-
Metal catalysts are widely used in industrial catalysis due to their high activity [1−4]. Understanding their structure-performance relationship is crucial for designing effective heterogeneous metal catalysts, as it provides insight into the active site and reaction mechanism. Oxides are frequently used as supports for metal catalysts in many significant industrial and environmental applications [5]. In real heterogeneous catalytic systems, metals supported on oxides exhibit complex structures, making it challenging to determine the actual active sites, as metal catalysts undergo structural transformations during reactions. Thus, obtaining a deeper understanding of metal-oxide interactions at the atomic scale is difficult, but such interactions can be studied using model catalytic systems with fewer variables and more controllable structures. In model catalytic systems, oxides have been used as supports for metal nanostructures to explore metal-oxide interactions [6, 7]. However, most oxide single crystals have weak conductivity, which limits their applicability in surface science characterizations. An adaptable method to investigate the properties of oxide materials involves the formation of thin oxide films on metal supports [7−9]. Well-ordered oxide films, such as FeO/Pt(111) [10, 11], NiO/Ag(100) [12], and CuxO/Cu(111) [13, 14], have been extensively studied over the past decades. The surface structure and thickness of these oxide films can be adjusted at the atomic scale. In this context, metal particles deposited on thin oxide films serve as ideal model catalytic systems for investigating the structure-performance relationships of oxide-supported metal catalysts [15−17]. Coordinatively unsaturated metal sites at the edges of the nano-islands or nano-clusters and the interface between metal and oxide film are often regarded as active sites in catalytic reactions due to the strong metal-support interaction (SMSI) [18−20]. For examples, Cu/ZnO/Al2O3 [21], which possess high activity for methanol synthesis and Au/TiOx [22], facilitates CO oxidation.
In the past decades, many attempts have been made to achieve the well-ordered growth of FeOx thin films [10, 23−26]. Through repeated cycles of iron deposition and subsequent oxidation, Weiss et al. [10] prepared a well-ordered film. They found that FeO grew layer by layer on Pt(111) up to a thickness of 2.5 monolayer (ML), and that Fe3O4(111) islands began to form after the completion of the second FeO layer. For the FeO submonolayer film grown on Pt(111), many experimental methods have been proposed. The preparation methods include depositing Fe in ultrahigh vacuum (UHV) and subsequently annealing in O2 proposed by Meyer [27], depositing Fe in O2 and subsequently annealing in UHV proposed by Chen [28], and depositing Fe in O2 and annealing in an O2 afterward proposed by Luo [29], and so on.
The ultrathin FeO film grown on a Pt(111) substrate exhibits a Moiré structure with different stacking configurations, including face-centered cubic (FCC), hexagonal close-packed (HCP), and TOP domains, where different Fe−O layer distances and surface dipoles have been reported [29−32]. Additionally, the ability of the FeO film grown on Pt(111) as a supporter for growth of metal particles has been demonstrated, Freund et al. and Shaikhutdinov et al. [33−35] found that Pd deposition at 10 K resulted in random nucleation of Pd particles, without any preferential decoration of steps edges or holes, and that Au atoms occupied HCP domains, because these positions presented the largest surface dipole. Luo et al. [32] found that Cu preferentially nucleated in the FCC domains of FeO and FeO2 at 100 K, which can be correlated with the p-band center of surface oxygen. In thermal stability experiments, Meyer et al. [27] found that upon annealing to 600 K, Pd atoms migrated through the FeO(111) film and formed a Pd intercalation structure.
In recent decades, Cu-containing FeOx materials have attracted significant attention. For example, in the high-temperature (310–450 °C) water-gas shift (WGS) process, Cu is introduced as a promoter [36, 37]. Zhu et al. [38] found that during the WGS reaction, a thin FeOx overlayer migrates onto the metallic Cu particles, creating Cu/FeOx interfaces. In order to investigate the structure and reactivity of Cu/FeO system at high temperatures, a detailed analysis of its thermal stability is required. Recently, Li et al. [18] used the FeO/Cu2O/Cu(100) inverse model catalytic system to study the role of the interface. They found that the presence of FeO islands promoted the reduction of copper oxide surface. Eventually, all oxygen atoms on the Cu surface are removed during UHV annealing above 800 K, forming the FeO/Cu(100) surface. This research provides valuable insights into the behavior of Cu/FeO system at high temperatures. However, the real catalyst consists of a small amount of Cu supported on the FeO substrate.
