311-6283http://orcid.org/0000-0002-3311-6283http://orcid.org/0000-0002-3311-6283http://orcid.org/0000-0002-3311-6283http://orcid.org/0000-0002-3311-6283http://orcid.org/0000-0001-5801-0971http://orcid.org/0000-0001-5801-0971http://orcid.org/0000-0001-5801-0971http://orcid.org/0000-0001-5801-0971http://orcid.org/0000-0001-5801-0971http://orcid.org/0000-0003-3548-5952http://orcid.org/0000-0003-3548-5952http://orcid.org/0000-0003-3548-5952http://orcid.org/0000-0003-3548-5952http://orcid.org/0000-0003-3548-5952http://orcid.org/0000-0003-3886-005Xhttp://orcid.org/0000-0003-3886-005Xhttp://orcid.org/0000-0003-3886-005Xhttp://orcid.org/0000-0003-3886-005Xhttp://orcid.org/0000-0003-3886-005Xhttp://orcid.org/0000-0002-7661-4936http://orcid.org/0000-0002-7661-4936http://orcid.org/0000-0002-7661-4936http://orcid.org/0000-0002-7661-4936http://orcid.org/0000-0002-7661-4936http://orcid.org/0000-0002-8123-1462http://orcid.org/0000-0002-8123-1462http://orcid.org/0000-0002-8123-1462http://orcid.org/0000-0002-8123-1462http://orcid.org/0000-0002-8123-1462http://crossmark.crossref.org/dialog/?doi=10.1038/s41467-024-45066-9&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-024-45066-9&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-024-45066-9&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-024-45066-9&domain=pdfmailto:ksuzuki@appchem.t.u-tokyo.ac.jpchanges and/or detaching from gold nanoparticles during use, leadingto destabilisation12–16. Therefore, the development of a universalmethodology for obtaining stable colloidal gold nanoparticles whilemaintaining high reactivity for various catalytic reactions is animportant yet challenging task.Metal oxide supports have occupied a central role in the field ofmetal nanoparticle catalysts owing to their ability to enhance the sta-bility of gold and other metal nanoparticles, control their electronicstates and achieve synergistic effects3,4,6,18. In this regard, poly-oxometalates (POMs), a type of anionic metal oxide nanoclusters withdiverse properties and functionalities19–26, have emerged as efficientprotecting agents for stabilising metal nanoparticles by means ofefficient coordination, electrostatic repulsion and steric hindrance(Fig. 1a and Supplementary Table 1, Entries 13 − 24)27–37. In addition toan exceptional steric effect, owing to which bulky POM ligands onlybind to a fraction of the metal surface, thus ensuring substrate access,the adaptable structures and properties of functional POM ligandsenable a molecular-level catalyst design for achieving fine-tuning cat-alysis and synergistic effects33–36. However, POM-protected goldnanoparticles sometimes undergo agglomeration in solution duringstorage and/or usage31,32,36,37, which canbe attributed todecompositionand structure transformation of POMs in the commonly used aqueousmedia or occupation of the vacant sites of POMsby alkalimetal cationsand solvent molecules, leading to destabilization of goldFig. 1 | Preparation of gold nanoparticles. a Representative methods for thepreparation of gold nanoparticles using thiol protection, organic polymer protec-tion andpolyoxometalate protection.bThiswork: a non-polar-solvent-basedmulti-dentate polyoxometalate protectionmethod for developing ultra-stable andhighlyreactive gold nanoparticle catalysts (TOA tetraoctylammonium).Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 2nanoparticles38–41. Moreover, the restriction to hydrophilic use besidesthe semi-stability issue of POM-protected gold nanoparticles has alsolimited the exploration of such a feasible molecular-level catalystdesign towards practical catalytic applications32. Accordingly, werecently applied utilisation of POMs to design various metal oxoclusters and metal clusters in organic solvents, where we have shownthat the above-mentioned troubles in aqueous media can beavoided42–44.Here we present a feasible method for obtaining ultra-stable andhighly reactive small gold nanoparticles (of ~3 nm)byemployingmulti-dentate POM ligands in a non-polar solvent system (e.g. toluene,Fig. 1b). Notably, the resultant small gold nanoparticles exhibitexceptional high stability in solution even under harsh conditions suchas high concentration (>5mM metal), long-term storing (>1 year),heating (~90 °C) and addition of bases (e.g. K2CO3 and Cs2CO3), whichare typically required in catalytic applications. These POM-stabilisedgold nanoparticles exhibit high catalytic performance in the aerobicoxidation of alcohols with wide substrate scope and high selectivity toaldehyde or ketone products without changes in the particle size.Additionally, these colloidal gold nanoparticles are effective for var-ious catalytic oxidation reactions using molecular oxygen (O2) as thesole oxidant. This methodology can be extended to various POMligands, metals (e.g. platinum, ruthenium, rhenium and rhodium) andsolvent systems (e.g. p-xylene and 1,2-dichloroethane) to producesmall metal nanoparticles stabilised by multi-dentate POM ligands,which demonstrates its wide applicability and versatility.Results and discussionDevelopment of a non-polar-solvent-based multi-dentate POMprotection methodGold nanoparticles protected by multi-dentate [SiW9O34]10− (SiW9)ligands in toluene (Au-TOASiW9) were prepared according to thefollowing method (Fig. 1b): first, separate aqueous solutions of chlor-oauric acid (HAuCl4) and the sodium salt of SiW9 (NaSiW9) wereprepared and transferred into toluene using tetraoctylammoniumbromide (TOAB) as a phase transfer agent. Afterwards, an aqueoussolution of sodium borohydride (NaBH4) was slowly added to theresulting solution leading to a fast colour change in the toluene phasefrom orange to dark red (Supplementary Fig. 1), followed by a phaseseparation process to yield toluene solution of Au-TOASiW9. A char-acteristic surface plasmon resonance band attributed to the goldnanoparticles of Au-TOASiW9 was observed at 524nm in theultraviolet–visible (UV–vis) spectrum (Supplementary Fig. 2). It shouldbe noted that an excess amount of TOAB and a minimal amount ofNaBH4 are required; the former facilitates the transfer of SiW9 asligands