Surface electronic structure and morphology of silver on iron oxide films

Surface Science 607 (2013) 124–129

Contents lists available at SciVerse ScienceDirect

Surface Science journal homepage: www.elsevier.com/locate/susc

Surface electronic structure and morphology of silver on iron oxide films Shuai Wang, Shuming Liu, Jiandong Guo, Qinlin Guo ⁎ Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, PO Box 603, Beijing 100190, China

a r t i c l e

i n f o

Article history: Received 25 June 2012 Accepted 22 August 2012 Available online 8 September 2012 Keywords: Iron oxide films Silver growth Surface plasmon Work function Surface electronic structure

a b s t r a c t The surface and electronic structure of silver grown on well defined FeO(111), Fe3O4(111) and Fe2O3(0001) films have been in situ studied using various surface analytical techniques. For Ag grown on FeO(111) or Fe2O3(0001) surfaces, the silver Mie plasmon resonance observed at about 3.8 eV at initial coverage indicates three dimensional growth mode of silver. Contrarily, no Mie plasmon peak detected for Ag on Fe3O4(111) at initial coverage of Ag suggests a wetting behavior. The main mechanism of Ag growth mode on iron oxide films and the interfacial interaction are discussed. The interfacial information is useful for noble metal–iron oxides system in chemical reactions, e.g. in photocatalysis and biology. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Oxide supported noble metal systems have been studied intensively due to their extreme importance in catalysis [1–5]. As the earth-abundant and environmentally safe material, iron oxides with unique magnetic and electronic properties are important oxide supports, and have been attracting considerable attention [6–12]. The high Curie temperature (858 K) and the predicted half metallic behavior of magnetite (Fe3O4) make it useful in spintronics and drug delivery [13]. Water splitting on the surface of hematite (Fe2O3) has been studied because Fe2O3 is corrosion-stable and inexpensive [3]. Since the interfacial behaviors including the charge transfer, interfacial diffusion play crucial roles on the properties of metal–oxide systems [4,14–16], a detailed understanding of the interface is helpful to improve the performance of real systems in chemical reaction. Thus, model surfaces of single crystals or well-ordered thin films of iron oxides grown on metal substrates including FeO(111) [6,17,18], Fe2O3(0001) [6,18–21],  Fe2O3(01 12) [22,23], Fe3O4(111) [6,17–20,24–28] and Fe3O4(100) [7,29–32] have been studied. It is known that some surfaces are of polarity [6,33,34]. The polar surface may not be stable and the stabilization of the polar surfaces may lead to different surface structures, which may bring some novel properties. For example, the Jahn–Teller stabilization of polar Fe3O4(001) results in irreversible water dissociation and the formed surface hydroxyls remain stable up to 500 K [7]. Polarity compensated (1 × 1) surface of 4 nm thick FeO(111) on Pt(111) was prepared with careful optimization of oxidation conditions [17]. The stabilization of FeO(111) may be due to the interlayer relaxation of FeO(111) films [35]. It was reported that the polarity of Fe2O3(0001) films are ⁎ Corresponding author. Tel.: +86 10 8264 9435; fax: +86 10 8264 9228. E-mail address: [email protected] (Q. Guo). 0039-6028/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.susc.2012.08.027

stabilized by strong modification of electronic structure near the surface [6]. As for the Fe3O4(111) films, strong interlayer relaxation in the surface region makes the polarity compensated [25,26]. Various properties of noble metal particles supported on welldefined iron oxide surfaces were also studied [36–46]. Encapsulation of Pt nanoparticles on Fe3O4(111) was observed, which is in relation with high adhesion energy between Pt and iron oxide supports [38,40]. The growth modes of Au, Ag and Pt on Fe3O4(001) and Fe3O4(111) were discussed based on the surface energy of deposited metals [43,44]. In addition, density functional theory (DFT) calculation shows that the synergistic effects at the interface between the noble metal bilayer and Fe2O3(0001) support can enhance catalytic activity [46]. Among the noble metals, silver is attractive for industrial applications due to its low cost and easy preparation. Besides the promising application in plasmonic photocatalyts, silver is useful in biological areas because of its good antibiotic properties. For instance, Ag–iron oxides' core/shell nanocomposites have great potential to be applied in targeting therapy area [10–12]. The antimicrobial properties of silver were found to be amplified in partly oxidized silver particles [47]. For silver nanoparticles on iron oxides, however, few model studies have been done and the chemical information, surface electron collective excitation and interfacial interaction are less known. In the present work, we prepared thin iron oxide films of FeO(111), Fe3O4(111) and Fe2O3(0001) on Mo(110) substrate. The growth and electronic structures of silver on these films with different morphologies are investigated by using various surface analytical techniques including low energy electron diffraction (LEED), Auger electron spectroscopy (AES), ultraviolet photoelectron spectroscopy (UPS), and electron energy loss spectroscopy (EELS). The experimental results show that, based on the surface plasmon, valence band structure, silver grows on FeO(111) and Fe2O3(0001) as three dimensional (3D)

S. Wang et al. / Surface Science 607 (2013) 124–129

125

clusters, but forms wetting layer on Fe3O4(111). The reasons for the different growth behaviors are discussed.

