Prepared by UHV Chemical Vap - American Chemical Society

Mar 25, 2014 - electron diffraction (AS-XPD). The goal is putting on a firm basis the description of chemical and structural characterization of the Z...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCC

Zr2O3 Nanostripes on TiO2(110) Prepared by UHV Chemical Vapor Deposition Askia E. Reeder, S. Agnoli, G. Andrea Rizzi,* and G. Granozzi Department of Chemical Sciences, University of Padova, Via Marzolo, 1, 35131 Padova, Italy ABSTRACT: Zirconium tetra-tert-butoxide was used as precursor for the growth of ZrO2 films on TiO2(110) in ultrahigh vacuum conditions. The composition of the film was characterized by X-ray photoelectron spectroscopy (XPS), while its thickness and growth mode were characterized by angle-resolved photoelectron spectroscopy (AR-XPS). The structure of the interface was studied in detail by angle-scanned X-ray photoelectron diffraction (AS-XPD). These data, compared with multiple-scattering spherical wave (MSSW) simulations suggest the growth, at the interface, of Zr2O3 nanostripes aligned along the [001] direction. The structure and composition of the interface are responsible for the lower acidity of the ZrO2/TiO2(110) surface if compared with that of the bare substrate.



INTRODUCTION Zirconia (ZrO2) has a wide range of existing and potential technological applications, e.g., in high temperature fuel cells1 and as a high dielectric material in electronic devices.2,3 The properties that establish its use in applications include high melting point, low thermal conductivity, high-k dielectric value, and high mechanical and corrosion resistance. In the field of catalysis, zirconia is employed both as support and active component. For example, it has been investigated as a catalyst for partial oxidation, dehydration, isomerization of hydrocarbons, dehydrogenation of alkanes, and the synthesis of olefins from alcohols.4,5 It has also been studied as a component of mixed oxide acid−base bifunctional catalysts6 and methanol synthesis catalysts,7 and as a thermally stable support in the catalytic combustion of methane.8 Its most important catalytic application is in three-way automotive catalytic converters, where it is considered essential to stabilize the ceria active component.9 Lately, it has been used in the photocatalytic reduction of greenhouse gases,10 especially in conjunction with titania,11 for environmental12 and energy applications:13 it has been demonstrated14 that while interfaces between identical oxides substantially decrease the number of trapped charges, ZrO2/TiO2 heterojunctions give rise to a clear beneficial effect to the photocatalytic activity. Actually, there is an effective interfacial charge separation across nanometer-sized ZrO2/TiO2 interfaces attributed to the diminished charge carrier recombination in ZrO2. In addition, ZrO2/TiO2 mixed oxides have attracted much attention because they show both better thermal stability and modified surface acidity properties with respect to the bare ZrO2,15 and the pertinent literature is quite extensive.16−19 Despite the interest for their catalytic and photocatalytic applications, a systematic characterization of the interface composition and properties of these mixed oxide systems is still © 2014 American Chemical Society

lacking. To our best knowledge, there are only a few examples in the literature where true substitutional Ti(1‑x)ZrxO2 mixed oxides have been characterized:20,21 in the first case the authors used a laser ablation procedure, whereas in the latter the samples were prepared by an inverse microemulsion method. On the other hand, there is no study in the literature where the ZrO2/TiO2 interface has been investigated in detail. We have therefore undertaken a study based on the premises of surface science aiming at the preparation and characterization of ultrathin films of ZrO2 on the most studied titania single crystal surface, i.e., the rutile TiO2(110). The growth has been carried out by ultrahigh-vacuum (UHV) chemical vapor deposition (CVD), and the deposits have been characterized by photoemission spectroscopy (XPS) and angle-scanned photoelectron diffraction (AS-XPD). The goal is putting on a firm basis the description of chemical and structural characterization of the ZrO2/TiO2 interface. We have chosen to grow the ZrO2 film by using UHV-CVD and zirconium tetra-tert-butoxide (ZTB) as precursor. By means of CVD, it is possible to achieve atomic layer control of the film thickness, uniform thickness on large areas, stoichiometric control, and excellent conformal step coverage of nonplanar samples. Bradley22 first used ZTB as a CVD precursor because of its volatility and facile decomposition, and proposed its use in the electronic and ceramic industry as precursor for the deposition of ZrO2. It is worth pointing out that the use of an alkoxide precursor for the deposition of an oxide thin film is a very convenient choice since these compounds are highly volatile, readily decompose at moderate high temperatures, and contain more than enough oxygen to satisfy the metal’s requirement in forming the oxide. Received: January 22, 2014 Revised: March 14, 2014 Published: March 25, 2014 8026