In this work, vapor-deposited Cu on well-defined FeO thin films grown on a Pt(111) substrate was used to study several fundamental issues related to the iron oxide supported copper catalysts. The nucleation sites and oxidation state of Cu adatoms were studied using scanning tunneling microscopy (STM) and X-ray photoelectron spectroscopy (XPS). The results indicate that Cu adatoms randomly nucleate over the whole FeO surface and tend to form metallic clusters. Given that thermal stability is a critical factor in catalyst evaluation, a series of annealing experiments were carried out. The results indicate that the Cu clusters on FeO/Pt(111) have no obvious change up to 600 K. However, annealing at 800 K leads to a dramatic structure change: all Cu clusters disappear and two new islands emerge on the surface. All the evidence points to the conclusion that two types of islands correspond to the second layer of FeO and a CuPt3 alloy. These results provide fundamental insights into the iron oxide-copper interactions. The obtained Cu/FeO/Pt(111) and CuPt3/FeO/Pt(111) systems can serve as model systems for studying the catalytic performance of Cu-FeO and CuPt3-FeO interface.
-
The studies were carried out in two Omicron UHV instruments. STM and infrared reflection absorption spectroscopy (IRAS) studies were carried out in one Omicron UHV chamber (with a base pressure of 5 × 10−10 mbar) equipped with Auger electron spectroscopy (AES), low-energy electron diffraction (LEED), STM, IRAS and a quadrupole mass-spectrometer for temperature-programmed desorption (TPD). The Pt(111) crystal was mounted on the Omicron sample holder. The temperature was measured by a K-type thermocouple spot-welded to the edge of the crystal. The Pt(111) single crystal was cleaned through cycles of Argon-ion sputtering (1.5 keV, 10 mA, 20 min), followed by annealing in UHV (1100 K, 15 min). The FeO film grown on the Pt(111) substrate was prepared by depositing Fe (99.9% purity) from an e-beam evaporator (EMT3, Omicron) in O2 atmosphere (P=2 × 10−7 mbar) onto the Pt(111) substrate at room temperature (RT), followed by annealing in 2 × 10−7 mbar O2 atmosphere at 600 K for 15 min. All STM measurements were carried out at RT with a bias voltage applied to the sample. The STM images were processed using the commercial software SPIP.
XPS measurements were performed in a separate Omicron UHV system (base pressure 1 × 10−10 mbar) equipped with LEED, XPS, TPD, STM. All XPS spectra were recorded using Al Kα radiation (1486.6 eV) and the binding energy (BE) values were calibrated with Pt 4f peak at 74.0 eV from the clean Pt(111) surface.
-
FIG. 1 shows the STM image of 1 ML FeO grown on Pt(111) surface. As can be seen, the mismatch between the unit cell of FeO(111) film (3.1 Å) and Pt(111) substrate (2.77 Å) results in a Moiré pattern with a periodicity of around 26.5 Å. The Moiré cell consists of three domains, with Fe atoms alternately occupying HCP, TOP, and FCC sites on the Pt(111) surface. The three domains can be distinguished by their distinct apparent heights, as shown in FIG. 1 [29−32].
FIG. 2 displays the STM images of 0.05 ML Cu deposited onto FeO/Pt(111) at RT. As shown in FIG. 2(a), Cu adatoms tend to aggregate into clusters. To better observe the Cu deposition sites, we zoomed in on the marked region. FIG. 2(b) shows that among seven Cu dots, three are on the HCP sites, three are on the FCC sites, and one is on the TOP sites. Therefore, we conclude that Cu deposition at RT results in random nucleation on the FeO/Pt(111) surface. This spatial arrangement is similar to that of Pd deposited on FeO/Pt(111), where Pd adatoms exhibit no preferred adsorption site [33]. Differently Au adatoms occupy the center of HCP-domains forming a hexagonal lattice with Au-Au distance coherent with the size of the Moiré cell [33, 34]. Giordano et al. [33] demonstrated that charge transfer from the adatoms to the substrate is associated with the ordering of the adatoms.