AES of iron oxides

2. Experiment

3. Results and discussion 3.1. Preparation of FeO(111), Fe3O4(111) and Fe2O3(0001) films We have systematically characterized as prepared thin films of FeO(111), Fe3O4(111) and Fe2O3(0001) by using various surface analytical techniques [18]. Fig. 1 shows the AES results. There are only the Auger peaks from Fe MNN, Fe LMM and O KLL in Fig. 1, indicating clean iron oxide surfaces. The Fe/O intensity ratio of these three films agrees well with those of FeO(111), Fe3O4(111) and Fe2O3(0001) as reported in Ref. [6]. The LEED patterns of FeO, Fe3O4 and Fe2O3 films are shown in Fig. 1 as insets. The unit cells indicated in insets all show hexagonal symmetry and correspond to (1× 1) surface unit cells of FeO(111), Fe3O4(111) and Fe2O3(0001) with surface lattice constant of 3.04, 5.94 and 5.03 Å, respectively [6]. The sharp LEED spots from FeO(111) and Fe3O4(111) films are indicative of flatter surfaces. Comparatively, for Fe2O3(0001), the relatively broad spots suggest existence of surface defects.

dN(E)/dE

Fe2O3(0001)

Fe3O4(111)

FeO(111)

O(KLL)

Fe(MNN) 0

200

400

Fe(LMM) 600

800

1000

Kinetic Energy (eV) Fig. 1. AES spectra of FeO(111), Fe3O4(111) and Fe2O3(0001) films. Insets show the corresponding LEED patterns with EP = 61 eV.

Fig. 2 gives their UPS spectra. In Fe oxides, the valence band consists of overlapping and hybridizing of O 2p and Fe 3d states. The O 2p state is mainly responsible for the most intense emission around binding energy (BE) of 4–7 eV. The characteristic peak at about 1.0 eV is attributed to Fe 3d 6L final state inferring Fe 2+ cations, which confirms formation of FeO films [48–50]. For Fe3O4, besides the

He I UPS of iron oxides

5

Fe 3d L Intensity (arb. units)

The experiments were performed in an ultrahigh vacuum system: ELS-22 (LEYBOLD-HERAUS GMBH) with base pressure of 2 × 10 −10 mbar equipped with LEED, AES, high resolution EELS and UPS. Single crystal of Mo(110) was used as substrate to grow iron oxide films. A C-type thermocouple was spot welded to the substrate for accurate temperature measurements. The clean Mo (110) surface was obtained by annealing at ~ 1100 K in ~ 10 −7 mbar O2 to remove the surface contamination (mainly carbon), followed by a subsequent flash to 1700 K without O2. No impurities were detected on the substrate by AES except a small residual oxygen signal, which was proved to have no influence on growth of the iron oxide films. Iron source was made of high purity Fe (99.994%) wire wrapped tightly around tungsten wire. The source was thoroughly degassed by thermal treatment before deposition. The deposition rates of Fe was about 0.5–1.0 monolayer per minute (ML/min) monitored by a quartz crystal oscillator. The FeO(111) films were prepared by evaporating Fe in ~10−7 mbar O2 at ~600 K, followed by annealing at 700–800 K without O2, while Fe3O4(111) films were obtained by evaporating Fe in ~10 −6 mbar O2 at ~600 K, followed by annealing at 700–800 K without O2 [18,24]. The Fe2O3(0001) films were obtained by oxidizing the FeO(111) films in ~10−5 mbar O2 at ~600 K for 2 h. The thicknesses of these iron oxide films are estimated to be around 7–15 nm considering the crystal structures and deposition rate of Fe. Silver was deposited on these iron oxide films at room temperature (RT). The source of silver was made of high purity silver wire (99.995%) either wrapped tightly around tungsten wire or loaded in a fully outgassed alumina crucible. The deposition rate of Ag was monitored by quartz crystal oscillator. The monolayer equivalent (MLE) was used to scale the coverage of deposited Ag with one monolayer packing density being 1.4 × 10 15 atoms/cm2 for Ag(111). In the UPS measurements, the He I (hv = 21.2 eV) source and the pass energy of 4 eV were used. The analyzer takeoff angle is about 15° with respect to the surface normal. The Fermi level (EF) was calibrated using either a clean Mo(110) crystal metal or thick Ni films. Work function of these iron oxide films was monitored by measuring the low energy onset of secondary electron distribution in the UPS spectrum with the sample biased at − 5 V. In EELS measurements, primary electron beam energy of 20 eV was used for investigations of surface electronic structures and plasmon excitations. The typical resolution is 60–70 meV from the full width at half maximum of the elastic peak. All data were collected at RT.

6

Fe 3d L

Fe2O3(0001)

Fe3O4(111)

FeO(111)

0

2

4

6

8

10

Binding Energy (eV) Fig. 2. UPS spectra of FeO(111), Fe3O4(111) and Fe2O3(0001) films. The dash lines at 1.0 and 2.5 eV correspond to Fe 3d6L and Fe 3d5L final states, respectively.

126

S. Wang et al. / Surface Science 607 (2013) 124–129

peak within 1 eV, another broad peak around 2.5 eV appears, which is due to Fe 3d 5L final state from Fe 3+ cations [48–50]. For Fe2O3, the emission within 1 eV of Fe 2+ cations is absent and the peak around 2.5 eV due to Fe 3+ cations is intense [48–50]. 3.2. Silver deposition on FeO(111), Fe3O4(111) and Fe2O3(0001) films The Ag was deposited on FeO(111) films at RT. The EELS spectra as a function of Ag coverage are shown in Fig. 3. For the clean FeO(111) films, a main loss peak at about 0.6 eV is attributed to a transition from Fe 3d to Fe 4s in Fe 2 + [51]. After depositing 0.16 MLE of Ag, a loss feature at about 3.83 eV is observed. As Ag coverage increases to 1.7 MLE, the peak at 0.6 eV from FeO attenuates and the peak at 3.83 eV shifts to 3.9 eV. For thick Ag with 20 MLE coverage, the peak position is also 3.9 eV. The appearance of this loss peak after Ag deposition is caused by the collective plasmon modes of silver [52–64]. For ultra-small metal particles, only Mie resonance plasmon exists as the collective plasmon mode and shows no dispersion. If the supported metal clusters are not spherical, the Mie resonance splits into two modes corresponding to an electron oscillation perpendicular (1,0 mode) and parallel (1,1 mode) to the substrate [52]. In EELS, only the (1,0) mode can be detected because the (1,1) mode electron oscillation in the surface plane will be quenched by its image dipole. For flat silver films or quite large silver clusters, the term ‘surface plasmon’ is defined as the collective plasmon mode because it shows dispersion relations, which is different from Mie resonance. It was reported that the two dimensional (2D) Ag monolayer will not give the surface plasmon [65]. Theoretical result also indicates that the surface plasmon resonance at about 3.8 eV does not exist for Ag monolayer on Al metal [66]. In our experiment, the detected collective plasmon modes mean that at initial deposition of Ag, about 0.16 MLE, the Ag does not form monolayer but grows as 3D clusters on FeO(111). We conclude that the peak at 3.83 eV is due to (1,0) Mie resonance of silver clusters, which is consistent with reported results [58]. For higher coverage of Ag, the LEED pattern of Ag(111) (1 × 1) structure is observed as shown in the inset of Fig. 3, indicating the

formation of either bigger silver particles or ordered silver films. The surface plasmon position depends on the parallel momentum, which is determined by primary beam energy, the incident and scattering angles. Since we used the same primary energy as well as the incident and scattering angles, the observed plasmon energy shift from 3.83 to 3.90 eV as a function of Ag coverage must be caused by other factors such as cluster size and the valence electron density [56–59,64–66]. Fig. 4 shows the valence band spectra by UPS with increasing Ag coverage on FeO(111) films. With Ag coverage increasing from 0.16 to 0.7 MLE, the peaks at 4.9 eV gradually increase in intensity. These peaks are attributed to Ag 4d band state of metallic Ag [67,68]. The emission from the FeO(111) substrate, however, is gradually suppressed due to the damping effect of the deposited Ag. At the coverage > 0.7 MLE, the 4d band splitting is observed, being attributed to crystal field effects of multilayer Ag films [69]. The valence band of thick Ag films (20 MLE) is consistent with the known UPS results of Ag(111) single crystal, in which low density of states (DOS) at EF and the peaks at 4–7 eV due to spin-orbit split d-band complex are observed [70–73]. At 0.16 MLE coverage, the emission of Ag 4d shows a shift of 0.3 eV to higher BE. The BE shift for metal clusters on oxide films can be attributed to the initial state and final state effects (particle size effects). However, since observed Mie resonance plasmon suggests that 3D Ag clusters are of metallic properties, the initial state effects' contribution is rather small. Thus, the shift of 0.3 eV is mainly due to the size effects at initial coverage. The inset of Fig. 4 shows the enlarged region around Fermi level. The DOS at EF is rather low for FeO(111), which facilitates distinguishing the DOS of Ag from the substrate. For 20 MLE Ag films, finite DOS at EF can be observed clearly, which is consistent with the valence band of bulk Ag. For Ag coverageb 0.7 MLE, the DOS near EF is very low. This result agrees with that of the band shift to higher BE, which again indicates particle size effects. For Ag coverage higher than 0.7 MLE, finite DOS near EF starts to appear, which indicates properties of bulk Ag. This means that Ag clusters grow larger and the particle size effects can be neglected.

He I UPS of Ag on FeO(111)

EELS of Ag on FeO(111) EP = 20 eV

20 1.7

Surface Plasmon 0.7

Ag/ MLE 20 ×64

1.7

Intensity (arb. units)

Loss Intensity (arb. units)

0.35 0.16

Ag/ MLE

FeO(111)

-0.5

0.0

0.5

1.0

20

B E (eV)

1.7

0.7

0.7 0.35

0.35

0.16

0.16

FeO(111) FeO(111) 0

1

2

3

4

5

6

7

Loss Energy (eV) Fig. 3. EELS spectra of Ag on FeO(111) as a function of Ag coverage. The inset shows the LEED pattern with 20 MLE Ag on FeO(111).

0

2

4

6

8

10

Binding Energy (eV) Fig. 4. UPS spectra of Ag on FeO(111) as a function of Ag coverage. The inset gives the enlarged region of DOS near Fermi level.

S. Wang et al. / Surface Science 607 (2013) 124–129

EELS of Ag on Fe3O4(111) EP = 20 eV

Surface Plasmon

Ag/ MLE

×64

Loss Intensity (arb. units)

20

5 ×64 ×8

1.7

1

0.7 0.35 0.16 Fe3 O 4(111) 0

1

2

3

4

5

6

7

Loss Energy (eV) Fig. 5. EELS spectra of Ag on Fe3O4(111) as a function of Ag coverage. The inset shows the LEED pattern with 1.7 MLE Ag on Fe3O4(111).

He I UPS of Ag on Fe3O4(111)

20

5 1.7 0.7

Ag/ MLE

0.35

Intensity (arb. units)

It is well known that, for metal clusters deposited on oxide support, size effects are important in the electronic structure as well as catalytic properties [64,74]. For Ag, intrinsic size effects consist of two major competing processes including spillout of the Ag 5s electrons and hybridization of Ag 5s and 4d electrons. These two processes have opposite effects on surface plasmon frequency shift. With decreasing size to 3–7 nm, the former one causes about 0.4 eV red shift in surface plasmon position [57,63], while the latter one causes blue shift of about 0.3 eV [58]. For Ag on FeO(111), the peak position of Ag plasmon at initial coverage shows red shift with energy less than 0.1 eV. This small shift points out that both of these two processes affect the Mie resonance plasmon of Ag. Ag was also deposited on Fe3O4(111) films. Fig. 5 gives the plasmon change as a function of Ag coverage. For clean Fe3O4(111) films, a broader peak around 4.7 eV can be attributed to transitions from Fe 3d to Fe 4s as well as from O 2p to Fe 3d in Fe 2+, respectively [22,23,51]. For Ag coverage less than 0.7 MLE, no Mie plasmon resonance loss peaks were observed, which indicates formation of 2D layer/cluster of silver [65,66]. The LEED pattern of 1.7 MLE Ag on Fe3O4(111) films is shown as inset in Fig. 5. The bigger and brighter spots are from Ag(111), and the intensity of the spots from Fe3O4(111) becomes very weak due to growth of Ag. This result suggests that initially deposited Ag prefers to wet Fe3O4(111) surface, i.e. the Ag grows on Fe3O4(111) as layer by layer. For coverage at 1 MLE, the plasmon peak starts to appear at about 3.87 eV, and its signal becomes stronger with the coverage increase. The valence band structure of Ag on Fe3O4(111) films is shown in Fig. 6. At initial coverage, b 0.7 MLE, no distinct peaks can be observed,

127

0.16 Fe3O4(111)

-0.5

0.0

0.5

20

1.0

B E (eV)

5 1.7 0.7 0.35 0.16 Fe3 O 4(111)

0

2

4

6

8

10

Binding Energy (eV) Fig. 6. UPS spectra of Ag on Fe3O4(111) as a function of Ag coverage. The inset gives the enlarged region of DOS near Fermi level.

which is different from that of Ag on FeO(111). For 1.7 MLE coverage, the emission in region 4 to 6 eV becomes stronger due to Ag 4d band, and a shoulder at 6 eV clearly appears being similar to that of Ag on FeO(111) (see Fig. 4). However, the emission in region 4 to 6 eV is broader and no distinct Ag 4d peak is observed at 4.8 eV compared with that of Ag on FeO(111). These quite different electronic structures are indicative of an interaction between 2D silver and Fe3O4(111) films. One may argue that the silver might be oxidized at the interface. It is reported that for the oxide state of silver, an emission peak at 3.2 eV below EF appears due to the Ag 5s, 5p-O 2p hybridization of valence electrons [72,75]. However, in our experiments no apparent shoulder is observed at ~3 eV, illuminating that the silver is not oxidized on Fe3O4(111) films. The inset of Fig. 6 shows the enlarged region around Fermi level. The low but finite DOS at EF for Fe3O4(111) is due to the t2g electrons with minority spin, which results in the conductivity of Fe3O4(111) above Verwey transition [76]. For Ag coverage less than 1.7 MLE, the DOS near EF does not increase. This is indicative of non-metallic properties of Ag due to interaction between Ag and Fe3O4(111). For coverage higher than 1.7 MLE, the DOS near EF starts to increase. This suggests metallic properties of a part of Ag, which is not affected by Fe3O4(111). For 20 MLE coverage, finite DOS near EF can be observed clearly, which is consistent with that of bulk Ag. For comparison, Ag was also grown on Fe2O3(0001) films. Fig. 7 presents the EELS spectra with increased Ag coverage. For clean Fe2O3(0001) films, a very broad peak centered at about 4.5 eV are observed, which is due to O 2p to Fe 3d transitions. This result is in good agreement with that of the ideal Fe2O3 single crystal [22,23]. At 0.16 MLE, a very weak peak from Mie plasmon resonance of silver appears, suggesting 3D growth of silver at initial coverage. As Ag coverage increases, the surface plasmon gradually becomes intense and locates at 3.87 eV, which is typical of surface plasmon of metallic Ag. The variation of valence band with increased Ag coverage on Fe2O3(0001) films is shown in Fig. 8. At 0.16 MLE of Ag deposition, a peak at 4.8 eV originated from Ag 4d band state appears. While the silver coverage increases to 1.7 MLE, the 4d band splitting due

128

S. Wang et al. / Surface Science 607 (2013) 124–129

EELS of Ag on Fe2O3(0001) EP = 20 eV

Surface Plasmon Ag/ MLE

Loss Intensity (arb. units)

×64

20

1.7

0.7

0.35

0.16

Fe2 O3 (0001)

0

1

2

3

4

5

6

7

Loss Energy (eV) Fig. 7. EELS spectra of Ag on Fe2O3(0001) as a function of Ag coverage.

to crystal field effects is also observed. The Ag 4d peak at initial coverage shows about 0.2 eV shift relative to the bulk value. Being similar to that of Ag on FeO(111), the band shift of Ag 4d band infers particle size effects for small Ag clusters at initial coverage. The enlarged region around the Fermi level in the inset of Fig. 8 shows similar behavior as that of Ag on FeO(111).

He I UPS of Ag on Fe2O3(0001)

20

1.7 0.7

The results of Ag growth on iron oxide films presented above clearly show that at initial coverage, Ag grows as 3D clusters on FeO(111) and Fe2O3(0001), while Ag wets on Fe3O4(111). The growth mode of Ag on iron oxides depends on many factors such as surface structure, defect, surface energy and interfacial interaction. Surface symmetry and lattice mismatch between deposited metal and oxide supports strongly affect the growth mode. For FeO(111), Fe2O3(0001), Fe3O4(111) and Ag(111), their surface symmetry is all hexagonal, which favors the growth of Ag in (111) orientation. In addition, the surface lattice mismatch between Ag(111) (2.89 Å) and O–O interatomic distance of FeO(111) (3.04 Å), Fe2O3(0001) (2.90 Å) and Fe3O4(111) (2.97 Å) is 5.2%, 0.3% and 2.8%, respectively [6,76]. The relatively large lattice mismatch between Ag(111) and FeO(111) favors the growth of 3D silver clusters. Therefore, our observation of (111) LEED pattern for higher coverage of Ag on FeO(111) is due to large Ag particles. Surface defects often act as the nucleate centers resulting in the 3D growth of metal clusters [78]. Thus, the defective surface of Fe2O3(0001) films favors the 3D silver clusters, even though the lattice mismatch between Ag(111) and Fe2O3(0001) is very small. For Fe3O4(111), a better lattice match as well as less surface defects favor the layer-by-layer growth of Ag. Besides surface structure and defect, the surface energies of the silver (γAg), the face of iron oxide (γOxide) and the interface (γAg/Oxide) are important issues. Define the total energy of the silver-iron oxide systems Δγ = γAg + γAg/Oxide − γOxide. Generally, for Δγ > 0, metal atoms tend to grow in 3D mode, while Δγ b 0 benefits to form wetting layer. Theoretically, the surface energies of FeO(111) (~0.7 J/m 2) and Fe3O4(111) (~ 0.8 J/m 2) are basically the same [77,79]. Thus, the wetting behavior of Ag on Fe3O4(111) and the 3D growth of Ag on FeO(111) suggest that the interfacial energy of Ag/Fe3O4(111) is lower than that of Ag/FeO(111). However, we noticed that the interfacial energy could also depend on other factors such as surface morphology, terminations, atomic densities and the interfacial interaction. In addition, the electron chemical potential of these surfaces is a factor that affects the interfacial interaction [6,15,37]. We have compared the electron chemical potential of FeO(111), Fe3O4(111), Fe2O3(0001) and Ag(111) surfaces by measuring their work functions. The electron chemical potential difference between Fe3O4(111) and Ag(111) is 1.3 eV, while the difference between FeO(111), Fe2O3(0001) and Ag(111) is 0.4 and 0.8 eV, respectively. The relatively large difference of electron chemical potential at the interface of Ag on Fe3O4(111) may lead to an interfacial interaction and affect the growth mode of silver.

0.35 0.16

Ag/ MLE

Intensity (arb. units)

Fe2O3(0001)

-0.5

0.0

0.5

20

1.0

B E (eV)

1.7 0.7 0.35 0.16

Fe2 O 3(0001)

0

2

4

6

8

10

Binding Energy (eV) Fig. 8. UPS spectra of Ag on Fe2O3(0001) as a function of Ag coverage. The inset gives the enlarged region of DOS near Fermi level.

4. Conclusions In this work, surface electronic structure and morphology of silver grown on well defined FeO(111), Fe3O4(111) and Fe2O3(0001) surfaces have been investigated. At initial deposition of Ag, the observation of Mie plasmon resonance at 3.8 eV suggests a 3D mode growth of Ag on FeO(111) and Fe2O3(0001). While for silver on Fe3O4(111), no Mie plasmon appears at initial coverage, inferring wetting layer of Ag films and an interfacial interaction. The large lattice mismatch between Ag and FeO(111) favors the 3D growth of Ag. The richness of surface defects hampers the 2D growth of silver on Fe2O3(0001). For Ag on Fe3O4(111), the better lattice match and less surface defects all favor the 2D growth mode of Ag. The present study provides some interfacial information for the use of noble metal–iron oxide system in photocatalysis and biological areas. Acknowledgments We gratefully acknowledge the support for this work by the Knowledge Innovation Program of the Chinese Academy of Sciences (Grant No. KJCX2-EW-W09) and 973 Program of China (2012CB921700).

S. Wang et al. / Surface Science 607 (2013) 124–129

References [1] K. Watanabe, D. Menzel, N. Nilius, H.-J. Freund, Chem. Rev. 106 (2006) 4301. [2] K. Awazu, M. Fujimaki, C. Rockstuhl, J. Tominaga, H. Murakami, Y. Ohki, N. Yoshida, T. Watanabe, J. Am. Chem. Soc. 130 (2008) 1676. [3] I. Thomann, B.A. Pinaud, Z. Chen, B.M. Clemens, T.F. Jaramilo, M.L. Brongersma, Nano Lett. 11 (2011) 3440. [4] Q. Fu, H. Saltsburg, M. Flytzani-Stephanopoulos, Science 301 (2003) 935. [5] W. Deng, A.I. Frenkel, R. Si, M. Flytzani-Stephanopoulos, J. Phys. Chem. C 112 (2008) 12834. [6] W. Weiss, W. Ranke, Prog. Surf. Sci. 70 (2002) 1. [7] G.S. Parkinson, Z. Novotný, P. Jacobson, M. Schmid, U. Diebold, J. Am. Chem. Soc. 133 (2011) 12650. [8] H. Ding, C. Shen, H. Chao, Z. Xu, C. Li, T. Yuan, X. Shi, H.-J. Gao, Chin. Phys. B 19 (2010) 066102. [9] C.H. Liu, Z.D. Zhou, X. Yu, B.Q. Lv, J.F. Mao, D. Xiao, Inorg. Mater. 44 (2008) 291. [10] D. Tang, R. Yuan, Y. Chai, J. Phys. Chem. B 110 (2006) 11640. [11] Z. Wei, Z. Zhou, M. Yang, C. Liu, Z. Zhao, D. Huang, Z. Chen, J. Gao, J. Mater. Chem. 21 (2011) 16344. [12] P. Gong, H. Li, X. He, K. Wang, J. Hu, W. Tan, S. Zhang, X. Yang, Nanotechnology 18 (2007) 285604. [13] Z. Zhang, S. Satpathy, Phys. Rev. B 44 (1991) 13319. [14] M. Baumer, H.-J. Freund, Prog. Surf. Sci. 61 (1999) 127. [15] Q. Fu, T. Wagner, Surf. Sci. Rep. 62 (2007) 431. [16] M. Sterrer, T. Risse, U.M. Pozzoni, L. Giordano, M. Heyde, H.-P. Rust, G. Pacchioni, H.-J. Freund, Phys. Rev. Lett. 98 (2007) 096107. [17] N. Spiridis, D. Wilgocka-Ślęzak, K. Freindl, B. Figarska, T. Giela, E. Młyńczak, B. Strzelczyk, M. Zając, J. Korecki, Phys. Rev. B 85 (2012) 075436. [18] M. Xue, S. Wang, K. Wu, J. Guo, Q. Guo, Langmuir 27 (2011) 11. [19] X. Deng, J. Lee, C. Matranga, Surf. Sci. 604 (2010) 627. [20] C. Schlueter, M. Lübbe, A.M. Gigler, W. Moritz, Surf. Sci. 605 (2011) 1986. [21] N.G. Condon, F.M. Leibsle, A.R. Lennie, P.W. Murray, D.J. Vaughan, G. Thornton, Phys. Rev. Lett. 75 (1995) 1961. [22] M.A. Henderson, Surf. Sci. 604 (2010) 1197. [23] M.A. Henderson, Surf. Sci. 515 (2002) 253. [24] Y.Q. Cai, M. Ritter, W. Weiss, A.M. Bradshaw, Phys. Rev. B 58 (1998) 5043. [25] M. Ritter, W. Weiss, Surf. Sci. 432 (1999) 81. [26] Sh.K. Shaikhutdinov, M. Ritter, X.-G. Wang, H. Over, W. Weiss, Phys. Rev. B 60 (1999) 11062. [27] K. Zheng, Q. Guo, E.G. Wang, Surf. Sci. 602 (2008) 2600. [28] Y. Joseph, C. Kuhrs, W. Ranke, W. Weiss, Surf. Sci. 433–435 (1999) 114. [29] M. Fonin, R. Pentcheva, Yu.S. Dedkov, M. Sperlich, D.V. Vyalikh, M. Scheffler, U. Rüdiger, G. Güntherodt, Phys. Rev. B 72 (2005) 104436. [30] B. Stanka, W. Hebenstreit, U. Diebold, S.A. Chambers, Surf. Sci. 448 (2000) 49. [31] N. Spiridis, B. Handke, T. Slezak, J. Barbasz, M. Zając, J. Haber, J. Korecki, J. Phys. Chem. B 108 (2004) 14356. [32] R. Pentcheva, F. Wendler, H.L. Meyerheim, W. Moritz, N. Jedrecy, M. Scheffler, Phys. Rev. Lett. 94 (2005) 126101. [33] P.W. Tasker, J. Phys. C 12 (1979) 4977. [34] J. Goniakowski, F. Finocchi, C. Noguera, Rep. Prog. Phys. 71 (2008) 016501. [35] G. Ketteler, W. Ranke, R. Schlogl, Phys. Chem. Chem. Phys. 6 (2004) 205. [36] Sh.K. Shaikhutdinov, R. Meyer, D. Lahav, M. Bäumer, T. Klüner, H.-J. Freund, Phys. Rev. Lett. 91 (2003) 076102. [37] R. Meyer, M. Bäumer, Sh.K. Shaikhutdinov, H.-J. Freund, Surf. Sci. 546 (2003) L813. [38] L. Giordano, G. Pacchioni, J. Goniakowski, N. Nilius, E.D.L. Rienks, H.-J. Freund, Phys. Rev. Lett. 101 (2008) 026102.

129

[39] K. Jordan, S. Murphy, I.V. Shvets, Surf. Sci. 600 (2006) 5150. [40] Z.-H. Qin, M. Lewandowski, Y.-N. Sun, S. Shaikhutdinov, H.-J. Freund, J. Phys. Chem. C 112 (2008) 10209. [41] K.T. Rim, D. Eom, L. Liu, E. Stolyarova, J.M. Raitano, S.-W. Chan, M. FlytzaniStephanopoulos, G.W. Flynn, J. Phys. Chem. C 113 (2009) 10198. [42] M. Lewandowski, Y.-N. Sun, Z.-H. Qin, S. Shaikhutdinov, H.-J. Freund, Appl. Catal. A 391 (2011) 407. [43] C. Gatel, E. Snoeck, Surf. Sci. 601 (2007) 1031. [44] C. Gatel, E. Snoeck, Surf. Sci. 600 (2006) 2650. [45] N. Spiridis, R.P. Socha, B. Handke, J. Haber, M. Szczepanik, J. Korecki, Catal. Today 169 (2011) 24. [46] K. Wong, Q.H. Zeng, A.B. Yu, J. Phys. Chem. C 115 (2011) 4656. [47] P.L. Taylor, O. Omotoso, J.B. Wiskel, D. Mitlin, R.E. Burrell, Biomaterials 26 (2005) 7230. [48] A. Fujimori, N. Kimizuka, M. Taniguchi, S. Suga, Phys. Rev. B 36 (1987) 6691. [49] Robert J. Lad, Victor E. Henrich, Phys. Rev. B 39 (1989) 13478. [50] A. Fujimori, M. Saeki, N. Kimizuka, Phys. Rev. B 34 (1986) 7318. [51] Xu-Xin Zhang, J. Schoenes, W. Reim, P. Wachter, J. Phys. C: Solid State Phys. 16 (1983) 6055. [52] U. Kreibig, W. Vollmer, Optical Properties of Metal Clusters, in: Springer Series Materials Science, Vol. 25, Springer-Verlag, Berlin, 1995. [53] J. Daniels, Z. Phys. 203 (1967) 235. [54] D. Martin, F. Creuzet, J. Jupille, Y. Borensztein, P. Gadenne, Surf. Sci. 377–379 (1997) 958. [55] T. Wenzel, J. Bosbach, F. Stietz, F. Träger, Surf. Sci. 432 (1999) 257. [56] M. Rocca, Surf. Sci. Rep. 22 (1995) 1. [57] Q. Guo, D.W. Goodman, Chin. Phys. 10 (2001) 80. [58] N. Nilius, N. Ernst, H.-J. Freund, Phys. Rev. Lett. 84 (2000) 3994. [59] F. Didier, J. Jupille, Surf. Sci. 307–309 (1994) 587. [60] Y. Yu, Y. Jiang, Z. Tang, Q. Guo, J. Jia, Q. Xue, K. Wu, E. Wang, Phys. Rev. B 72 (2005) 205405. [61] M. Rocca, Y. Li, F. Buatier de Mongeot, U. Valbusa, Phys. Rev. B 52 (1995) 14947. [62] R. Contini, J.M. Layet, Solid State Commun. 64 (1987) 1179. [63] J.R. Heath, Phys. Rev. B 40 (1989) 9982. [64] K. Luo, X. Lai, C.-W. Yi, K.A. Davis, K.K. Gath, D.W. Goodman, J. Phys. Chem. B 109 (2005) 4064. [65] J.M. Layet, R. Contini, J. Derrien, H. Lüth, Surf. Sci. 168 (1986) 142. [66] Ansgar Liebsch, in: Electronic Exicitations at Metal Surfaces, Plenum Press, New York, 1997, p. 126. [67] I. Matsuda, H.W. Yeom, K. Tono, T. Ohta, Surf. Sci. 438 (1999) 231. [68] P. Stracke, S. Krischok, V. Kempter, Surf. Sci. 473 (2001) 86. [69] J.G. Tobin, S.W. Robey, D.A. Shirley, Phys. Rev. B 33 (1986) 2270. [70] A. Goldmann, D. Westphal, R. Courths, Phys. Rev. B 25 (1982) 2000. [71] G. Borstel, M. Neumann, M. Wöhlecke, Phys. Rev. B 23 (1981) 3121. [72] A.I. Boronin, S.V. Koscheev, G.M. Zhidomirov, J. Electron Spectrosc. Relat. Phenom. 96 (1998) 43. [73] A. Elbe, G. Meister, A. Goldmann, Surf. Sci. 397 (1998) 346. [74] C.T. Campbell, S.C. Parker, D.E. Starr, Science 298 (2002) 811. [75] L.S. Kibis, V.I. Avdeev, S.V. Koscheev, A.I. Boronin, Surf. Sci. 604 (2010) 1185. [76] J. Garcia, G. Subias, J. Phys. Condens. Matter 16 (2004) R145. [77] J.H. Larsen, J.T. Ranney, D.E. Starr, J.E. Musgrove, C.T. Campbell, Phys. Rev. B 63 (2001) 195410. [78] C.T. Campbell, Surf. Sci. Rep. 27 (1997) 1. [79] S.H. Overbury, P.A. Bertrand, G.A. Somorjai, Chem. Rev. 75 (1975) 547.