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

data.29 The thickness has been defined assuming the monolayer (ML) as the thickness of tetragonal ZrO2 half-cell length along [001] (2.63 Å). XPS experiments were performed in a multitechnique UHV chamber equipped with a 5 degree of freedom (x, y, z, polar angle θ, and azimuthal angle φ) manipulator and the possibility to heat the sample up to 1000 K, a four grid rear-LEED, and a double anode OMICRON (Mg/Al) X-ray source interfaced with a 5 channeltrons VG ESCALAB electron hemispherical analyzer. All the XPS measurements were performed at RT using nonmonochromatized Al Kα radiation (hν = 1486.6 eV) and a pass energy of 50 eV, for the survey, and 20 eV for the high-resolution spectra. In an AS-XPD experiment,30 the photoelectron intensity I is monitored as a function of the polar (θ) and azimuthal (φ) angles. We consider the anisotropy function χ(θ, φ) = I(θ, φ)/(I0(θ) − 1), where I0(θ) is the average evaluated from I(θ,φ) over φ in the 0−360° range and the plot of χ versus (θ, φ) is called anisotropy 2π stereographic pattern. The center of the 2π plots corresponds to the surface normal, the radial section displays a polar scan, the circular section an azimuthal scan, and the anisotropy is given by the corresponding color scale. The presence of anisotropy is an indication of an ordered arrangement around the emitter. Multiple-scattering spherical wave (MSSW) simulations of the AS-XPD patterns were obtained by the EDAC software.31

Moreover, the organic groups can be easily removed in the form of volatile nonreactive products of reaction. Since there are no metal−carbon bonds a low amount of carbon contamination and absence of carbides is expected. A first application of this precursor in the formation of a thin film was done by Cameron at al.23 by using Si(100) as substrate because of the importance of ZrO2 in high-k gate dielectric for MOSFETs.24−26 Also the decomposition reaction of ZTB was thoroughly studied by using IR spectroscopy: a reaction mechanism was proposed showing that ZrO2 is formed on the surface via hydrogen β-elimination from the tert-butoxide ligand, forming isobutylene (or isobutene) as main byproduct. The interfacial composition with Si, when using ZTB as precursor, was studied in detail by P. G. Karlsson et al.27,28 Similar studies of ZrO2/TiO2 interfaces, nevertheless, are still lacking.



EXPERIMENTAL SECTION An epipolished rutile TiO2 crystal (MaTecK) was mounted on a rotating sample holder with two Mo clips holding the substrate and a K-thermocouple in place. The sample was cleaned via sputtering treatments with 2.0 kV Ar+ flux (P = 5 × 10−7 mbar) at an angle of 45° with respect to the surface normal for 20 min. This was followed by annealing in UHV for 5 min at 923 K to form a stoichiometric surface and wide (110) terraces, while the introduction of O bulk vacancies assured good sample conductivity. Various cleaning cycles were undertaken until the surface was free from contamination and a well-defined (1 × 1) low energy electron diffraction (LEED) pattern was visible. The deposition of ZrO2 was carried out via UHV-CVD as an alternative and easier way with respect to evaporation of metallic Zr in oxygen background. The zirconium tetra-tertbutoxide precursor (ZTB, Sigma-Aldrich) is a yellow liquid that easily evaporates at room temperature (RT). Moreover, the precursor readily decomposes to ZrO2 when adsorbed on a heated substrate, so the deposition can be carried out without the assistance of oxygen. Since ZTB is extremely hygroscopic, the precursor was inserted in a UHV vial in a drybox, in order to avoid ambient moisture. After this, the UHV vial was connected via a gas line to the preparation chamber. The vial and precursor were freeze−pump−thawed for several cycles in order to eliminate contaminating gases. ZTB was introduced into the chamber via a leak valve connected to a bellow and a nozzle, 3 cm away from the substrate surface. During the deposition a quadrupole mass spectrometer (QMS, PFEIFFER Prisma 80) was used to monitor isobutylene, the main product resulting from the decomposition on the surface of ZTB, as already verified by Cameron et al.23 QMS was used together with a pressure gauge to correctly read the pressure during depositions. The experimental procedure to effectively deposit ZrO2/ TiO 2(110) films required a long optimization of the experimental parameters. It was shown by other authors27,28 that the ZTB decomposition process might form, at different temperatures, various carbonaceous residues, especially at low temperature, when the decomposition rate is particularly slow. Our deposition was carried out at three different temperatures 680, 740, 770 K and in five stages of a minute each at a pressure of 1.3 × 10−7 mbar. The results were traced by using XPS. In order to allow an estimate of the thickness and the growth rate of the grown ZrO2 films, the “patched overlayer” growth model has been assumed to analyze angle-resolved XPS (AR-XPS)



RESULTS AND DISCUSSION ZrO2/TiO2(110) Film Deposition and Chemical Characterization. In this section we will analyze the XPS spectra obtained from the ZrO2 films of increasing thickness (1.3−6.3 ML) grown with a substrate temperature of 680, 740, and 770 K, respectively. For space saving, in Figure 1 we report only the XPS data acquired at the lowest temperature. The Zr 3d peaks are reported in Figure 1a: the Zr 3d5/2 peak is centered at a binding energy (BE) of 182.9 eV, while for thicker deposits it shifts up to 183.3 eV. A similar behavior was observed in a previous paper32 where the Zr 3d5/2 was centered at 183 eV for the thicker films and showed ca. 0.3 eV shift to lower BE for thinner films. The Ti 2p peak (Figure 1b) did not show any significant shift, thus excluding band bending. In all cases the Ti 2p signal is characterized by the presence of a weak shoulder at 457.8 eV that highlights the presence of Ti3+ (see the inset in Figure 1b). Although not purveyed in this paper, depositions undertaken at higher temperatures show progressively more reduction of the Ti4+ centers. This evidence gives support to the hypothesis that oxygen diffusion from the titania support is activated at the higher temperatures. We also observe (Figure 1c) a shift of 0.3 eV in the O 1s peak, from 530.4 to 530.7 eV with increasing thickness. The BE values relative to the O 1s peaks in the case of the pure oxides are quite close: 530 eV for TiO2 and 530.5 for ZrO2. Thus, we assume the shift is due to the increased amount of deposited ZrO2, although this observation offers no help in discerning whether the interface is a mixed oxide. The O 1s peak shows also a weak shoulder at ca. 532 eV. At higher thickness this shoulder becomes progressively smaller, but never disappears. This indicates the presence of oxygen atoms not in proper bulk configuration,33 or as pointed out by Karlsson et al.,28 this shoulder is most probably due to hydroxyl groups formed upon precursor decomposition. The C 1s (Figure 2) spectral line-shape is rich with information about the ZTB decomposition on the TiO2 surface. By looking at the XPS spectra of ZrO2(1.3 ML)/ 8027

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

Figure 2. Al Kα XPS C 1s spectra referring to ZrO2(1.3 ML)/ TiO2(110) films obtained at 680 K. Components corresponding to aliphatic and oxygen bound carbon atoms are indicated by the C1, C2, and C3 labels. The peak corresponding to carbonate or oxalate component is indicated by the C4 label. This component disappears at the highest deposition temperature of 770 K.

related to a carbon atom next to the carbonyl one, and, finally, the C3 component corresponds to a carbonyl carbon atom. The last component, C4, as suggested by the very high BE value can be assigned only to oxalate or carbonate species. This would suggest an interaction with the oxygen atoms from both substrate and overlayer at step sites. This is further reinforced by the fact that the C4 feature is strongly dependent on the deposition temperature, and disappears above 750 K. Finally, the fact that we observe no components at BE lower than 284 eV, corresponding to carbides, suggests that iso-butene does not form either Ti−C or Zr−C bonds, but rather its byproducts remain trapped at the interface. The amount of carbonaceous species, as deduced from XPS, decreases with increasing deposition temperature passing from 15% to 20% at 740 and 680 K down to less than 10% for the highest deposition temperature (we assume that the carbon contamination concentrated mainly at the interface layer). Interestingly, the carbon amount decreases by increasing the film thickness at 770 K, while in the case of the lower deposition temperatures thicker films have higher amount of carbonaceous species. In order to have some evidence of the presence of a mixed oxide at the interface, we report in Figure 3 the behavior of Ti 2p and Zr 3d signals at increasing deposition time (red curve and dots refer to Zr 3d signal, black curve and dots refer to Ti 2p signal). In order to be independent from X-ray source photon flux and sample position the Ti 2p and Zr 3d signals were normalized with respect to the O 1s peak intensity that we expect to be virtually constant considering the identical stoichiometry of the two oxides. If a sharp interface is formed, one would expect the Ti signal to decrease by following a perfect exponential attenuation, while it is immediately evident from the data reported in Figure 3 that this is not the case. The Ti 2p signal appears rather constant at deposition times higher than 2 and 3 min, in the case of films grown at 680 K (Figure 3a) and 740 K (Figure 3b), respectively, while in the case of the highest growth temperature (770 K) the signal is constant only after deposition times higher than 4 min. The highest deposition temperature corresponds to the highest deposition rate, as deduced by the Zr 3d/Ti 2p ratio. The rather constant intensity of the Ti 2p signal, obtained in the case of the lowest growth temperature and lowest deposition rate, is due to the diffusion of Ti4+ ions through a thinner ZrO2 film (assuming the diffusion rate of Ti4+ almost constant in the studied temperature range). These data, nevertheless, do not allow us

Figure 1. Al Kα XPS spectra from ZrO2/TiO2(110) grown at 680 K. Zr 3d, Ti 2p, and O 1s signals refer to a ZrO2 increasing coverage, as indicated by the arrow. On the left side of graph b the fitting of the Ti 2p3/2 component is shown in order to highlight the presence of the weak signal at 457.8 eV corresponding to Ti3+.

TiO2(110) deposits obtained at 680 K we observe different carbonaceous components. Iso-butylene (or iso-butene) was evidenced by Cameron et al.23,34 as the major byproduct of the interaction of ZTB on Si(100) via β-elimination from tertbutanol. In particular, Karlsson et al.28 noted that bonds form with the Si surface, and observed mainly Si−C compounds and bound methyl groups, while C−O bonds were found only on Si(111). In our work due to the oxidic nature of the surface we observe, according to the XPS measurements, different reaction products. Robbins et al.35 studied the reactivity of isobutene with TiO2(110) and verified that methacrolein formed on the surface. Assuming the β-elimination mechanism occurs, isobutene can be formed after decomposition of ZTB on the surface. Thus, a possible interpretation of the C 1s spectra reported in Figure 2 can be the following: the components labeled C1, C2, and C3 are associated to carbonaceous species containing aliphatic and oxygen bound carbon atoms, where C1 refers to the methyl or methylene carbon atoms, C2 can be 8028

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

Figure 3. Ti 2p and Zr 3d intensities for ZrO2/TiO2(110) films as function of the deposition times. Graphs a−c correspond to increasing deposition temperatures. All intensities values are normalized to the O 1s intensity.

ML)/TiO2(110) ultrathin film prepared at 770 K (in order to minimize the carbon contamination). Preliminary considerations suggest that, of the three zirconia polymorphs (i.e., monoclinic, tetragonal, and cubic), the tetragonal one has the lowest mismatch value with respect to the TiO2(110) surface. A possible epitaxial relation (as depicted in Figure 4) can be ZrO2(001)[100]//TiO2(110)[11̅1] with a mismatch of about 3% along direction [11̅1] and about 9% along direction [115̅ ]̅ of TiO2. If the hypothesis of a tetragonal matched film would be true, one could expect to see a LEED pattern with a rotated superstructure with respect to the TiO2(110) pattern. On the contrary, we observe that LEED

to discriminate between a truly mixed oxide at the interface (with substitution of Zr4+ by Ti4+) or if a simple mixture of the two oxides is obtained. We actually assumed the second hypothesis to estimate the thickness and the growth rate (r) of the ZrO2 films. From these data, reporting the ln(r) as function of 1/T, it is possible to estimate the activation energy for the decomposition of ZTB on TiO2(110): the obtained value is 56 ± 12 kJ/mol, whereas for the same reaction on Si(100) a value of about 49 kcal/mol was found.23 As a final comment we would like to point out that our data show a different trend from those found by Cameron et al.23 in their paper. These authors, in fact, claim that carbon poisoning slowed the film growth as more carbon was deposited at higher temperatures. Karlsson et al.,28 as well, claimed that the decrease in the growth rate was due to the formation of an extended interface layer, followed by the formation of ZrO2 that prevented the diffusion of Si on the surface. In their case, Si was assumed to favor the decomposition of ZTB. In our findings the film growth is faster and with less carbon at higher temperatures. Since titania is a reducible oxide, we can rationalize the difference taking into account the diffusion of bulk oxygen atoms from the substrate to the film, forming oxidized carbon species (CO and CO2) that could easily desorb from the surface. Structural Characterization of the ZrO2/TiO2 Interface. Once the growth and the chemical characterization of the ZrO2 ultrathin film has been put on a firm basis, we moved toward a structural characterization of the ZrO2/TiO2 interface using electron diffraction techniques (i.e., LEED and AS-XPD) in UHV. To this end we have taken the lowest coverage ZrO2(1.3

Figure 4. Epitaxial relation between tetragonal ZrO2(001) and TiO2(110). Blue circles correspond to Ti atoms, red circles to O atoms. The tetragonal ZrO2 unit cell is highlighted in black, while two adjacent cells are indicated by the green transparent rectangle. 8029

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

images are showing diffused patterns, for coverage up to 1.3 ML, identical to those of TiO2(110) (1 × 1). This evidence rules out any hypothesis of a matched epitaxial ZrO2(001) film on TiO2(110). Thus, we extended our investigation to AS-XPD scans in order to achieve some information about the Zr ions crystallographic sites. We remind the reader that AS-XPD is known for being a very useful and straightforward tool based on elastic electron-atom scattering to identify the possible formation of surface alloys and locally ordered ultrathin films by examining the short-range order around each emitting atom.30,36,35,37 Differently from LEED, the electrons participating to the scattering events are originated inside the sample by photoemission from core levels. Any signal originated on notordered emitter atoms is averaged to zero. In Figure 5 we report the forward scattering (FS) directions for a clean TiO2(110) surface, where a and b represent side

Figure 5. Side views of rutile-TiO2(110) with principal forward scattering (FS) directions, indicated by the arrows. The polar angles are measured with respect to the surface. Blue-gray atoms = Ti, red atoms = O. Figure 6. Azimuthal XPD scans obtained from (a) a clean TiO2(110) surface (Ti 2p signal) and (b, c) ZrOx(1.3 ML)/TiO2(110) (Ti 2p and Zr 3d signals).

views of the surface aligned with the [001] and [11̅0] directions, respectively. The arrows indicate the main FS events with takeoff angles measured with respect to the surface. In Figure 6 we report the Ti 2p azimuthal scans for different polar angles θ obtained from the following: (a) a clean TiO2(110) surface, (b) after growth of 1.3 ML of ZrO2, and (c) the Zr 3d azimuthal scans from the same sample. If we look at the Ti 2p azimuthal scan it is very easy to identify the [100] and [11̅0] directions containing the two symmetry planes of this surface and corresponding respectively to Ti−O bonds and Ti− Ti metal−metal FS directions. When the analysis direction corresponds to the Ti−O bonds we see a FS peak at θ = 38− 39°, while if the analysis direction is aligned with the Ti−Ti direction we find a FS peak at θ = 45°. The azimuthal scans referring to the Zr 3d signal, although somewhat similar to the Ti 2p ones, show a different pattern, and in particular, we notice that the FS peak aligned with direction [11̅0], that is so evident in the case of the Ti 2p at θ = 45° (Figure 6a,b), is hardly visible in the case of Zr 3d (Figure 6c). This means that Zr4+ and Ti4+ are not occupying the same crystal sites and there is no formation of a substitutional oxide. This observation is not unexpected since Ti and Zr have very different sizes, and Zr, in ZrO2, is characterized by 8-fold coordination of oxygen ions. In Figure 7a we report the complete 2π stereographic plot obtained for Zr 3d from θ = 24° to 60°. The 2π plot shown in Figure 7a is somewhat similar to the one measured in the case of VOx nanoclusters grown on TiO2(110) at coverage as low as 0.3 ML.38,39 In that case, photoemission and XPD techniques, combined with density functional theory calculations, allowed the identification of exotic V4O6 nanoclusters, holding vanadyl

groups, although the vanadium oxidation state is formally +3. In our case we have a higher surface coverage, compatible with the one found in the case of RuOx nanostripes grown on TiO2(110).40 In particular, we can notice that the only strong diffraction feature appears at θ = 40° and aligned with direction [001] of the substrate. Other, less pronounced features, can be seen at θ = 46°, aligned with direction [11̅0], θ = 30° and shifted by 32−33° from direction [110̅ ], and at θ = 56° and shifted by 45° from direction [11̅0] (see arrows in Figure 7a). This pattern was then compared with a series of MSSW simulations searching for the best agreement with a specific structural model. A satisfactory agreement between experimental and simulated patterns is obtained by assuming the formation of Zr2O3 nanostripes, similar to the model suggested by Blanco-Rev et al.41 in the case of (1 × 2) reconstruction of TiO2(110) or by Rodriguez et al.40 for RuOx nanostructures on TiO2(110). This model is reported in Figure 8. We have found that the best agreement is obtained when some Ti ions of the topmost layer of TiO2(110) are substituted by Zr. In particular, the substitution of Ti ions bound to the bridging oxygen of TiO2(110) surface with Zr is effective in reducing the R-factor, while the contribution of Zr substituting 5-fold Ti ions is negligible. In Figure 7b we report the best simulation with the Zr2O3 nanostripes model: it reproduces the strong FS feature corresponding to the formation of Zr−O bonds with oxygen in a bridging configuration and also the weaker features indicated by the arrows. The corresponding R-factor value is 0.6, better 8030

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

the surface. Probably other not-ordered patches are present in the film. We also tested a further model, i.e., Zr2O4 nanostripes model (Figure 7f), but the agreement is very poor, as can be easily seen in the figure. As a final comment we would like to point out that the formation of an interface layer with a formally Zr3+ species is also in tune with the XPS data that show a progressive shift of the Zr 3d peak from 182.9 to 183.3 eV in the case of thicker deposits.42,43 In order to have a further validation of this model, we have exposed the surface of the ZrO2(1.3 ML)/TiO2(110) film to 100 L of pyridine and compared the adsorbed amount with that chemisorbed on a clean TiO2(110) surface. On TiO2(110), pyridine adsorbs on the 5-fold coordinated Ti atom, the acid site, and 4-fold coordinated Ti atoms at step edges.44 In the case of the ZrOx nanostripe models, the 5-fold Ti acid sites are covered by the ZrOx nanostructures, while the only possible adsorption sites could be on the areas beside the nanostripes. In Figure 9 we report the XPS N 1s signal recorded after exposure to 100 L of pyridine of a TiO2(110) surface and ZrO2(1.3 ML)/TiO2(110). By taking into account the Ti atom density of TiO2(110) (1.05 × 1015 at/cm2) and Zr atom density of tetragonal ZrO2(001) (7.33 × 1014 at/cm2) it is easy to verify that 1.3 ML of ZrO2 corresponds to 0.9 ML of rutile-TiO2, and therefore, one expects that the uncovered 5-fold Ti acid sites, where pyridine could adsorb, are only 1/10 of the amount on TiO2(110). The amount adsorbed in our ZrO2(1.3 ML)/ TiO2(110) sample, according to the N 1s XPS area, is about 1/5 of that adsorbed on TiO2(110). This approximate model would indicate that some molecules of pyridine are adsorbed on uncovered 5-fold coordinated and step edge Ti sites, as expected, but also on disordered ZrO2 patches and/or on lateral sites of the nanostripes. It has to be noted that in general ZrO2/TiO2 composites exhibit a higher acidity with respect to the single components,45 so the unusual more basic properties observed in the present case are connected to the special structural configuration of the interface, where the acid sites (5fold coordinated Ti atoms) are mostly covered by the Zr2O3 nanostripes. The ability to grow these special nanostructures therefore offers a simple way to modify in a controlled way the surface chemistry since the amount and distance between basic and acid sites is the most critical parameter controlling the absorption and decomposition of molecules.46

Figure 7. (a) Stereographic projections of XPD curves obtained from of ZrO2(1.3 ML)/TiO2(110). The data are acquired from θ = 24° to 60° from the surface. (b−f) MSSW simulations of XPD patterns obtained assuming different models (see text).



CONCLUSION ZTB was a quite convenient precursor for the formation of zirconia films. The obtained zirconia films for coverage higher than 3−4 ML were free from carbonaceous species. The optimization of the deposition procedure allowed us to obtain a satisfactorily, reproducible, linear film growth, even by varying the pressure. The interface showed some carbon contamination that, although unbound to metal atoms of either substrate or film (no metal carbides), nevertheless warped the deposit crystal structure. Contamination at the interface was found to be unavoidable and spectacularly tough as depositions in an oxygen environment or long oxidized annealing only partially removed it. This may not have necessarily been a bad thing as previous experiments47,48 have demonstrated that the presence of carbon is useful in zirconia and mixed ZrO2−TiO2 based catalysts in the dehydrogenation of alkanes by preventing particle sintering and improving selectivity. We have then studied the formation of an interface layer, demonstrating that it is consisting of Zr2O3 nanostripes aligned along direction [001] with partial substitution of Ti ions by Zr

Figure 8. Zr2O3 nanostripe model corresponding to XPD simulated pattern reported in Figure 7b. The Zr−O bond length is assumed to be 2.10 Å; Zr atoms are blue-green, Ti atoms are gray, and O atoms are red.

than those obtained assuming other models: Zr2O3 nanostripes (Figure 7d, no substitutional Zr in the topmost layer, R-factor = 0.77), Zr3O6 nanostripes (Figure 7c, with substitutional Zr ions in the topmost layer, R-factor = 1.06), or Zr3O6 nanostripes (Figure 7e, no substitutional Zr ions in the topmost layer, Rfactor = 0.78). It is interesting that simulation of Figure 7d does not reproduce the weak features at grazing angles. The not fully satisfactory numerical agreement between theory and experiment (i.e., high R-factor values) is due to the much higher anisotropy value obtained from theory and is indicative that the nanostripes do not represent the unique structure present in 8031

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

Figure 9. N 1s signal after exposure to 100 L of pyridine for (a) clean TiO2(110) and (b) ZrO2(1.3 ML)/TiO2(110) surface. (2) Peacock, P. W.; Robertson, J. Behavior of Hydrogen in High Dielectric Constant Oxide Gate Insulators. Appl. Phys. Lett. 2003, 83, 2025−2027. (3) Copel, M.; Gribelyuk, M.; Gusev, E. Structure and Stability of Ultrathin Zirconium Oxide Layers on Si(001). Appl. Phys. Lett. 2000, 76, 436−438. (4) Tanabe, K.; Yamaguchi, T. Acid-Base Bifunctional Catalysis by ZrO2 and Its Mixed Oxides. Catal. Today 1994, 20, 185−197. (5) Yamaguchi, T. Application of ZrO2 as a Catalyst and a Catalyst Support. Catal. Today 1994, 20, 199−217. (6) Fung, J.; Wang, I. The Reaction Mechanism of C6 Hydrocarbons over Acid−Base Bifunctional Catalysts, TiO2−ZrO2. J. Catal. 1996, 164, 166−172. (7) Schilke, T. C.; Fisher, I. A.; Bell, A. T. In Situ Infrared Study of Methanol Synthesis from CO2/H2 on Titania and Zirconia Promoted Cu/SiO2. J. Catal. 1999, 184, 144−156. (8) Yang, S.; Maroto-Valiente, A.; Benito-Gonzalez, M.; RodriguezRamos, I.; Guerrero-Ruiz, A. Methane Combustion over Supported Palladium Catalysts: I. Reactivity and active phase. Appl. Catal., B 2000, 28, 223−233. (9) Farrauto, R. J.; Heck, R. M. Catalytic Converters: State of the Art and Perspectives. Catal. Today 1999, 51, 351−360. (10) Kohno, Y.; Tanaka, T.; Funabiki, T.; Yoshida, S. Reaction Mechanism in the Photoreduction of CO2 with CH4 over ZrO2. Phys. Chem. Chem. Phys. 2000, 2, 5302−5307. (11) Kitiyanan, A.; Sakulkhaemaruethai, S.; Suzuki, Y.; Yoshikawa, S. Structural and Photovoltaic Properties of Binary TiO2−ZrO2 Oxides System Prepared by Sol−gel Method. Compos. Sci. Technol. 2006, 66, 1259−1265. (12) Paz, Y. Application of TiO2 Photocatalysis for Air Treatment: Patents’ Overview. Appl. Catal., B 2010, 99, 448−460. (13) O’Regan, B.; Gratzel, M. A Low-Cost, High-Efficiency Solar Cell Based on Dye-sensitized Colloidal TiO2 films. Nature 1991, 353, 737− 740. (14) Siedl, N.; Elser, M. J.; Bernardi, J.; Diwald, O. Functional Interfaces in Pure and Blended Oxide Nanoparticle Networks: Recombination versus Separation of Photogenerated Charges. J. Phys. Chem. C 2009, 113, 15792−15795.

in the outermost layer of the substrate, as evidenced by ASXPD scans and MSSW simulations. These nanostructures are indeed very similar to the Ti2O3 stripes that are the building blocks of (1 × 2) reconstruction of TiO2(110),41 and share also strong similarities with the vanadia nanostructures recently observed on this same surface.38 This seems to indicate that the formation of this nanostripe motif is a quite general phenomenon and can be exploited to grow new materials characterized by unexpected properties.38,40 As a matter of fact, also in the present case zirconium atoms show a quite unique coordination and an unusual oxidation state (Zr3+) suggesting that this system could have special chemical properties. In fact, by using pyridine as a probe molecule, the Zr2O3-nanostripes/ TiO2(110) exhibited profoundly modified acid/basic properties with respect to the TiO2(110) surface and typical TiO2/ZrO2 composites.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work has been funded by the Italian Ministry of Instruction, University and Research (MIUR) through the FIRB Project RBAP115AYN “Oxides at the nanoscale: multifunctionality and applications”, and through the fund “Programs of national relevance” (PRIN-2009).



REFERENCES

(1) Jiang, S. P.; Chan, S. H. A Review of Anode Materials Development in Solid Oxide Fuel Cells. J. Mater. Sci. 2004, 39, 4405− 4439. 8032

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033

The Journal of Physical Chemistry C

Article

(36) Fadley, C. S. X-ray Photoelectron Spectroscopy: Progress and Perspectives. J. Electron. Spectrosc. 2010, 178, 2−32. (37) Rizzi, G. A.; Reeder, A. E.; Agnoli, S.; Granozzi, G. Epitaxial MoOx Nanostructures on TiO2(110) Obtained Using Thermal Decomposition of Mo(CO)6. Surf. Sci. 2006, 600, 3345−3351. (38) Artiglia, L.; Agnoli, S.; Vittadini, A.; Verdini, A.; Cossaro, A.; Floreano, L.; Granozzi, G. Atomic Structure and Special Reactivity Toward Methanol Oxidation of Vanadia Nanoclusters on TiO2(110). J. Am. Chem. Soc. 2013, 135, 17331−17338. (39) Agnoli, S.; Sambi, M.; Granozzi, G.; Castellarin-Cudia, C.; Surnev, S.; Ramsey, M. G.; Netzer, F. P. The Growth of Ultrathin Films of Vanadium Oxide on TiO2(110). Surf. Sci. 2004, 562, 150− 156. (40) Yang, F.; Kundu, S.; Vidal, A. B.; Graciani, J.; Ramirez, P. J.; Senanayake, S. D.; Stacchiola, D.; Evans, J.; Liu, P.; Sanz, J. F.; Rodriguez, J. A. Determining the Behavior of RuOx Nanoparticles in Mixed-Metal Oxides: Structural and Catalytic Properties of RuO2/ TiO2(110) Surfaces. Angew. Chem., Int. Ed. 2011, 50, 10198−10202. (41) Blanco-Rey, M.; Abad, J.; Rogero, C.; Mendez, J.; Lopez, M. F.; Martin-Gago, J. A.; de Andres, P. L. Structure of Rutile TiO2 (110)-(1 × 2): Formation of Ti2O3 Quasi-1D Metallic Chains. Phys. Rev. Lett. 2006, 96, 055502. (42) Luo, S. X.; Wang, F. M.; Shi, Z. S.; Xin, F. Preparation and Photocatalytic Activity of Zr Doped TiO2. Mater. Res. Innovations 2009, 13, 64−69. (43) Xue, K. H.; Blaise, P.; Fonseca, L. R. C.; Nishi, Y. Prediction of Semimetallic Tetragonal Hf2O3 and Zr2O3 from First Principles. Phys. Rev. Lett. 2013, 110, 065502. (44) Suzuki, S.; Yamaguchi, Y.; Onishi, H.; Sasaki, T.; Fukui, K.; Iwasawa, Y. Study of Pyridine and Its Derivatives Adsorbed on a TiO2(110)-(1 × 1) Surface by Means of STM, TDS, XPS and MD Calculation in Relation to Surface Acid-Base Interaction. J. Chem. Soc., Faraday Trans. 1998, 94, 161−166. (45) Manriquez, M. E.; Lopez, T.; Gomez, R.; Navarrete, J. Preparation of TiO2-ZrO2 Mixed Oxides with Controlled Acid-Basic Properties. J. Mol. Catal. A: Chem. 2004, 220, 229−237. (46) Pang, C. L.; Lindsay, R.; Thornton, G. Chemical Reactions on Rutile TiO2(110). Chem. Soc. Rev. 2008, 37, 2328−2353. (47) Trunschke, A.; Hoang, D. L.; Radnik, J.; Brzezinka, K. W.; Brückner, A.; Lieske, H. Transition Metal Oxide/Carbon Composite Catalysts for n-Alkane Aromatization: Structure and Catalytic Properties. Appl. Catal., A 2001, 208, 381−392. (48) Hoang, D. L.; Preiss, H.; Parlitz, B.; Krumeich, F.; Lieske, H. Zirconia/Carbon Composites as Monofunctional Catalysts in C6+ Alkane Aromatization. Appl. Catal., A 1999, 182, 385−397.

(15) Zou, H.; Lin, Y. S. Structural and Surface Chemical Properties of Sol−Gel Derived TiO2−ZrO2 Oxides. Appl. Catal., A 2004, 265, 35− 42. (16) Reddy, B. M.; Khan, A. Recent Advances on TiO2−ZrO2 Mixed Oxides as Catalysts and Catalyst Supports. Catal. Rev. 2005, 47, 257− 296. (17) Liang, L.; Sheng, Y.; Xu, Y.; Wu, D.; Sun, Y. Optical Properties of Sol−Gel Derived ZrO2−TiO2 Composite Films. Thin Solid Films 2007, 515, 7765−7771. (18) Chary, K. V. R.; Sagar, G. V.; Naresh, D.; Seela, K. K.; Sridhar, B. Characterization and Reactivity of Copper Oxide Catalysts Supported on TiO2−ZrO2. J. Phys. Chem. B 2005, 109, 9437−9444. (19) Livraghi, S.; Olivero, F.; Paganini, M. C.; Giamello, E. Titanium Ions Dispersed into the ZrO2 Matrix: Spectroscopic Properties and Photoinduced Electron Transfer. J. Phys. Chem. C 2010, 114, 18553− 18558. (20) Mitu, B.; Vizireanu, S.; Birjega, R.; Dinescu, M.; Somacescu, S.; Osiceanu, P.; Pârvulescu, V.; Dinescu, G. Comparative Properties of Ternary Oxides of ZrO2−TiO2−Y2O3 Obtained by Laser Ablation, Magnetron Sputtering and Sol−Gel Techniques. Thin Solid Films 2007, 515, 6484−6488. (21) Hernández-Alonso, M. D.; Coronado, J. M.; Bachiller-Baeza, B.; Fernández-García, M.; Soria, J. Influence of Structural and Surface Characteristics of Ti1‑xZrxO2 Nanoparticles on the Photocatalytic Degradation of Methylcyclohexane in the Gas Phase. Chem. Mater. 2007, 19, 4283−4291. (22) Bradley, D. C. Metal Alkoxides as Precursors for Electronic and Ceramic Materials. Chem. Rev. 1989, 89, 1317−1322. (23) Cameron, M. A.; George, S. M. ZrO2 Film Growth by Chemical Vapor Deposition Using Zirconium Tetra-tert-butoxide. Thin Solid Films 1999, 348, 90−98. (24) Lin, Y.-S.; Puthenkovilakam, R.; Chang, J. P.; Bouldin, C.; Levin, I.; Nguyen, N. V.; Ehrstein, J.; Sun, Y.; Pianetta, P.; Conard, T.; Vandervorst, W.; Venturo, V.; Selbrede, S. Interfacial Properties of ZrO2 on Silicon. J. Appl. Phys. 2003, 93, 5945−5952. (25) Huang, S.-S.; Wu, T.-B. Thickness Dependence of Structural and Electrical Characteristics of ZrO2 Thin Films as Grown on Si by Chemical-vapor Deposition. J. Vac. Sci. Technol., B 2004, 22, 1940− 1948. (26) Puthenkovilakam, R.; Chang, J. P. Valence Band Structure and Band Alignment at the ZrO2/Si Interface. Appl. Phys. Lett. 2004, 84, 1353−1355. (27) Karlsson, P. G.; Richter, J. H.; Blomquist, J.; Uvdal, P.; Grehk, T. M.; Sandell, A. Metal Organic Chemical Vapor Deposition of Ultrathin ZrO2 Films on Si(100) and Si(111) Studied by Electron Spectroscopy. Surf. Sci. 2007, 601, 1008−1018. (28) Karlsson, P. G.; Göthelid, E.; Richter, J. H.; Sandell, A. Initial Stages of ZrO2 Chemical Vapor Deposition on Si(100)-(2 × 1) From Zirconium Tetra-tert-butoxide. Surf. Sci. 2008, 602, 1803−1809. (29) Fadley, C. S. Angle-Resolved X-ray Photoelectron Spectroscopy. Prog. Surf. Sci. 1984, 16, 275−388. (30) Granozzi, G.; Sambi, M. Angle-Scanned Photoelectron Diffraction: Probing Crystalline Ultrathin Films. Adv. Mater. 1996, 8, 315−326. (31) de Abajo, F. J. G.; Van Hove, M. A.; Fadley, C. S. Multiple Scattering of Electrons in Solids and Molecules: A Cluster-Model Approach. Phys. Rev. B 2001, 63, 075404. (32) Zhou, Y.; Kojima, N.; Sasaki, K. Growth and Dielectric Properties of Tetragonal ZrO2 Films by Limited Reaction Sputtering. J. Phys. D: Appl. Phys. 2008, 41, 175414. (33) Sakamoto, K.; Zhang, H. M.; Uhrberg, R. I. G. Observation of Two Metastable Oxygen Species Adsorbed on a Si(111)-(7 × 7) Surface: Reinterpretation of The Initial Oxidation Process. Phys. Rev. B 2003, 68, 075302. (34) Gould, B. J.; Povey, I. M.; Pemble, M. E.; Flavell, W. R. Chemical Vapour Deposition of ZrO2 Thin Films Monitored by IR Spectroscopy. J. Mater. Chem. 1994, 4, 1815−1819. (35) Robbins, M. D.; Henderson, M. A. The Partial Oxidation of Isobutene and Propene on TiO2(110). J. Catal. 2006, 238, 111−121. 8033

dx.doi.org/10.1021/jp500759h | J. Phys. Chem. C 2014, 118, 8026−8033