At elevated temperatures, catalysts may undergo structural transformations, leading to changes in their catalytic properties. To investigate the thermal stability of Cu/FeO/Pt(111) model catalyst, 0.4 ML Cu was deposited onto FeO/Pt(111) followed by annealing at different temperatures in UHV. FIG. 3 displays STM images together with insets that illustrate the corrugation profiles along the blue line. As shown in FIG. 3(a), 0.4 ML Cu deposited on FeO/Pt(111) forms well-distributed clusters with 5.1 nm lateral size and 2.3 Å average height. After annealing at 500 K, the average size of Cu is 5.0 nm; annealing at 600 K, the average size of Cu is 3.9 nm; annealing at 700 K, the average size of Cu is 2.8 nm, together with some 2D islands of about 10 Å lateral size. So, the annealing did not cause sintering of the Cu particles, instead, it caused penetrating through FeO. Upon further increasing the annealing temperature at 800 K, two types of islands of different heights form. As seen in the areas marked by black dotted lines in FIG. 3(e), these islands exhibit a hexagonal shape, which is charateristic of the FeO film epitaxial growth on the Pt(111) surface, and their Moiré structure has the same periodicity and orientation as the underlying FeO film (FIG. 3(f)). As shown in inset, the height of these islands with respect to the monolayer FeO film is about 2.3 Å which is similar to the second layer of FeO islands observed on the first layer of FeO film (FIG. S1 in Supplementary materials, SM). Therefore, these islands can be considered as monolayer FeO islands located above the monolayer FeO film. In addition to the second layer FeO islands, there are some irregularly shaped islands (marked by red dotted lines) with a height of 1.4–1.5 Å with respect to the FeO film. The high-resolution STM image (FIG. 3(f)) shows that these islands with irregular shape form “worm-like” surface structures. However, there is insufficient information to determine the composition of these irregular islands.
In order to study the electronic properties of Cu in this Cu/FeO/Pt(111) system after different annealing processes, XPS spectra of Cu 2p and Cu LMM Auger region were collected. As seen in FIG. 4(a), no characteristic features of Cu2+, specially, the shakeup satellites between the Cu 2p1/2 and Cu 2p3/2 doublets [39], were observed after Cu deposition and the subsequent stepwise annealing. The Cu 2p3/2 peak at 932.6 eV can be assigned to either Cu0 or Cu+. To further identify the oxidation state of Cu, we examined the Cu LMM Auger spectra. As seen in FIG. 4(b), peaks were observed at 918.5 eV, with no peak at 916.7 eV, confirming that Cu did not undergo oxidation during the entire process [39].
FIG. 5 shows the XPS spectra of Cu 2p3/2 after stepwise annealing using different electron take-off angles. The take-off angle θ is the angle between the trajectory of a electron as it leaves the surface and the surface plane. A small take-off angle provides more surface information while a larger angle provides more bulk information. The schematic illustration of surface sensitivity and photoelectric process in AR-XPS can be seen in Scheme S1 (SM). All spectra indicate that the peak of as-deposited Cu is located at 932.7 eV indicating that the Cu adatoms are in the Cu0 state. After annealing at elevated temperatures, the Cu 2p3/2 peak becomes less intense and shifts to a lower BE. After annealing at 800 K, the peak shifts to 932.1 eV, which can be assigned to Cu in a CuxPt alloy [40]. The peak areas of Cu 2p are plotted in FIG. 5(d). The data show that the peak intensity remains almost unchanged from RT to 600 K but decreases sharply after annealing at 700 K and 800 K. Therefore, we conclude that Cu atoms diffuse into the Pt(111) substrate, forming a CuxPt alloy during the annealing process. We also considered the possibility that Cu diffusion might alter the chemical environment of Fe. However, as seen in FIG. S2 (SM), we did not observe any significant changes in the Fe 2p during Cu diffusion, which excludes the possibility of forming an alloy between Cu and Fe.
Based on the STM images and XPS results, two schematic structures were proposed (Scheme S2 in SM). The first model was raised with referring to a similar structure reported for Pd/FeO/Pt(111) system, where some Pd atoms did not fully diffuse through FeO film, resulting in a metastable structure [27]. In this model, the “worm-like” structure is assumed to consist of a mixture of Cu atoms and the FeO film, while the ordered islands correspond to Cu located beneath the FeO film. Another possible structure would be that the irregular islands are Cu-Pt alloys, with the FeO film partially squeezed away, forming a second layer of FeO.
