Doped metal clusters as bimetallic AuCo nanocatalysts: insights into structural dynamics and correlation with catalytic activity by in situ spectroscopy

Co-doped Au25 nanoclusters with different numbers of doping atoms were synthesized and supported on CeO2. The catalytic properties were studied in the CO oxidation reaction. In all cases, an enhancement in catalytic activity was observed compared to the pure Au25 nanocluster catalyst. Interestingly, a different catalytic performance was obtained depending on the number of Co atoms within the cluster. This was related to the mobility of atoms within the cluster’s structure under pretreatment and reaction conditions, resulting in active CoAu nanoalloy sites. The evolution of the doped Au clusters into nanoalloys with well-distributed Co atoms within the Au cluster structure was revealed by combined XAFS, DRIFTS, and XPS studies. Overall, these studies contribute to a better understanding of the dynamics of doped nanoclusters on supports upon pretreatment and reaction, which is key information for the future development and application of bimetallic nanocluster (nanoalloy) catalysts.

Synthesis of Au25(SC2H4Ph)18. The synthesis was carried out following a protocol by Shivare et al. 2 50ml of THF and 500mg of HAuCl4 •3H2O were mixed with 1.2eq. of TOAB and stirred for 10 min. Then, 0.85 ml of phenylethyl mercaptan was added to the solution and stirred until transparent. 480 mg of NaBH4 in 10 ml of ice-cold water was added at once, leading to a dark brown reaction mixture. The solution was stirred for 4 days under ambient conditions, before the solvent was evaporated and the precipitate was washed several times with methanol. The clusters were then separated by Size Exclusion Electronic Supplementary Material (ESI) for Faraday Discussions. This journal is © The Royal Society of Chemistry 2022 Chromatography (SEC) and their purity evaluated by Ultraviolet-Visible (UV-Vis) spectroscopy and matrix-assisted laser desorption ionization (MALDI) mass spectrometry.

Characterization Techniques
UV-Vis spectra of nanoclusters dissolved in CH2Cl2 were recorded on a Perkin Elmer Lambda 750 UV-Vis spectrometer. Diffuse Reflectance Spectroscopy (DRS) of the nanoparticle catalysts was performed using the same instrument coupled to a 60 mm integration sphere.

UV-Vis spectra and MALDI mass spectra of the Au25 nanocluster sample in solution
Both the UV-Vis spectrum of [Au25(SC2H4Ph)18]and the dominant peak at m/z = ≈7394 in the MALDI mass spectrum are in good agreement with the reported data. 2 Figure S1. UV-Vis (left) and MALDI mass spectrum (right) of Au25(SC2H4Ph)18. Figure S2. CO K-edge XANES of bimetallic cluster with reference materials Figure S3. In situ IR spectra during CO oxidation reaction from room temperature to 250°C. X-ray photoelectron spectroscopy (XPS) measurements of the spent catalysts were performed on a lab based in situ NAP-XPS system equipped with a Phoibos 150 NAP hemispherical analyzer and a XR 50 MF X-ray source (microfocus), all SPECS GmbH. Spectra were recorded with monochromatic Al Kα radiation and data were analysed with the CasaXPS software. Peaks were fitted after linear background subtraction with Gauss−Lorentz sum functions. The spectra were referenced to the Fermi edge and the C1s signal. Peak positions and full width at half-maximum (FWHM) were left unconstrained. Au4f peaks were fitted with 3.7 eV doublet separation and a fixed ratio of 4:3 for Au4f 7/2 and Au4f 5/2 . Figure S4. Fitted Au4f spectra (left) and C1s spectra (right) of the Au25/CeO2 sample as well as the samples extracted with Acetonitrile, DCM and Acetone.
The XPS analysis shows that the undoped reference sample (Au/CeO2) has a binding energy for the Au4f 7/2 transition of 83.0 eV. This is about 1 eV lower than typical values. This behavior could be due to nanosize or charge transfer effects. In addition, only one Au species was found in this sample. The C1s transition shows that three different carbon species. The main species, with a binding energy of 284.4 eV, can be attributed to graphitic carbon, 3 which probably formed during the reaction. Two species with higher binding energies were found, one at 286.9 eV and one at 291.3 eV. The compound at 286.9 eV is most likely a carbon bound to an electronegative partner such as oxygen or sulfur. 4,5 However, since no sulfur was present on the surface, a deposited carbo-oxide species is most likely. This species has a peak area of around 35 % of the main component. The species with the highest binding energy at 291.3 eV indicated a highly oxidized carbon species with a very electron deficient carbon. This species cannot be further defined without more thorough investigation. However, since this species is difficult to detect, having only about 3% of the peak area of graphitic carbon, we did not pursue it further.
For the AuCo/CeO2 Acetonitrile, the Au4f spectra showed the presence of two different Au species. While the major species with a Au4f 7/2 binding energy of 85.4 eV was not observed by us, the minor species has a very similar Au4f 7/2 binding energy to the undoped reference sample (83.1 eV vs. 83.0 eV). The C1s spectra looks similar to the Au25/CeO2 sample. The two species with the lower binding energies have their maxima at 284.4 eV and 286.6 eV only divagating from the undoped sample by 0.3 eV for the species with the higher binding energy. However, the peak ratio has shifted as the second component has a peak area around 5 % bigger than that of the graphitic carbon. Again, a species with even higher binding energy was found at 290.3 eV. However, this component is also more pronounced and has a peak area of about 45% of the graphitic carbon. It is possible that this is adsorbed CO2 on CeO2 derived from the reaction. 6 The AuCo/CeO2 DCM sample has only one Au species. The Au4f 7/2 binding energy of 84.0 eV fits to the expected value for metallic Au. 7 The C1s spectra, however, reveals the presence of carbidic carbon on the surface with a binding energy of 281.5 eV. This species was not detected in the other samples. Graphitic carbon is also present here at 284.4 eV. In addition, a species with higher binding energy (289.1 eV) is present, accounting for about 30% of the peak area of graphitic carbon. This could indicate a carbonate species. No Co could be detected for the sample as well as for the other Co-doped sample, most likely due to the low loading of the samples.
The sample extracted with AuCo/CeO2 Acetone shows the Au species found also in the Acetonitril sample with a binding energy for the Au4f7/2 of 85.4 eV. In contrast, the two minor species seen in the previous measurement could not be found in this sample. However, it should be noted that due to differential charging of the Aceton extracted sample in UHV conditions the measurement was performed under 1 mbar of N2. This led to a decrease in signal intensity and the minor species could therefore be not visible due to the worse signal to noise ratio. The C1s region is dominated by graphitic carbon at a binding energy of 284.4 eV. Another contribution to the spectra could be identified with a species at a binding energy at 289.2 eV which was also observed in other samples.