Tailored Vapor-Phase Growth of Cu - American Chemical Society

ARTICLE pubs.acs.org/Langmuir

Tailored Vapor-Phase Growth of CuxO
TiO2 (x = 1, 2) Nanomaterials Decorated with Au Particles Davide Barreca,*,† Giorgio Carraro,‡ Alberto Gasparotto,‡ Chiara Maccato,‡ Oleg I. Lebedev,§ Anna Parfenova,‡ Stuart Turner,|| Eugenio Tondello,‡ and Gustaaf Van Tendeloo|| †

CNR-ISTM and INSTM, Department of Chemistry, University of Padova, Via Marzolo, 1, 35131 Padova, Italy Department of Chemistry, University of Padova and INSTM, Via Marzolo, 1, 35131 Padova, Italy § Laboratoire CRISMAT, UMR 6508, CNRS-ENSICAEN, Bd. Mar
echal Juin, 6, 14050 Caen CEDEX 4, France EMAT, University of Antwerp, Groenenborgerlaan, 171, B-2020 Antwerp, Belgium

)

‡

bS Supporting Information ABSTRACT: We report on the fabrication of CuxO
TiO2 (x = 1, 2) nanomaterials by an unprecedented vapor-phase approach. The adopted strategy involves the growth of porous CuxO matrices by means of chemical vapor deposition (CVD), followed by the controlled dispersion of TiO2 nanoparticles. The syntheses are performed on Si(100) substrates at temperatures of 400
550 °C under wet oxygen atmospheres, adopting Cu(hfa)2 3 TMEDA (hfa =1,1,1,5,5,5-hexafluoro-2,4-pentanedionate; TMEDA = N,N,N0 ,N0 -tetramethylethylenediamine) and Ti(O-iPr)2(dpm)2 (O-iPr = isopropoxy; dpm = 2,2,6,6-tetramethyl-3,5heptanedionate) as copper and titanium precursors, respectively. Subsequently, finely dispersed gold nanoparticles are introduced in the as-prepared systems via radio frequency (RF)sputtering under mild conditions. The synthesis process results in the formation of systems with chemical composition and nano-organization strongly dependent on the nature of the initial CuxO matrix and on the deposited TiO2 amount. The decoration with low-size gold clusters paves the way to the engineering of hierarchically organized nanomaterials.

’ INTRODUCTION A key goal of the modern nanofabrication frontiers is the growth of multifunctional oxide systems with p
n heterojunctions, a promising research field for a plethora of advanced applications.1
4 In this context, p-type Cu2O or CuO semiconductors (Eg ≈ 2.1 and 1.2 eV, respectively)5
7 and n-type TiO2 (Eg ≈ 3.5 eV)7
9 have received considerable attention for the preparation of CuxO
TiO2 (x = 1, 2) nanomaterials, enabling to favorably exploit the inherent nanostructure advantages and the intimate contact between CuxO and TiO2,10,11 that results in an improved charge carrier separation.7
10,12 In particular, the overdispersion of TiO2 on copper oxide matrices enables to join the TiO2 stability toward photoleaching processes7,13,14 and the favorable CuxO bandgap, that can shift absorption toward the vis range. Overall, these features result in remarkable technological advantages for gas sensing15
17 and photoactivated applications (i.e., pollutant degradation4,12,13,18 and water splitting)10,14,19 as well as for a variety of catalytic processes.20,21 To date, various routes have been adopted to obtain CuxO
TiO2 nanosystems, ranging from impregnation,19 photoand electrochemical depositions,7,10,11 sol
gel and hydrothermal processes,9,12,20,22
24 sputtering,25,26 and chemical vapor deposition (CVD).8,27
30 Nevertheless, solid state reactions, leading ultimately to copper titanates, might take place during the system preparation12,14,22,25,31 and detrimentally affect the nature of the CuxO
TiO2 p
n junction and the system properties. As a consequence, the development of synthetic strategies enabling r 2011 American Chemical Society