In order to determine the surface structure, CO was chosen as the probe molecule to study the properties of adsorption sites. A series of IRAS spectra of different surfaces exposed to a saturated amount of CO at RT are shown in FIG. 6. For the 1 ML FeO grown on Pt(111), no CO adsorption features are observed, indicating that the FeO film covers the whole Pt(111) surface. The lack of CO adsorption on the Cu deposited FeO/Pt(111) surface is expected, as CO desorbs from copper below RT [41, 42]. For the sample annealing at 800 K, the IRAS spectrum shows a strong peak at 2080 cm−1 and a weak peak at 1853 cm−1, which are different from the saturated CO adsorption peaks on Pt(111) surface (2090 cm−1 and 1845 cm−1). Therefore, the “worm-like” surface observed by STM cannot be attributed to Cu atoms mixed in the FeO film or to a pure Pt surface. Becker et al. [43] reported CO adsorption on CuxPt alloy surfaces, as shown in Table I. The CO adsorption peaks on CuPt3(111) are in accordance with our IRAS results for CO adsorption on the surface exhibiting the “worm-like” structure. This leads to the conclusion that the “worm-like” structure is a CuPt3 alloy. The STM image of Cu deposited on Pt(111) at 300 K followed by annealing in UHV at 800 K shows some “worm-like” structures (FIG. S3 in SM), which further supports our conclusion.
The results presented above provide two model systems for studying the catalytic properties of the interface between metals and iron oxides. When Cu is deposited on FeO/Pt(111), copper clusters are formed and randomly distributed across the FeO film, creating a Cu-FeO interface that can be used to study its role in catalytic reactions, such as the WGS reaction. However, this model catalyst remains stable only up to 700 K, beyond which structural transformations degrade the Cu-FeO interface. It is worth noting that annealing Cu/FeO/Pt(111) at 800 K produces a new structure that features a CuPt3-FeO interface. This structure offers a new model system to study a catalyst composed of CuPtx alloy and iron oxide, with potential applications in CO oxidation, methanol oxidation [44], and other reactions.
-
Our study on the growth of Cu on monolayer FeO on Pt(111) shows that the Cu adatoms tend to form clusters and randomly occupy HCP, FCC, and TOP sites on the FeO film, resulting in metallic Cu nano-clusters on the FeO/Pt(111) surface. To study the thermal stability of this Cu/FeO/Pt(111) system, stepwise annealing in UHV was carried out. Only minor changes were observed after annealing at 500 K and 600 K, suggesting that this model system can be used to study the catalytic properties in the temperature range from RT to 600 K. Annealing at 700 K causes a significant reduction in the amount of Cu adatoms and surface morphology changes. After annealing at 800 K, Cu diffuses into the Pt substrate, forming a Cu-Pt alloy that is partially exposed on the surface as “worm-like” islands. Meanwhile, some regions of the FeO film are displaced, forming FeO islands on top of the monolayer FeO film, resulting in a CuPt3/ FeO/Pt(111) structure.
This study gives insight into the nature and extent of the interactions between Cu clusters and the FeO film. The obtained Cu/FeO/Pt(111) and CuPt3/FeO/ Pt(111) structure may serve as model catalysts for exploring synergistic effect of iron oxide-metal (alloy) interface.
Supplementary materials: A STM image of 1.2 ML FeO grown on Pt(111), schematic illustration of surface sensitivity and photoelectric process in angle resolved photoelectron spectroscopy (AR-XPS), XPS spectra for Fe 2p before Cu deposition and after stepwise annealing, schematic diagram of Cu/FeO/Pt(111) system after annealing at 800 K, a STM image of Cu-Pt alloy.
Thermally Induced Transformation from Cu/FeO/Pt(111) to CuPt3/FeO/Pt(111)
- Received Date: 11/12/2024
- Available Online: 27/04/2026
-
Key words:
- FeO film /
- Cu nanoclusters /
- Iron oxide-metal interface /
- Scanning tunneling microscopy /
- X-ray photoelectron spectroscopy
Abstract: Understanding the precise structure is the initial step of exploring synergistic effect between transition metal and oxide support in heterogeneous catalysis. In this work, the growth and thermal stability of Cu on FeO/Pt(111) were systematically investigated using various surface science techniques, including scanning tunneling microscopy (STM), angle-resolved X-ray photoelectron spectroscopy (AR-XPS) and infrared reflection absorption spectroscopy (IRAS). Room temperature deposition results in randomly distributed Cu clusters on the FeO surface, which undergoes negligible changes at low annealing temperatures (below 600 K). After annealing at 800 K, the Cu/FeO/Pt(111) system transforms to a new structure composed of two types of islands with different heights and morphologies. The irregular islands are identified as CuPt3 alloys and the islands with a Moiré structure are the second layer of FeO, thus forming a CuPt3/FeO interface. Our work provides two model systems, Cu/FeO/Pt(111) and CuPt3/FeO/Pt(111), which can be used to study the catalytic properties of the Cu-FeO or CuPt3-FeO interface using surface science approaches.
首页
登录
注册


DownLoad: