Adsorption of Au and Pd on Ruthenium-Supported Bilayer Silica

Abstract

Adsorption of Au and Pd over bilayer SiO₂/Ru has been investigated using scanning-probe microscopy, X-ray photoemission spectroscopy (XPS), and density functional theory (DFT). Low temperature (∼5 K) atomic force (AFM) and scanning tunneling microscopy (STM) measurements reveal the presence of small adsorption features after exposing the samples to small doses of either metal. In the case of Pd, we note a homogeneous distribution of adsorbates across the entire surface, which consists of both amorphous and crystalline phases. Au, however, adsorbs only over amorphous areas and domain boundaries, which possess larger pores than can be found in the ordered portions of the film. DFT calculations reveal that this difference is rooted in the pore-size-dependent barriers for diffusion of the two metals into the film, where they can then bind stably at the Ru interface. Auger parameter analysis of the Pd 3d and Au 4f core-levels from atoms binding in this manner show upward orbital-energy shifts, which, according to the results of theoretical calculations, originate from effects similar to those causing surface core-level-shifts for such metals. Further analysis of the computational results shows that such atoms donate electron density to the Ru support, which is consistent with XPS results that show band-bending effects related to decreases in the work-function of the sample after adsorbing either metal. Additional features in the XPS studies suggest that a secondary binding mechanism, mediated by cluster formation over the SiO₂ film, becomes increasingly favorable as temperature and loading increase

Exploring Pd adsorption, diffusion, permeation, and nucleation on bilayer SiO2/Ru as a function of hydroxylation and precursor environment: From UHV to catalyst preparation

“…For the most extreme case (0.05 ML), we see little change in the Pd 3d spectrum upon heating to 200 K, indicating that the Pd structures formed following exposure at 100 K remain stable throughout that temperature range. By contrast, we note downward BE shifts associated with the silica-bound particles above 300 K, which reflects mobilization of Pd in this temperature range to result in progressively larger supported clusters [13,26,27]. As this trend continues at higher temperatures, the Pd eventually ends up in a state virtually equivalent to that formed when exposing the sample to 1 ML at 100 K, such that the two components are no longer distinguishable and produce a single 3d 5/2 XPS feature centered at ~335.6 eV.…”

“…Following liquid exposure, samples were blown dry in a flow of helium before reintroduction to the UHV system following placement within and subsequent evacuation of the transfer-chamber. As an extension of previous work dedicated to the characterization of small concentrations of Pd deposited on SiO 2 /Ru at low temperature [13,25], we conducted a series of temperature-dependent XPS measurements consistent with that shown in Figure 1a for Figure 1c. Note that the curved background is attributed to a broad Ru feature present in the absence of Pd, and that this has been accounted for via comparison with uncovered SiO 2 /Ru samples.…”

“…Note that the curved background is attributed to a broad Ru feature present in the absence of Pd, and that this has been accounted for via comparison with uncovered SiO 2 /Ru samples. As more Pd is deposited over the film (Figure 1b), the likelihood of surface nucleation becomes increasingly probable, such that the relative abundance of silica-bound Pd grows steadily relative to that diffusing through the film, and this coincides with a growth in the average size of those clusters [13]. These trends are reflected in Figure 1b as a continual growth of the higher BE feature relative to its lower BE counterpart and a gradual shift of that feature to lower BE with smaller peak widths as Pd exposure increases, such that the contribution from Ru-bound Pd becomes essentially negligible and the peak location of the SiO 2 -bound Pd becomes more consistent with bulk metal at higher loadings [13].…”

“…For the latter, these phases typically coexist to varying degrees as separate domains within the bilayer films depending on the specific preparation conditions employed [9][10][11]. In the case of the vitreous domains, the presence of differently sized pores leads to the creation of membranes capable of selectively passing or blocking the diffusion of various atoms and molecules through the film depending on the presence or absence of silica rings larger than various permeation thresholds, as has been shown for various gases, like CO and D 2 [12], and metal atoms, like Pd and Au [13,14].…”

“…As shown previously, an ability to alternate between the crystalline and vitreous phases of the film could provide a means of tailoring its permeability towards various adsorbates [13], but this has unfortunately proven difficult to reliably control in practice and necessitates sample temperatures far above those practical for many potential applications before bond rearrangement within the film becomes notable (≈1200K). By contrast, chemically controlled manipulation of surface termination, such as the creation or removal of hydroxyl groups, which have been shown to act as anchor sites for various metal species diffusing over other oxide supports [15,16], may provide a better means of tailoring diffusion processes on and through the film under a wider range of conditions.…”