careful control of the system properties and functional behavior, especially in photocatalysis and gas sensing, is still an open challenge. Herein, we present an unprecedented synthesis protocol for the preparation of CuxO
TiO2 (x = 1, 2) nanomaterials endowed with tunable properties (see Scheme 1). In particular, Cu2O and CuO nanostructures were prepared by CVD under O2 þ H2O reaction atmospheres, using Cu(hfa)2 3 TMEDA as copper molecule source.6,32
34 A great deal of attention was devoted to the development of single-phase porous copper oxide nanostructures, in order to optimize the subsequent dispersion of TiO2 achieving an intimate contact between the two oxides. A well-established titanium precursor, Ti(O-iPr)2(dpm)2, was employed in this processing step.35 A further innovative and unprecedented issue addressed in the present work was the decoration of CuxO
TiO2 nanosystems with Au nanoparticles (NPs), whose peculiar reactivity and functionality as electron sinks contribute to enhance carrier lifetime and material performances especially for photocatalysis and gas sensing.36 In order to prevent undesired modifications of the oxide host matrix, in this investigation gold introduction was performed by radio frequency (RF)-sputtering under mild

Received: February 23, 2011 Revised: April 1, 2011 Published: April 25, 2011 6409

dx.doi.org/10.1021/la200698t | Langmuir 2011, 27, 6409–6417

Langmuir Scheme 1. Schematic Representation of the CVD/RF-Sputtering Synthetic Route Adopted for the Preparation of CuxO
TiO2
Au Nanosystems

ARTICLE

Table 1. Synthesis Conditions Adopted for the Preparation of the Systems Prepared in the Present Worka

a

synthesis conditions, optimizing the metal particle dispersion and spatial distribution. Despite that various research efforts have been dedicated to TiO2 sensitization by metal particles,36,37 only a few studies were devoted to modification of copper oxides with Au NPs (for instance, for gas sensing).38 Recently, the preparation of powdered CuxO
TiO2
Au catalysts was described,39 whereas no literature reports on the synthesis of such supported systems are available to date.

’ EXPERIMENTAL SECTION Nanomaterial Synthesis. The precursor Cu(hfa)2 3 TMEDA used for CuxO (x = 1, 2) deposition was prepared following a recently reported procedure.32,34 Copper oxide nanostructures were grown in a custom-built cold-wall CVD apparatus40,41 consisting of a quartz chamber, a resistively heated susceptor, and an external reservoir for precursor vaporization. Starting from our recent results,6 the synthesis of pure Cu2O and CuO nanostructures was performed at temperatures of 400 and 550 °C, with total pressures of 3.0 and 10.0 mbar, respectively (Table 1, column (a)). In each case, the deposition time was set at 120 min and the precursor vaporized at 70 °C. Vapor transport toward the deposition zone was performed through gas lines heated at 120 °C by an O2 flow (purity = 6.0; flow rates = 20 and 100 sccm for Cu2O and CuO, respectively). An auxiliary oxygen flow (20 and 100 sccm for the two cases) was introduced separately into the reaction chamber after passing through a water reservoir maintained at 50 °C. Under these conditions, H2O partial pressure was estimated to be ≈0.3 (1.5) mbar for a total reactor pressure of 3.0 (10.0) mbar.35,42 Depositions were performed on p-type Si(100) substrates (MEMC, Merano, Italy, 10 mm 10 mm  1 mm), subjected to a previously described precleaning procedure.13 Subsequent CVD of TiO2 was accomplished from Ti(O-iPr)2(dpm)2 (99%, Aldrich), vaporized at 80 °C. The growth process was carried out at a total pressure of 10.0 mbar and a substrate temperature of 400 °C,

For the sake of clarity, Au-free samples are marked in gray.