into the toluene phase until reaching the cation exchangeequilibrium (Supplementary Fig. 3) and the latter prevents the reversetransfer of TOASiW9 into the aqueous phase, in which 10 equivalentsof TOAB and 4 equivalents of NaBH4 relative to HAuCl4 were found tobe the optimal conditions (Supplementary Table 2). Transmissionelectron microscopy (TEM) observation of Au-TOASiW9 indicated anaverage particle size (2.9 nm) comparable to that of dodecanethiol-protected gold nanoparticles (Au-dodecanethiol, 2.6 nm), which weresynthesised via the classic Brust–Schiffrin method for comparativepurposes (Fig. 2a, e)10. Additionally, gold nanoparticles protected onlyby the surfactant TOAB (Au-TOAB, 2.5 nm) and fully occupied POM[SiW12O40]4− (Au-TOASiW12, 3.2 nm) were also prepared (Fig. 2g, h).Generally, colloidal gold nanoparticles suffer from stability issuessuch as concentration limitation due to the salting-out effect andagglomeration during use2,5,9,12. In contrast, this methodology allowsemployingmetal precursors at concentrations exceeding 5mM, whichare significantly higher than those used in previously reported meth-ods (Supplementary Table 1, Entries 13 − 25). The stability of the small(~3 nm) gold nanoparticles prepared in this study was evaluated underlong-term storage, heating treatment and addition of bases, which aregeneral requirements for practical catalytic applications insolution1–6,12–15. Au-TOASiW9 exhibited exceptional stability, retainingits particle size and size distribution during storage in toluene at roomtemperature (~25 °C) for over 1 year (Fig. 2b). In contrast, Au-TOABunderwent agglomeration within 2months (Supplementary Fig. 4).Additionally, Au-TOASiW9 displayed an extraordinary stability underheating conditions, maintaining its particle size after continuous stir-ring at 90 °C for 24 h, whereas the particle size of Au-dodecanethiolincreased, and Au-TOAB and Au-TOASiW12 agglomerated and pre-cipitated under the same conditions (Fig. 2c, f–h). Au-TOASiW9 alsoshowed strong resistance to the addition of bases such as Cs2CO3,whereas Au-TOAB and Au-dodecanethiol agglomerated and partiallyprecipitated under the same conditions (Supplementary Fig. 5). Theprecipitation of Au-dodecanethiol can be ascribed to the oxidation ofthiol ligands to disulfides followed by detachment from the goldnanoparticles14, which highlights the stability issue of the ligandsthemselves during use. Notably, even after participating in catalyticaerobic alcohol oxidations in the presence of a base, Au-TOASiW9retained its particle size and size distribution (see Fig. 2d and sub-sequent discussion). Taken together, these results indicate that Au-TOASiW9 is muchmore stable than the well-known Au-dodecanethioland exhibits resistance to heating and addition of bases, which areessential conditions in catalysis, and multi-dentate POM ligands arekey for achieving high stability.To demonstrate the applicability and versatility of this practicalapproach, it was successfully expanded to other organic solvents, suchas p-xylene and 1,2-dichloroethane, and to various types of POMs,including trivacant SiW9, fully occupied SiW12, monovacant[SiW11O39]8− (SiW11) and divacant [SiW10O36]8− (SiW10), furnishingsmall (~3 nm) gold nanoparticles under similar synthetic conditions(Supplementary Fig. 6). Furthermore, small nanoparticles (<5 nm) ofdifferent metals, including platinum, ruthenium, rhenium and rho-dium were successfully synthesised (Supplementary Fig. 7).Next, zeta potentialmeasurements were performed to investigatethe surface state of the gold nanoparticles (Fig. 3a). The observednegative zeta potential for the POM-protected gold nanoparticlesindicated the formation of anionic POM layers surrounding the goldnanoparticles31–33. Au-TOASiW9 exhibited the most negative zetapotential, suggesting the strongest interparticle electrostatic repul-sion that effectively prevents agglomeration. The aberration-correctedannular dark-field scanning TEM (ADF-STEM) images of Au-TOASiW9showed the presence of shells on the surface of the gold nanoparticles(Fig. 3b, Supplementary Fig. 8), and elemental mapping using energy-dispersive X-ray spectroscopy (EDS) further confirmed the formationof an Au-core–POM-shell-like structure (Fig. 3c). Through titrationexperiments of dodecanethiol to Au-TOASiW9 with inspiration byprevious reports29,30, it was estimated that around 30 SiW9 ligandssurrounded a 3 nm gold particle, and surface coverage can be esti-mated as 47% (Supplementary Fig. 9, see explanation in detail). Sub-sequently, X-ray photoemission spectroscopy (XPS) showed that thebinding energy of the Au 4f7/2 region of Au-TOASiW9 (82.3 eV) wasnotably lower than that of bulk Au (84.0 eV) and Au-TOAB withoutPOM protection (83.2 eV, Fig. 3d, Supplementary Fig. 10), indicatingthe anionic status of the POM-protected gold nanoparticles stemmingfrom the electronic interaction between anionic POM ligands and goldnanoparticles31–36. Notably, the binding energy of Au-TOASiW9(82.3 eV) was even lower than those of reported anionic gold nano-particles, PVP-protected gold nanoparticles (Au:PVP, 82.7 eV) andSiW11-protected gold nanoparticles (82.8 eV), which were obtained inaqueous solution17,33. No obvious difference in the binding energieswas found between Au-TOASiW12 and Au-TOAB, highlighting theimportance of a robust electronic interaction between multi-dentatePOM ligands and gold nanoparticles in this system. Moreover, POM-protected gold nanoparticles with similar particle sizes possessedsequentially modulated electronic states (Fig. 3d), providing a feasibleArticle https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 3and effective tool for adjusting the activity of gold nanoparticle cata-lysts, as previously discussed in detail36.To confirm the structures of the POM ligands after hybridisationwith gold nanoparticles, solid samples of TOASiW9 and Au-TOASiW9obtained by evaporating the toluene solvent were characterised. Infra-red (IR) spectroscopies showed that Au-TOASiW9 exhibited similarbands to those of TOASiW9 and NaSiW9 regarding characteristicpeaks in the region from 800 to 1000 cm−1, but differed from those ofsodium tungstate, indicating that the structure of SiW9was preserved(Supplementary Fig. 11)34,36. According to Raman spectroscopies, theFig. 2 | Characterisation and stability test of gold nanoparticles. TEM imagesand size distribution histograms: a Au-TOASiW9. b Au-TOASiW9 after storing for1 year. c Au-TOASiW9 after heating at 90 °C for 24 h. d Au-TOASiW9 after alcoholoxidation in the presence of Cs2CO3 under the conditions shown in Table 1, entry 2.e Au-dodecanethiol. f Au-dodecanethiol after heating at 90 °C for 24h. g Au-TOASiW12 (the bottom picture shows the formation of precipitates from Au-TOASiW12 after heating at 90 °C for 24h). h Au-TOAB (the bottom picture showsthe formation of precipitates from Au-TOAB after heating at 90 °C for 24h).Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 4Fig. 3 | Investigations of the structure and electronic state of Au-TOASiW9.a Zeta potential of Au-TOASiW9, Au-TOASiW10, Au-TOASiW11, Au-TOASiW12,Au-TOAB and Au-dodecanethiol in toluene. b Representative atomic-resolutionannular dark field scanning transmission electron microscopy (ADF-STEM) imageof Au-TOASiW9. c EDS elemental mapping images of Au-TOASiW9. d XPS spectraof various ligand-protected gold nanoparticles and bulk Au (the arrow indicates thetrend of changes in binding energy). e W L3-edge XANES spectra with associatedsecond derivatives of TOASiW9 and Au-TOASiW9. f k3-WeightedW L3-edge EXAFSspectra of TOASiW9 and Au-TOASiW9. g, Relative energy of SiW9-protected goldnanoparticles and models of the optimised structures for different orientationsof SiW9.Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 5characteristic peak of W=Od bonding from POM structures wasobserved at 965 cm−1 for Au-TOASiW9 andTOASiW9, similar to thatofthe tetrabutylammonium salt of SiW9 (TBA4H6SiW9O34, TBASiW9) at970 cm−1 but slightly shifted from that of the sodium salt of SiW9(Na10SiW9O34, NaSiW9) at 940 cm−1 (Supplementary Fig. 12). This canbe ascribed to the presence of sodium cations near the POM anions inthe caseofNaSiW9 increasing theW=Obonding length andweakeningthe bonding strength45. To analyse the precise POM structure, X-rayabsorption fine structure analyses were conducted. Similar patterns ofAu-TOASiW9 and TOASiW9 in the second derivative of white-lineregion and W L3-edge k-space EXAFS spectra, indicating that POMsmaintained their structures after hybridizationwith gold nanoparticles(Fig. 3e, f)36,46. Then, similar patterns of TOASiW9 and NaSiW9 butcompletely different from those of the potassium salt of SiW12(K4SiW12O40, KSiW12) and WO3 in the second derivative of white-lineregion indicated that there existed no obvious structural changes inthe {WO6} octahedra (Supplementary Fig. 13a, b). The W L3-edge k-space EXAFS spectra showed no significant changes between TOA-SiW9 and NaSiW9 while in contrast to that of KSiW12 and WO3,strongly indicating that POM maintained intact structures as well(Supplementary Fig. 13c). In the R-space EXAFS spectra, the peaks atR = 1.2, 1.7 and 3.2 Å assignable to terminalW=O, bridgingW−O−WandW−W, respectively, exhibited no drastic changes from NaSiW9 toTOASiW9, further supporting the preservation of POM structures inthis method (Supplementary Fig. 13d). Finally, structures of POMsduring synthesis of gold nanoparticles were confirmed throughdeliberately transferring them into aqueous phase (SupplementaryFig. 14, Supplementary Table 2, Entry 5). In the IR spectra, the char-acteristic peaks of SiW9 in the region of 500 − 1000 cm−1 were wellconsistent between NaSiW9 and POMs after mixing with gold pre-cursors and sodium borohydride respectively, indicating their intactstructures in current method (Supplementary Fig. 14). These resultsdemonstrate that POM ligands remain stable in this synthetic system,effectively protecting the metal nanoparticles.The interaction between SiW9 and gold nanoparticles was furtherinvestigatedbyperformingfirst-principles calculations.Considering thatthe interaction of bulky tetraoctylammonium (TOA) cations and toluenesolvent molecules with POMs was negligible, structural optimisationswere conducted using one multi-dentate SiW9 ligand with several dif-ferent orientations on the surface of 2-nm-large gold nanoparticlewithout TOA cations and solvent molecules (Fig. 3g). The most stableorientationwas identifiedas that inwhichmulti-dentateSiW9 interactedwith a gold nanoparticle at the vacant site of SiW9. These resultsdemonstrate that the coordination of gold nanoparticles at the vacantsites of SiW9 effectively contributes to the nanoparticle stabilisationand, together with the experimental results, highlighting the essentialrole of an adequate coordination of POM ligands in protecting metalnanoparticles during synthesis, storage and use. Hence, the develop-ment of ultra-stable small metal nanoparticles via POM protection wassuccessfully achieved in a non-polar solvent for the first time (Fig. 1b).Oxidation reactions catalysed by ultra-stable small goldnanoparticlesThe high concentration and extraordinary stability of the developedcolloidal gold nanoparticles even under heating conditions and in thepresence of bases prompted us to explore their performance in liquid-phase catalytic reactions. Anionic gold nanoparticles have beenproved effective in activating O2, which renders them potential cata-lysts in aerobic oxidation reactions; however, only a few studies havebeen conducted so far17,33,36. Therefore, the low binding energy of theAu 4f7/2 region of Au-TOASiW9 motivated us to investigate the cata-lytic performance for various aerobic oxidation reactions in solution,particularly those requiring heating treatment or addition of bases,starting with aerobic alcohol oxidation as a typical model reaction ofgold nanoparticle catalysts.Among the examined POM-protected gold nanoparticle catalysts,Au-TOASiW9 exhibited the highest activity towards the aerobic oxi-dation of benzyl alcohol (1a) in the presence of K2CO3 as a base toselectively furnish benzaldehyde (2a) as a product (Table 1 and Sup-plementary Table 3). A sequentially modulated activity was founddepending on the type of POM in line with the electronic states of goldnanoparticles, which can be correlated with the effective activation ofO2 on anionic gold nanoparticles (Table 1, Entries 1–5; Fig. 3d andSupplementary Fig. 15). In contrast, gold nanoparticles including Au-TOASiW12, Au-TOAB and Au-dodecanethiol exhibited significantly lowcatalytic activity (Fig. 1a; Table 1, Entries 5–7). Notably, no significantTable 1 | Selective aerobic oxidation of benzyl alcohol (1a) to benzaldehyde (2a)aOO2 (1 atm), K2CO3Toluene, ~25 °C, 24 hColloidal gold nanoparticles (Au: 4 mol%)OH1a 2aEntry Catalyst Yield (%) Particle size (nm)Before use After use1 Au-TOASiW9 75 2.9 2.82 Au-TOASiW9b 92 2.9 2.93 Au-TOASiW10 17 3.4 3.34 Au-TOASiW11 24 2.9 3.05 Au-TOASiW12 4 3.2 4.36 Au-TOAB 3 2.5 5.67 Au-dodecanethiol <1 2.6 2.78 Au-dodecanethiolb <1 2.6 Precipitationc9 TOASiW9 <1 – –10 Au-TOASiW9, Ard <1 2.9 −aReaction conditions: 1a (0.25mmol), 3mL toluene solution of colloidal gold nanoparticles (Au: 4mol%), K2CO3 (0.5mmol), room temperature (~25 °C), O2 (1 atm), 24 h.bCs2CO3 (0.5mmol) was used instead of K2CO3 (0.5mmol).cDisulfide was detected in the reaction solution after the reaction.dUnder Ar (1 atm). All the reaction yields were determined via GC analysis using biphenyl as an internal standard.Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 6changes were observed in the particle size of the gold nanoparticlesprotected with SiW9, SiW10 and SiW11 even after the reaction, and nosignificant decrease was observed in the catalytic activity of Au-TOA-SiW9during the reaction (Supplementary Figs. 16 and 17). Furthermore,Raman spectra showed that Au-TOASiW9 after the catalytic oxidationof 1a exhibited similar characteristic bands with Au-TOASiW9, indi-cating that the structure of SiW9 was preserved (SupplementaryFig. 12). In contrast, an evident agglomeration of gold nanoparticles wasobservedduring the reactionusing fully occupiedSiW12 as aprotectingligand (Supplementary Fig 17). These results indicated that thepresenceof vacant (coordination) sites in the POM ligands is essential for pro-tecting the metal nanoparticles. Nevertheless, Au-TOAB underwentmore severe agglomeration than Au-TOASiW12, indicating that POMligands can generally function as stabilising agents in comparison toconventional organic substances31,32. Considering that the particle sizeof Au-dodecanethiol was kept after the reaction (Table 1, Entry 7; Sup-plementary Fig. 17), its low catalytic activity was likely due to difficultiesin substrate access and strong bonding of thiol ligands to the surface ofgold nanoparticles12,13. Meanwhile, Au-TOASiW9 exhibited higher cata-lytic activitywhenusingCs2CO3 as a base than in thepresence ofK2CO3,and 2a was still selectively produced from 1a while maintaining thesmall particle size (Table 1, Entry 2; Fig. 2d). In contrast, the reactionbarely proceeded using Au-dodecanethiol and Cs2CO3 due to the oxi-dation of the thiol ligands, which led to precipitation of gold nano-particles (Table 1, Entry 8)14.Additionally, TOASiW9 exhibited no activity and Au-TOASiW9did not promote the reaction under argon (Ar) atmosphere, confirm-ing gold nanoparticles as the active sites andO2 as the terminal oxidantin this catalysis (Table 1, Entries 9 and 10). The kinetic isotope effect(KIE) was then examined for Au-TOASiW9-catalysed oxidation of 1a.Under O2 atmosphere (1 atm), a much higher reaction rate wasobserved for 1a than benzyl-α,α-d2 alcohol (kH/kD = 3.2, SupplementaryFig. 18), indicating that C−Hcleavage can be the turnover limiting step.When the reactionwas carried out under air atmosphere (O2, 0.2 atm),no significant KIE was observed (kH/kD = 1.2, Supplementary Fig. 19),suggesting that O2 adsorption and/or activation can be turnover lim-iting step under air atmosphere (Supplementary Fig. 15).To investigate the effect of cations, preparation of gold nano-particles protected with SiW9 was examined using different cationssuch as tetrahexylammonium (THA), tetradecylammonium (TDA), andcetyltrimethylammonium (CTA) instead of TOA. Although TDA can beused for preparation of gold nanoparticles, incomplete phase transferwas observed for THA and CTA owing to less hydrophobicity of metalprecursors and emulsion formation, respectively (SupplementaryFig. 20a). Theobtainedgoldnanoparticles usingTDAcations andSiW9(Au-TDASiW9) possessed similar particle sizes of 3 nm and similarcatalytic reactivity to those of Au-TOASiW9, suggesting that coun-tercations did not necessarily facilitate the reaction (SupplementaryFig. 20b, c). These results provide direct evidence that multi-dentatePOM ligands do not only contribute to stabilising the small (~3 nm)gold nanoparticles during their preparation but also allow retainingtheir catalytically active sites, enabling the modulation of the electro-nic states of gold nanoparticles for activity control.Furthermore, the Au-TOASiW9-catalysed reaction demonstrated abroad substrate scope, enabling the conversion of various primary andsecondary alcohols to thecorresponding aldehydeandketoneproducts,respectively (Fig. 4a, Supplementary Fig. 21).WhenCs2CO3was used as abase at room temperature (~25 °C), Au-TOASiW9 efficiently promotedthe oxidation of benzyl alcohol as well as benzylic alcohols with eitherelectron-withdrawing or electron-donating groups, affording the corre-sponding benzaldehydes (2a–2i). Au-TOASiW9 successfully catalysedthe oxidation of a heteroaromatic alcohol to the corresponding alde-hyde (2j). In addition, an α,β-unsaturated alcohol afforded the corre-sponding α,β-unsaturated aldehyde (2k), and aromatic and aliphaticsecondary alcohols also gave the corresponding ketones (2l–2q).Finally, the applicability of Au-TOASiW9 towards various oxida-tion reactions using O2 as the sole oxidant was confirmed (Fig. 4b,Supplementary Fig. 22). In addition to the high reactivity and wideapplicability in alcohol oxidation reactions, Au-TOASiW9 catalysedthe oxidative dehydrogenation of N-methyl piperidone in toluene inthe presence of Cs2CO3 as a base (Supplementary Fig. 23a). Subse-quently, the cross-dehydrogenative coupling (CDC) reaction of aterminal alkyne and a hydrosilane in the presence of O2 as thehydrogen acceptor efficiently proceeded using Au-TOASiW9 toafford the desired alkynylsilane without formation of hydrosilylationproducts. In contrast, when the CDC reaction was performed underthe same conditions using typical supported gold nanoparticle cata-lysts, the undesirable hydrosilylation reaction occurred to a certainextent (Supplementary Fig. 23b)47. The observed high selectivity tothe CDC product when using Au-TOASiW9 can be attributed to a fastAu–hydride oxidation induced by activated oxygen species on theanionic gold nanoparticles preventing the hydrosilylation reaction.Moreover, in a regioselective alkynylation of a tertiary amine, Au-TOASiW9 demonstrated comparable reactivity to the reported cat-alytic system using supported gold nanoparticles in the presence ofZnBr2 (Supplementary Fig. 22c)48. These findings demonstrate thatthismethodology provides a universal protocol for the preparation ofcolloidal metal nanoparticle catalysts simultaneously exhibiting highreactivity, stability and selectivity that can competewith conventionalsupported metal nanoparticle catalysts.In summary, we have developed a non-polar-solvent-based multi-dentate POM protection strategy for obtaining ultra-stable and cata-lytically active colloidal gold nanoparticles. These small gold nano-particles exhibited remarkable tolerance towards high concentrationconditions (>5mM metal), long-term storage (>1 year), heating treat-ment (>90 °C) and addition of bases (e.g. Cs2CO3) without undergoingchanges in particle size and size distribution. They exhibited a highreactivity and selectivity in various catalytic oxidation reactions usingO2 as the sole oxidant, including alcohol oxidation, piperidone dehy-drogenation, terminal alkyne–hydrosilane cross-dehydrogenativecoupling and tertiary amine alkynylation. The robust electronic andmoderate steric effects of multi-dentate POM ligands are consideredessential to achieve an extraordinary catalytic performance. The wideapplicability to various POM ligands, solvent systems and metalnanoparticles, and the broad reaction scope render this approachhighly promising for solving the compromise between reactivity andstability of metal-nanoparticle-based materials in diverse fieldsincluding catalysis, biochemistry, photochemistry, coordinationchemistry, pharmaceuticals, physiochemistry and materials science.MethodsInstruments and reagentsGas chromatography (GC) analyses were conducted on ShimadzuGC2014 equipped with a flame ionization detector (FID) and anInertCap-5 capillary column (30m × 0.25mm × 0.25 μm) using Shi-madzu CR8A Chromatopac Data Processor for area calculations. GCmass spectrometry (GC-MS) analyses were performed by ShimadzuGCMS-QP2020 equipped with an InertCap-5 MS/NP capillary column(30m × 0.25mm × 0.25 μm) at an ionization voltage of 70 eV.Inductively coupled plasma atomic emission spectroscopy (ICP-AES)analyses were conducted by Shimadzu ICPS-8100. Transmissionelectron microscope (TEM) observations were conducted by JEM-2000EX and JEM-2010F at an acceleration voltage of 200 kV, andscanning transmission electron microscopy (STEM) observationswere conducted by JEM-ARM200F Thermal FE at an accelerationvoltage of 200 kV and JEM-ARM300CF with a cold FE at an accelera-tion voltage of 300 kV. The illumination semi-angle and the collectionsemi-angle for atomic-resolution annular dark field (ADF-)STEM ima-ges were acquired 30mrad and 32–200mrad, respectively. W L3-edgeX-ray absorption spectroscopy (XAS) was carried out at the BL01B1Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 7beamline of SPring-8. X-ray absorption fine structure (XAFS) mea-surements were conducted in transmission mode using a Si(111)double-crystal monochromator. The X-ray absorption near-edgestructure (XANES) and extended X-ray absorption fine structure(EXAFS) spectra were analysed using xTunes programme49. Pre-edgebackgrounds were subtracted using a McMaster equation. EXAFSbackgrounds were subtracted using a cubic spline method (splinerange = 5). W L3-edge EXAFS spectra in k-space were obtained as k3-weighted χ spectra after normalization. The X-ray photoelectronspectroscopy (XPS) was performed on ULVAC-PHI PHI5000 Versa-ProbeIII at the Advanced Characterisation Nanotechnology Platformof The University of Tokyo. The samples were embedded in In foil andbrought in the introduction chamber. The fitting of experimental datawas conducted using a Multipak software (version 9.2.0.5, by Ulvac-phi, inc.) in which the Shirley method was used for the backgroundand a Gauss−Lorentz type function was performed for fitting. Thebinding energies were calibrated by using the C 1 s signal of C−Cbonding at 284.9 eV. Under these conditions, the Au 4f7/2 signal ofbulk Au (CAS No. 7440-57-5) was located at 84.0 eV. Solution-stateultraviolet–visible (UV–Vis) spectra were measured on JASCO V-770spectrometer with a 1 cm quartz cell at room temperature (~25 °C).Zeta-potential measurement were conducted on Malvern ZetasizerNanoZS at a backscatter mode and a working voltage of 40V wasadopted. Infra-red (IR) spectra weremeasured on a JASCO FT/IR-4100using the attenuated total reflection method. Raman spectra weremeasured on a JASCO NRS-5100. All chemical reagents were obtainedfrom Tokyo Chemical Industry, Aldrich, Kanto Chemical, or FUJIFILMWako Pure Chemical (reagent grade) without pretreatment. Inorganicsalts of SiW9, SiW10, SiW11, SiW12 (NaSiW9, Na10SiW9O34;K8SiW10O36; K8SiW11O39; K4SiW12O40) and a tetrabutylammonium saltFig. 4 | Reaction scope. a Substrate scope of Au-TOASiW9-catalysed aerobicalcohol oxidation. Reaction conditions: alcohol (0.25mmol), 3mL toluene solutionof Au-TOASiW9 (Au: 4mol%), Cs2CO3 (0.5mmol), ~25 °C (room temperature), O2(1 atm), 24 h. b Schematic of various aerobic oxidation reactions catalysed by Au-TOASiW9. Detailed reaction conditions are described in Method section.Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 8of SiW9 (TBASiW9, (C16H36N)4H6SiW9O34) were prepared accordingto the reported procedures40,50.Preparation of gold nanoparticlesAu-TOASiW9 was prepared as follows: an aqueous solution of HAuCl4(20mL, 5mM) was mixed with a solution of TOAB in toluene (20mL,50mM). The two-phase mixture was vigorously stirred until all theHAuCl4was transferred into the organic layer, and an aqueous solutionofNaSiW9 (20mL, 5mM) was then added to the organic layer and theresulting solution was stirred for 30min, followed by a phase-separation to yield the organic layer. A freshly prepared aqueoussolution of NaBH4 (20mL, 20mM) was slowly added, and the organicphase was immediately separated and filtrated twice using hydro-phobic filters to remove residual water, affording a toluene solution ofAu-TOASiW9. For optimization of TOAB and NaBH4 usage (Supple-mentary Table 2), the amounts of NaSiW9 and HAuCl4 were holdconstant, and the same procedures were used with the exception ofvarying the amounts of TOAB and NaBH4.Au-TOASiW10, Au-TOASiW11, Au-TOASiW12 and Au-TOAB wereprepared using similar procedures as that for Au-TOASiW9, except forusing different POMs (K8SiW10O36, K8SiW11O39, K4SiW12O40) or no POMin the case of Au-TOAB. Gold nanoparticles in other organic solventswerepreparedusing the sameprocedures asdescribed forAu-TOASiW9in toluene but using p-xylene and 1,2-dichloroethane instead of toluene.POM-protected platinum, ruthenium and rhenium nanoparticles(Pt-TOASiW9, Ru-TOASiW9, Re-TOASiW9) were prepared as descri-bed above for Au-TOASiW9, except for using Na2PtCl6, K2RuCl5 andK2ReCl6, respectively, instead of HAuCl4. POM-protected rhodiumnanoparticles (Rh-TOASiW9) was prepared via a slightly modifiedmethodology using RhCl3 without conducting phase separation untilNaBH4 reduction was complete. Au-dodecanethiol was preparedaccording to the Brust–Schiffrin method10, except by decreasing theconcentration for comparison in catalytic test as follows: an aqueoussolution ofHAuCl4 (5mL, 18mM)wasmixedwith a solution of TOAB intoluene (20mL, 20mM). The two-phasemixturewas vigorously stirreduntil all the HAuCl4 was transferred into the organic layer, and dode-canethiol (0.9mmol, 10 equivalents to Au) was then added to theorganic phase. A freshly prepared aqueous solution of NaBH4 (5mL,200mM)was slowly addedwith vigorous stirring. After further stirringfor 3 h, the organic phase was separated and filtrated twice usinghydrophobic filters to remove residual water to give a toluene solutionof Au-dodecanethiol, that was directly usedwithout further treatment.Titration experimentsUV–vis titration procedure, based on changes in SPR absorbance29,30,was used to quantify the replacement of TOASiW9 by thiolates on thesurfaces of the goldnanoparticles. A toluene solutionof dodecanethiol(5mM) was gradually added to a toluene solution of Au-TOASiW9(0.5mM Au), and the changes of the absorbance of SPR bands weremonitored by UV–vis spectra. The temperature was maintained at25.0 ± 0.1 °C.Procedure for catalytic reactionsThe aerobic alcohol oxidation reaction was conducted as follows: 1a(0.25mmol), biphenyl as an internal standard (0.25mmol), Au-TOASiW9 (2mL toluene solution containing 0.01mmol Au),appended toluene for better dispersion of the base (1mL) and amagnetic stirrer bar were added to a Pyrex glass reactor, which wasthen purged with O2 gas and sealed with a screw cap. The solutionwas stirred at room temperature (~25 °C) for 24 h. After completionof the reaction, the substrate conversions and product yields weredetermined via GC analysis. For the reaction under Ar (1 atm), freeze-pump-thaw cycles were carried out and the reactor was connected toa balloon filled with an Ar gas.The oxidative dehydrogenation of piperidone was conducted asfollows: 1-methyl-4-piperidone (0.25mmol), biphenyl as an internalstandard (0.25mmol), Au-TOASiW9 (5mL toluene solution containing0.025mmol Au), and a Teflon-coated magnetic stirrer bar were addedto a Pyrex glass reactor, whichwas purged with O2 gas and then sealedwith a screw cap. The solution was stirred at 70 °C for 24 h. Aftercompletion of the reaction, the substrate conversions and productyields were determined via GC analysis.The cross-dehydrogenative coupling reaction of a terminal alkyneand a hydrosilane was conducted as follows: ethynylbenzene(0.25mmol), triethylsilane (0.3mmol), biphenyl as an internal stan-dard (0.25mmol), Au-TOASiW9 (2mL toluene solution containing0.01mmol Au) and a Teflon-coatedmagnetic stirrer barwere added toa Pyrex glass reactor, which was purged with O2 gas and sealed with ascrew cap. The solution was stirred at 90 °C for 24 h. After completionof the reaction, the substrate conversions and product yields weredetermined via GC analysis.The regioselective alkynylation of a tertiary amine was conductedas follows: 1-methylpiperidine (0.5mmol), ethynylbenzene(0.25mmol), ZnBr2 (0.25mmol), biphenyl as an internal standard(0.25mmol), Au-TOASiW9 (2mL toluene solution containing0.01mmol Au) and a Teflon-coatedmagnetic stirrer barwere added toa Pyrex glass reactor, which was then purged with O2 and sealed with ascrew cap. The solution was stirred at 90 °C for 24 h. After completionof the reaction, the substrate conversions and product yields weredetermined via GC analysis.Computational methodsFirst-principles density functional theory calculations were performedby using the CONQUEST code. Double-ζ plus polarization pseudoatomic orbital (PAO) basis functions were used with norm-conservingpseudopotentials. 5 s and 5p semi-core PAOs were used for Au and W.The PAO ranges [bohr] are as follows: Au [5 s, 5p, 5d, 6 s, 6p] = [2.81,3.36, (7.12, 3.74), (7.12, 3.74), 7.12], Si [3 s, 3p, 3d] = [(7.12, 4.02), (7.12,4.02), 7.12], W [5 s, 5p, 6 s, 5d, 6p] = [3.22, 3.85, (7.97, 4.18), (7.97, 4.18),7.97], and O [2 s, 2p, 3d] = [(4.91, 2.58), (4.91, 2.58), 4.91]. The PBEexchange-correlation functional was used. The geometry of SiW9-protected Au nanoparticle (denoted as Au-SiW9) with differentorientations was optimised, in which Au was modelled as a nano-particle of diameter about 2 nm (consisting of 309 atoms) in Ohsymmetry. According to the formal charge of SiW9, we set the chargeof the unit cell to be –10 for Au-SiW9 with neutralizing backgrounduniform charge density. The cubic unit cells of 75.6×75.6×75.6 bohr3were used for all systems. The grid spacing was 0.21 bohr.Data availabilityThe data that support the findings of this study are available from thecorresponding author upon request.References1. Ciriminna, R., Falletta, E., Pina, C. D., Teles, J. H. & Pagliaro, M.Industrial applications of gold catalysis. Angew. Chem. Int. Ed. 55,14210–14217 (2016).2. Zhang, Y., Cui, X., Shi, F. & Deng, Y. Nano-gold catalysis in finechemical synthesis. Chem. Rev. 112, 2467–2505 (2012).3. Sankar, M. et al. Role of the support in gold-containingnanoparticles as heterogeneous catalysts. Chem. Rev. 120,3890–3938 (2020).4. van Deelan, T. W., Mejía, C. H. & de Jong, K. P. Control of metal-support interactions in heterogeneous catalysts to enhance activityand selectivity. Nat. Catal. 2, 955–970 (2019).5. Mikami, Y., Dhakshinamoorthy, A., Alvaro, M. & García, H. Catalyticactivity of unsupported gold nanoparticles. Catal. Sci. Technol. 3,58–69 (2013).Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 96. Clavero, C. Plasmon-induced hot-electronic generation at nano-particle/metal-oxide interfaces for photovoltaic and photocatalyticdevices. Nat. Photonics 8, 95–103 (2014).7. Anker, J. N. et al. Biosensing with plasmonic nanosensors. Nat.Mater. 7, 442–453 (2008).8. Sanchis-Gual, R., Coronado-Puchau, M., Mallah, T. & Coronado, E.Hybrid nanostructures based on gold nanoparticles and functionalcoordination polymers: chemistry, physics and applications inbiomedicine, catalysis and magnetism. Coord. Chem. Rev. 480,215025 (2023).9. Zhao, P., Li, N. & Astruc, D. State of the art in gold nanoparticlesynthesis. Coord. Chem. Rev. 257, 638–665 (2013).10. Brust, M., Walker, M., Bethell, D., Schiffrin, D. & Whyman, R.Synthesis of thiol-derivatised gold nanoparticles in a two-phaseliquid–liquid system. J. Chem. Soc. Chem. Commun. https://doi.org/10.1039/C39940000801 (1994).11. Häkkinen, H. The gold-sulfur interface at the nanoscale. Nat. Chem.4, 443–455 (2012).12. Niu, Z. & Li, Y. Removal and utilization of capping agents in nano-catalysis. Chem. Mater. 26, 72–83 (2014).13. Lu, L., Zou, S. & Fang, B. The critical impacts of ligands on hetero-geneous nanocatalysis: a review. ACS Catal. 11, 6020–6058 (2021).14. Kang, J. et al. Oxidative elimination and reductive addition of thiol-terminated polymer ligands to metal nanoparticles. Angew. Chem.Int. Ed. 61, e202202405 (2022).15. Zhang, L., Wei, Z., Meng, M., Ung, G. & He, J. Do polymer ligandsblock the catalysis ofmetal nanoparticles? Unexpected importanceof bindingmotifs in improving catalytic activity. J. Mater. Chem. A8,15900–15908 (2020).16. Larson, T. A., Joshi, P. P. & Sokolov, K. Preventing protein adsorptionand macrophage uptake of gold nanoparticles via a hydrophobicshield. ACS Nano 6, 9182–9190 (2012).17. Tsunoyama, H., Ichikuni, N., Sakurai, H. & Tsukuda, T. Effect ofelectronic structures of Au clusters stabilized by poly(N-vinyl-2-pyrrolidone) on aerobic oxidation catalysis. J. Am. Chem. Soc. 131,7086–7093 (2009).18. Mitsudome, T. et al. Design of a silver-cerium dioxide core-shellnanocomposite catalyst for chemoselective reduction reactions.Angew. Chem. Int. Ed. 51, 136–139 (2012).19. Pope, M. T. Heteropoly and Isopoly Oxometalates (Springer, 1983).20. Sadakane, M. & Steckhan, E. Electrochemical properties of poly-oxometalates as electrocatalysts. Chem. Rev. 98, 219–238 (1998).21. Izarova, N. V., Pope, M. T. & Kortz, U. Noble metals in poly-oxometalates. Angew. Chem. Int. Ed. 51, 9492–9510 (2012).22. Miras, H. N., Yan, J., Long, D.-L. & Cronin, L. Engineering poly-oxometalates with emergent properties. Chem. Soc. Rev. 41,7403–7430 (2012).23. Lv, H. et al. Polyoxometalate water oxidation catalysts and theproduction of green fuel. Chem. Soc. Rev. 41, 7572–7589 (2012).24. Lechner, M., Güttel, R. & Streb, C. Challenges in polyoxometalate-mediated aerobic oxidation catalysis: catalyst development meetsreactor design. Dalton Trans. 45, 16716–16726 (2016).25. Weinstock, I. A., Schreiber, R. E. & Neumann, R. Dioxygen in poly-oxometalatemediated reactions.Chem.Rev. 118, 2680–2717 (2018).26. Suzuki, K., Mizuno, N. & Yamaguchi, K. Polyoxometalate photo-catalysis for liquid-phase selective organic functional group trans-formations. ACS Catal. 8, 10809–10825 (2018).27. Lin, Y. & Finke, R. G. Novel polyoxoanion- and Bu4N+-stabilized,isolable, and redissolvable, 20−30-Å Ir~300−900 nanoclusters: thekinetically controlled synthesis, characterization, and mechanismof formation of organic solvent-soluble, reproducible size, andreproducible catalytic activity metal nanoclusters. J. Am. Chem.Soc. 116, 8665–8353 (1994).28. Aiken, J.D.&Finke,R.G.Nanocluster formation synthetic, kinetic, andmechanistic studies. The detection of, and then methods to avoid,hydrogen mass-transfer limitations in the synthesis of polyoxoanion-and tetrabutylammonium-stabilized, near-monodisperse 40 ± 6 ÅRh(0) nanoclusters. J. Am. Chem. Soc. 120, 9545–9554 (1998).29. Wang, Y. et al. Self-assembly and structure of directly imagedinorganic-anion monolayers on a gold nanoparticle. J. Am. Chem.Soc. 131, 17412–17422 (2009).30. Wang, Y., Zeiri, O., Neyman, A., Stellacci, F. & Weinstock, I. A.Nucleation and island growth of alkanethiolate ligand domains ongold nanoparticles. ACS Nano 6, 629–640 (2012).31. Wang, Y. & Weinstock, I. A. Polyoxometalate-decorated nano-particles. Chem. Soc. Rev. 41, 7479–7496 (2012).32. Xia, K., Yamaguchi, K. & Suzuki, K. Recent advances in hybridmaterials of metal nanoparticles and polyoxometalates. Angew.Chem. Int. Ed. 62, e202214506 (2023).33. Zhang,M., Hao, J., Neyman, A.,Wang, Y. &Weinstock, I. A. Influenceof polyoxometalate protecting ligands on catalytic aerobic oxida-tion at the surface of gold nanoparticles in water. Inorg. Chem. 56,2400–2408 (2017).34. Martin, C. et al. Redox-active hybrid polyoxometalate-stabilizedgold nanoparticles. Angew. Chem. Int. Ed. 59, 14331–14335 (2020).35. Kikkawa, S. et al. Dual functional catalysis of [Nb6O19]8– modifiedAu/Al2O3. Chem. Commun. 58, 9018–9021 (2022).36. Xia, K. et al. Supported anionic gold nanoparticle catalystsmodifiedusing highly negatively charged multivacant polyoxometalates.Angew. Chem. Int. Ed. 61, e202205873 (2022).37. Martín, S. et al. Integrated synthesis of gold nanoparticles coatedwith polyoxometalate clusters. Inorg. Chem. 58, 4110–4116 (2019).38. Gumerova, N. I. & Rompel, A. Polyoxometalates in solution: spe-ciation under spotlight. Chem. Soc. Rev. 49, 7568–7601 (2020).39. Misra, A., Kozma, K., Streb, C. & Nyman, M. Beyond charge balance:counter-cations in polyoxometalate chemistry. Angew. Chem. Int.Ed. 59, 596–612 (2020).40. Minato, T., Suzuki, K., Yamaguchi, K. & Mizuno, N. Alkoxides of tri-vacant lacunary polyoxometalates. Chem. Eur. J. 23, 14213–14220(2017).41. Li, C., Mizuno, N., Yamaguchi, K. & Suzuki, K. Self-assembly ofanionic polyoxometalate–organic architectures based on lacunaryphosphomolybdates and pyridyl ligands. J. Am. Chem. Soc. 141,7687–7692 (2019).42. Yonesato, K. et al. Surface-exposed silver nanoclusters insidemolecular metal oxide cavities. Nat. Chem. 15, 940–944 (2023).43. Yonesato, K. et al. Controlled assembly synthesis of atomicallyprecise ultrastable silver nanoclusters with polyoxometalates. J.Am. Chem. Soc. 141, 19550–19554 (2019).44. Minato, T. et al. Robotic stepwise synthesis of hetero-multinuclearmetal oxo clusters as single-molecule magnets. J. Am. Chem. Soc.143, 12809–12816 (2021).45. Su, Y., Balmer,M. L. & Bunker, B. C. Raman spectroscopic studies ofsilicotitanates. J. Phys. Chem. B 104, 8160–8169 (2000).46. Yamazoe, S., Hitomi, Y., Shishido, T. & Tanaka, T. XAFS study oftungsten L1- and L3 edges: structural analysis of WO3 species loa-ded on TiO2 as a catalyst for photo-oxidation of NH3. J. Phys. Chem.C 112, 6869–6879 (2008).47. Yamaguchi, K., Wang, Y., Oishi, T., Kuroda, Y. & Mizuno, N. Het-erogeneously catalyzed aerobic cross-dehydrogenative couplingof terminal alkynes and monohydrosilanes by gold supported onOMS-2. Angew. Chem. Int. Ed. 52, 5627–5630 (2013).48. Yatabe, T. & Yamaguchi, K. Regiospecific α-methylene functionali-zation of tertiary amines with alkynes via Au-catalysed concertedone-proton/two-electron transfer to O2. Nat. Commun. 13,6505 (2022).49. Asakura, H. et al. xTunes: A newXASprocessing tool for detailed andon-the-fly analysis. Radiat. Phys. Chem. 175, 108270 (2020).50. Ginsberg, A. P. (Ed) Inorg. Synth. 27, Ch. 3 (John Wiley & Sons,Inc, 1990).Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 10https://doi.org/10.1039/C39940000801https://doi.org/10.1039/C39940000801AcknowledgementsWe gratefully acknowledge the financial support from JST FOREST(JPMJFR213M for K.S., JPMJFR2033 for R.I.), JST PRESTO (JPMJPR18T7 forK.S., JPMJPR19T9 for S.Y., JPMJPR20T4 for A.N., JPMJPR227A for T.Y.),JSPS KAKENHI (22H04971 for K.Ya), and the JSPS Core-to-Core pro-gramme. XAFS measurements were conducted at SPring-8 with theapproval of the Japan Synchrotron Radiation Research Institute (pro-posal numbers: 2023A1732, 2023A1554, 2022B1860, 2022B1684). A partof this work was supported by Advanced Research Infrastructure forMaterials and Nanotechnology in Japan (ARIM) of the Ministry of Edu-cation, Culture, Sports, Science and Technology (MEXT), Grant NumberJPMXP1222UT0184 and JPMXP1223UT0029. We thank Ms. Mari Morita(The University of Tokyo) for assistance with the STEM-EDS analysis.Author contributionsK.S. conceived and directed the project. K.X. designed and performedmost of experiments including synthesis, analysis and catalytic reac-tions. K.Ya. and T.Y. contributed to catalytic reactions and the project.S.K., S.Y. and K.Yo. carried out XAFS measurements and analysis. A.N.performed first-principles calculations. R.I., N.S. and Y.I. contributed tomicroscopy analysis. K.X. and K.S.wrote themanuscript, with input fromall the co-authors.Competing interestsThe authors declare no competing interests.Additional informationSupplementary information The online version containssupplementary material available athttps://doi.org/10.1038/s41467-024-45066-9.Correspondence and requests for materials should be addressed toKosuke Suzuki.Peer review information Nature Communications thanks the anon-ymous reviewers for their contribution to the peer review of this work. Apeer review file is available.Reprints and permissions information is available athttp://www.nature.com/reprintsPublisher’s note Springer Nature remains neutral with regard to jur-isdictional claims in published maps and institutional affiliations.Open Access This article is licensed under a Creative CommonsAttribution 4.0 International License, which permits use, sharing,adaptation, distribution and reproduction in any medium or format, aslong as you give appropriate credit to the original author(s) and thesource, provide a link to the Creative Commons license, and indicate ifchanges were made. The images or other third party material in thisarticle are included in the article’s Creative Commons license, unlessindicated otherwise in a credit line to the material. If material is notincluded in the article’s Creative Commons license and your intendeduse is not permitted by statutory regulation or exceeds the permitteduse, you will need to obtain permission directly from the copyrightholder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.© The Author(s) 2024Article https://doi.org/10.1038/s41467-024-45066-9Nature Communications |          (2024) 15:851 11https://doi.org/10.1038/s41467-024-45066-9http://www.nature.com/reprintshttp://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/ Ultra-stable and highly reactive colloidal gold nanoparticle catalysts protected using multi-dentate metal oxide nanoclusters Results and discussion Development of a non-polar-solvent-based multi-dentate POM protection�method Oxidation reactions catalysed by ultra-stable small gold nanoparticles Methods Instruments and reagents Preparation of gold nanoparticles Titration experiments Procedure for catalytic reactions Computational methods Data availability References Acknowledgements Author contributions Competing interests Additional information