with 40 and 80 sccm flow rates for the precursor mass transport (O2) and the auxiliary (O2 þ H2O) inlet, respectively. The choice of the growth temperature was based on previous experiments on the bare Si(100) substrates, indicating this value as the minimum one necessary to obtain crystalline TiO2 in the anatase phase.35 The use of higher temperatures was discarded in order to prevent undesired modifications of the underlying Cu2O and CuO matrices. TiO2 content in the obtained specimens was varied by using different deposition times, that is, 5 and 20 min (compare Table 1, column (b)). All the others parameters were set as for the CuxO deposition. At the end of each experiment, the resulting specimens were cooled down at room temperature before contacting the external atmosphere. Finally, Au deposition on CuxO
TiO2 (Table 1, column (c)) was performed by using a custom-built RF plasmochemical reactor (ν = 13.56 MHz) using electronic grade Ar (purity = 5.0) as plasma source.43 A gold target (2 in. diameter, 0.1 mm thick; BALTEC AG, 99%) was fixed on the RF electrode, while the Si(100)-supported samples were mounted on a second grounded electrode. The substrate temperature was measured via a thermocouple inserted into the resistively heated sample holder. Gold dispersion was performed under the following conditions: substrate temperature = 60 °C; RF-power = 5 W; total pressure = 0.38 mbar; Ar flow rate = 10 sccm; duration = 10 min. The obtained CuxO
TiO2
Au systems were not subjected to ex situ thermal treatments to prevent the occurrence of solid state reactions leading to the possible formation of copper titanate ternary phases.31 Characterization. X-ray photoelectron and X-ray excited Auger electron spectroscopy (XPS and XE-AES) analyses were carried out by using a Perkin-Elmer Φ5600ci spectrometer at pressures lower than 1  10
8 mbar, using a non-monochromatized Al KR source (hν = 1486.6 eV). After a Shirley-type background subtraction, raw XPS spectra were fitted using a nonlinear least-squares deconvolution adopting Gaussian
Lorentzian peak shapes. The reported binding energies (BEs, standard deviation = (0.2 eV) were corrected for charging effects by assigning to the C1s line of adventitious carbon a position of 284.8 eV.44 The Cu(I) and Cu(II) oxidation states were discriminated using the Auger R parameter.5,6,44 Quantitation was performed using sensitivity factors provided by Φ V5.4A software. 6410

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir

ARTICLE

In-depth analyses were carried out by Arþ sputtering at 4.5 kV and partial pressure of 5  10
8 mbar. Glancing incidence X-ray diffraction (GIXRD) patterns were collected at a constant incidence angle of 1° via a Bruker D8 Advance diffractometer equipped with a G€obel mirror, using a Cu KR X-ray source powered at 40 kV and 40 mA. The angular accuracy was 0.001°, and the angular resolution was better than 0.01°. the Scherrer equation was used to estimate the mean crystallite size. Field emission scanning electron microscopy (FE-SEM) images were collected by using a Zeiss SUPRA 40VP instrument equipped with an Oxford INCA x-sight X-ray detector for energy dispersive X-ray spectroscopy (EDXS) investigation. Plane-view and cross-sectional micrographs were recorded with accelerating voltages between 10 and 20 kV. Transmission electron microscopy (TEM) and electron diffraction (ED) experiments were carried out on a Tecnai G2 30 UT microscope operated at 300 kV with a 0.17 nm point resolution. High angle annular darkfield-scanning transmission electron microscopy (HAADF-STEM), high resolution (HR)-TEM, and EDXS experiments were performed on a FEI Titan 80-300 “cubed” microscope fitted with an aberrationcorrector for the imaging lens and the probe forming lens, as well as a monochromator. The HAADF detector inner semiangle used was ≈50 mrad. Cross-sectional specimens for TEM were prepared by mechanically grinding down to the thickness of approximately 20 μm, followed by Arþ ion-beam milling using a Balzers RES 101 GVN apparatus.

’ RESULTS AND DISCUSSION The optimized synthesis conditions for the present specimens are reported in Table 1. The following two sections of the present paragraph are focused on CuxO
TiO2 and CuxO
TiO2
Au nanomaterials, respectively. For comparison, pure Cu2O and CuO samples (# N°1 and # N°7), as long as Cu2O
Au and CuO
Au ones (# N°6 and # N°12), were also synthesized and analyzed. CuxO
TiO2 Nanomaterials (# N°2, 3 and # N°8, 9). Preliminary information on the system composition as a function of preparation conditions was gained by the combined use of XPS and XE-AES measurements. The attention was initially focused on the electronic states of copper and titanium, in order to verify the presence of possible chemical interactions between the single oxide components. To this aim, Figure 1 displays surface Cu2p and Ti2p photoelectron peaks. As a matter of fact, copper signals could be clearly observed even for TiO2 deposition times of 20 min, suggesting that no complete coverage of the underlying matrix took place under the adopted conditions. Irrespective of the synthesis parameters, for all Au-free CuxO
TiO2 nanosystems, the Cu2p spin
orbit components were appreciably broadened, due to the surface copresence of Cu(I) and Cu(II) contributions (Figure 1, top).45 In fact, the Cu2p3/2 signals could be decomposed by two contributing bands centered at BE = 932.2 and 934.4 eV, corresponding to Cu2O and CuO.6,18,22,26,44,46 The presence of the latter was also confirmed by the appearance of shake-up satellites located at BE ≈ 9.0 eV higher than the main spin
orbit components.5,12,18,21,24 In the case of Cu2O
TiO2 systems, such a phenomenon could be justified taking into account the high reactivity of copper(I) oxide during the deposition process, inducing its partial surface oxidation upon functionalization with TiO2 in O2 þ H2O atmospheres.4,7 Nevertheless, calculation of the copper Auger R parameter for Cu2O
TiO2 systems provided values of 1848.6 and 1849.0 eV for 5 and 20 min of TiO2 deposition (# N°2 and # N°3, respectively), values very close to that of Cu(I) oxide, indicating that the presence of Cu2O was still predominant.5,6,28 In a different way, for CuO
TiO2 materials,

Figure 1. Surface Cu2p (top) and Ti2p (bottom) photoelectron peaks for CuxO
TiO2 systems synthesized adopting the conditions of Table 1. In each case, the spectral intensity has been normalized for comparison purposes.

the copresence of Cu(I) and Cu(II) oxidation states arose from a partial surface reduction, an apparently unexpected phenomenon since titania deposition took place under an oxidizing environment. Nevertheless, similar effects at the CuO
TiO2 interface have been observed by various investigators.18,20,45 In this case, the Auger R parameter was 1849.6 and 1850.0 eV for TiO2 deposition times of 5 and 20 min (specimens # N°8 and # N°9, respectively), that is, closer to the values expected for CuO.6 The appearance of Cu(I) contribution could be due to the presence of O-deficient titania, in which coordinatively unsaturated Ti(IV) centers act as oxygen getters at the interface with CuO.18,22
24,28 The oxygen defectivity of the obtained samples was further confirmed by analyzing the O1s peak (not reported), which, beside the main lattice component (I) centered at BE = 530.1 eV, showed a second contribution (II) at BE = 531.4 eV, attributable to chemisorbed
OH groups,4,6,7,23,25,28,44,47 likely saturating O vacancies in the analyzed systems. It is worth noting that, upon TiO2 deposition, a significant decrease of the (II)/(I) ratio (from ≈0.8, in CuxO, to ≈0.2, in CuxO
TiO2) took place, indicating a corresponding decrease of
OH group content on the sample surface. Such a variation suggested that these moieties acted as nucleation centers for TiO2 particle dispersion. As explained later,
OH presence is also likely responsible for “dragging” TiO2 particles into the inner matrix regions. For all samples, the Ti2p3/2 BE of 458.6 eV (Figure 1, bottom) was in line with the presence of pure TiO2, enabling to exclude the presence of copper titanates or other ternary 6411

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir

Figure 2. GIXRD patterns of (top) Cu2O
TiO2 and (bottom) CuO
TiO2 nanomaterials, synthesized according to Table 1. The patterns of bare Cu2O (# N°1) and CuO (# N°7) systems are also reported for comparison. The markers indicate the expected Cu2O (cuprite, left triangle), CuO (tenorite, b), and TiO2 (anatase, V) reflections. The top figure inset shows an enlargement around the most intense Cu2O signal.

phases.7,9,12,18,20,25,28,47 This finding was further supported by structural analyses (see below). GIXRD patterns of CuxO
TiO2 specimens are displayed in Figure 2, together with the corresponding CuxO systems (# N°1 and 7). For Cu2O
TiO2 samples (Figure 2, top), only signals related to the Cu2O crystalline phase (cuprite)48 could be observed [for specimen # N°1: 2θ = 29.6° (110), 36.4° (111), and 42.3° (200)]. No diffraction peaks corresponding to TiO2 were detected, due to its high dispersion and relatively low amount.9,35,45,49 Nevertheless, a careful inspection of the most intense (111) signal (Figure 2, inset) revealed a progressive shift toward higher angular values [maximum Δ(2θ) = þ0.2°] on increasing the TiO2 deposition time. This effect could be attributed to a distortion of the Cu2O crystal lattice induced by the interstitial doping with Ti(IV).24 In agreement with XPS results, no reflections related to Cu
Ti
O mixed phases were present, confirming that the adopted synthesis conditions were mild enough to prevent the formation of copper titanates. The average crystallite dimensions for Cu2O were calculated to be 30 nm. In the case of CuO
TiO2 systems (Figure 2, bottom), GIXRD patterns were dominated by CuO (tenorite) reflections [2θ = 32.5°, 35.5°, 38.7°, 46.3°, and 48.8° due to (110), (002)/ (111), (111)/(200), (112), and (202) planes, respectively],50

ARTICLE

Figure 3. Plane-view (left) and cross-sectional (right) FE-SEM micrographs of CuxO
TiO2 nanomaterials deposited adopting the conditions of Table 1. The marked lines correspond to the EDXS scans displayed in Figure 4.

while no signals related to crystalline TiO2 or to mixed Cu
Ti
O phases could be observed, in analogy with the previous case. Nevertheless, upon functionalization with TiO2, a weak signal at 2θ = 36.5° due to the (111) Cu2O reflection48 was observed, and its intensity increased with the TiO2 deposition time. This phenomenon indicated a partial CuO f Cu2O reduction, in agreement with XPS and XE-AES data. At variance with Cu2O
TiO2 specimens, in this case no appreciable shift of copper(II) oxide XRD signals could be observed and the mean CuO nanocrystal size was 20 nm. In order to investigate the system nano-organization and its interrelations with the chemical composition, both plane-view and cross-sectional FE-SEM images were recorded (Figure 3). As a matter of fact, the system morphology was mainly influenced by the nature of the pristine copper oxide matrix. For Cu2O-based samples (Figure 3, # N°2 and 3), the micrographs were dominated by interconnected nanopyramidal aggregates with a mean size of 130 nm, whose uniform distribution over the substrate surface gave rise to relatively compact structures. A comparison of these images with those pertaining to pure Cu2O (# N°1, not shown) indicated that TiO2 deposition conditions were indeed sufficiently mild to avoid marked modifications of the underlying 6412

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir oxide matrix. Correspondingly, cross-sectional micrographs were characterized by pseudocolumnar structures aligned perpendicularly to the growth surface. A careful inspection enabled to discern a homogeneous dispersion of low-sized titania particles (L ≈ 15 nm), suggesting a conformal coverage of Cu2O matrices by the overdispersed TiO2. Variation of TiO2 deposition time from 5 to 20 min led to a parallel increase of the overall nanodeposit thickness from 190 to 210 nm. A significant variation in the nano-organization occurred upon switching to CuO-based systems (Figure 3, # N°8 and 9). As a matter of fact, a highly porous morphology with nanoflower-like structures (L ≈ 400 nm) was obtained. The cross-sectional view showed an even distribution of such nanoaggregates over the whole substrate surface. Similarly to the case of Cu2O
TiO2 samples, an increase of the overall mean thickness took place upon increasing the TiO2 deposition time from 5 to 20 min, i.e., on going from sample # N°8 (600 nm) to # N°9 (700 nm). The nano-organization of CuO-TiO2 specimens appears extremely appealing in view of photocatalytic/sensing applications thanks to the high system active area, in line with the high surface roughness evidenced by atomic force microscopy (AFM) measurements (Supporting Information Figure S4). In addition, the system porosity suggested a uniform dispersion of TiO2 into the CuO supporting matrix. To confirm this hypothesis, a compositional investigation by in-depth EDXS and XPS analyses was undertaken. As a matter of fact, whereas XPS information is averaged over larger sampling areas and can also be influenced by the surface roughness, EDXS line-scans along the overall deposit thickness can provide useful complementary indications on the local chemical composition. Representative EDXS profiles are plotted in Figure 4. The analyses clearly showed titanium presence throughout the nanodeposit thickness, that could be related to a vapor infiltration of the Ti precursor upon deposition on copper oxides. As regards Cu2O
TiO2 (Figure 4, top; sample # N°2), titanium was mainly confined to an outermost region ≈50 nm thick, whereas for CuO
TiO2 (Figure 4, bottom; sample # N°8) a more uniform distribution was observed. This difference highlighted that the use of the more porous CuO enhanced the in-depth TiO2 penetration. These observations are further supported by evaluating the intensity ratio between the O KR1 and the Cu KR1 X-ray lines. For sample # N°2, the obtained value was close to the stoichiometric one for Cu2O (O:Cu ≈ 1:2) for depth values higher than ≈50 nm, whereas it was appreciably higher in the outermost sample region (O:Cu ≈ 1:1). This difference provided a qualitative indication on the higher TiO2 concentration in the external region due to O from titanium oxide, as confirmed by XPS depth profiles (compare Figure 5). In a different way, for CuO
TiO2 (sample # N°8), the O:Cu value always matched the expected one for CuO (O:Cu ≈ 1:1) throughout the investigated region, since the enhanced TiO2 in-depth dispersion did not induce any appreciable alteration of the oxygen/copper ratio. It is worthwhile noticing that similar results were not affected by the specific line-scan position, indicating a good lateral homogeneity of the present nanodeposits. In order to gain deeper insight into the relative distribution of copper(I, II) and titanium oxides, CuxO
TiO2 systems (specimens # N°2 and 8, Figure 5, top) were subjected to XPS depth profile analysis, that provided key information on the elemental atomic percentages as a function of the sputtering time. As a general rule, the carbon signal (not reported) fell to noise level after 2 min erosion, indicating the high purity of all the

ARTICLE

Figure 4. Representative EDXS scans along the line marked in Figure 3 for selected CuxO
TiO2 nanomaterials. The blue, red, and green traces correspond to Cu KR1, O KR1, and Ti KR1 X-ray signals, respectively.

investigated samples. The presence of an appreciable amount of Cu even after the interface with the Si substrate can be ascribed to the concurrence of morphological variations induced by Arþ bombardment and the roughness of the present nanosystems, resulting in an enhanced deposit/substrate intermixing.44 Irrespective of TiO2 deposition time, for Cu2O
TiO2, the outermost system region was characterized by a net predominance of Ti content with respect to the Cu one. In particular, for sample # N°2 (overall thickness = 190 nm, compare FE-SEM data), a rough estimation yielded a value of ≈30 nm for the TiO2 accumulation region. Upon increasing the sputtering time, the copper atomic percentage increased at the expense of that of titanium, even though the latter was still present in the inner system regions. Such an observation highlighted that, despite the rather compact nature of Cu2O matrices (see above and Figure 3), in-depth TiO2 dispersion was indeed effective. For a titania deposition time up to 20 min (sample # N°3, data not shown), both the content and the penetration of TiO2 into the nanodeposit underwent an appreciable enhancement. A different situation occurred regarding CuO
TiO2 systems (compare Figure 5, specimen # N°8, overall thickness = 600 nm). At the sample surface, Ti was still the prevailing species, with an average atomic percentage of 25% irrespective of titania deposition time. In this case, TiO2 content was predominant in the outermost ≈60 nm. On increasing the sputtering time, Ti content underwent a slow decrease and, at variance with the previous case, was still appreciable up to the increase of the Si substrate signal. A similar trend proved that TiO2 was indeed successfully dispersed throughout CuO matrices thanks to their higher porosity with respect to Cu2O ones. As already anticipated, further effects synergistically contributing to the in-depth 6413

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir

ARTICLE

Figure 5. Representative XPS depth profiles for selected CuxO
TiO2 (top) and CuxO
TiO2
Au (bottom) specimens deposited adopting the conditions of Table 1: (0) Cu, (/) Ti, (4) Si, and (b) Au.

distribution of titanium oxide are the inherent CVD infiltration power and the presence of
OH groups acting as catalytic centers for the Ti precursor decomposition.6,9 For sample # N°9, corresponding to a TiO2 deposition time of 20 min (data not shown), the main difference was an increase of the TiO2 overall content and of the corresponding deposit thickness, in agreement with crosssectional FE-SEM results (see above and Figure 3). The results obtained from XPS profiles on CuxO
TiO2 systems are summarized in Supporting Information (Figure S3, top) in order to evidence both the impact of the copper oxide matrix and TiO2 deposition time on the in-depth composition. CuxO
TiO2
Au Nanomaterials (# N°4, 5 and # N°10, 11). On the basis of the obtained results, efforts were subsequently dedicated to RF-sputtering of gold NPs on CuxO
TiO2 specimens (see Table 1). Surface XPS and XE-AES analyses did not evidence any appreciable variation with respect to the homologous Au-free samples regarding Cu oxidation states (compare Figure 1, top and Supporting Information Figure S1, top). The metallic state of gold in all CuxO
TiO2
Au specimens was confirmed by the position of surface Au4f photoelectron peaks [BE(Au4f7/2) = 84.4 eV; Supporting Information Figure S1, bottom].39,44,47 GIXRD patterns (Supporting Information Figure S2) did not reveal the presence of gold diffraction peaks, suggesting thus a high dispersion of metal NPs into the obtained systems. No substantial difference in CuxO peak positions after gold deposition could be detected, indicating that RF-sputtering of gold did not result in any significant alteration of phase composition. In line with this observation, the mean crystallite sizes were the same as those previously reported (30 and 20 nm for Cu2O and CuO, respectively). Nevertheless, a major difference with respect to Au-free samples was a systematic intensity reduction of all diffraction peaks (at least by 30%) after RFsputtering of metal particles. As indicated by FE-SEM (see

below), such an effect, accompanied by a reduction of the overall deposit thickness of ≈30%, could be related to the plasmochemical treatment of the obtained specimens, in spite of the mild conditions adopted for gold deposition. In fact, bombardment and compaction phenomena of CuxO
TiO2 might occur, resulting, in turn, in an enhanced structural disorder, that is, in a decrease of the overall diffracted intensity. Similar effects have been reported regarding plasma exposure of CVD ZnO and sol
gel lanthanum oxide systems,51,52 even under conditions softer than the present ones. The above observations are supported by the analysis of FE-SEM pictures for Au-containing samples reported in Figure 6. With regard to Cu2O
TiO2
Au (sample # N°4), significant variations occurred upon plasma exposure, leading to a compact morphology. In fact, the lateral nanopyramid size and overall thickness were reduced to 90 and 100 nm, respectively (compare samples # N°4, Figure 6 and # N°2, Figure 3). In a different way, the morphology of CuO
TiO2
Au systems was reminiscent of the pristine gold-free ones. For sample # N°10 (Figure 6), the nanoflower lateral dimensions and thickness were evaluated to be 350 and 470 nm, respectively. High magnification micrographs (see insets) displayed a uniform decoration by low-sized (≈6 nm) gold nanoparticles, whose dispersion and distribution were further investigated by TEM analyses (see below and Figure 8). Important information on in-depth elemental distribution was provided by XPS depth profiling (Figure 5, bottom; samples # N°4 and 10). With respect to specimens # N°2 and 8 (Figure 5, top), a main difference was the lower sputtering time necessary to observe a net increase of the Si substrate signal. This feature was in agreement with the above-discussed plasmochemical modifications, producing, in turn, an overall thickness decrease. Despite this alteration, TiO2 in-depth dispersion was still relatively similar to that observed for the homologous Au-free specimen. 6414

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir

ARTICLE

Figure 6. Plane-view (left) and cross-sectional (right) FE-SEM micrographs of CuxO
TiO2
Au nanomaterials deposited adopting the conditions of Table 1.

Figure 8. (a) HAADF-STEM image of a CuxO
TiO2
Au specimen (# N°10). (b) HR-TEM image of Au nanoparticles. The particle on the right top of the micrograph is imaged along the [011] zone axis orientation, as evidenced by the fast Fourier transform (FFT) in the inset. The size distribution of Au NPs is also displayed.

Figure 7. Cross-sectional TEM images of representative CuxO
TiO2
Au nanomaterials. The ED patterns (right) are the superimposition of the Si(100) substrate and polycrystalline CuxO systems.

For the Cu2O-based sample (# N°4, overall thickness = 100 nm), the presence of gold and TiO2 was essentially confined in external ≈5 and 10 nm regions, respectively, due to the relatively compact system morphology (see also Supporting Information Figure S3, bottom). In a different way, the CuO-containing specimen (# N°10) was characterized by an even Au distribution throughout the investigated depth, with an average content of ≈1%. As already noted, this result was mainly due to the synergy between the copper(II) oxide matrix porosity and the infiltration power characterizing plasmochemical techniques, whose concurrence enables to obtain a more uniform system composition. In fact, in this case, no external accumulation regions for Au or TiO2 could be clearly observed. To attain a deeper insight into the system nanostructure, with particular regard to the spatial distribution of gold NPs, a TEM study was performed on selected representative samples. Figure 7 displays low magnification bright-field TEM images and corresponding ED patterns of CuxO
TiO2
Au nanomaterials. As a matter of fact, the polycrystalline system nature was confirmed and the obtained indications were in line with those provided by FE-SEM observations (Figure 6). As already noted, the main difference was dictated by the nature of the pristine copper oxide system. Cu2O-based specimens (see # N°4, Figure 7, top)

exhibited a rather compact structure, with lateral aggregate sizes of ≈70 nm and overall height of ≈100 nm. The corresponding ED pattern was consistent with the presence of Cu2O as the main phase.6 In agreement with the GIXRD data, neither TiO2 nor Au signals were detected due to their moderate amount.9 The CuO-based systems (specimen # N°10, Figure 7, bottom) were characterized by more randomly shaped nanosized grains. The system appeared to be highly porous and exhibited appreciable topography and thickness variations. The pertaining ED pattern, indexed according to the monoclinic CuO structure,6 was characterized by rings with diffuse halos, suggesting a higher structural disorder/smaller grain size with respect to the previous case. It should be noted that no unambiguous structural evidence of TiO2 or/and Au was detected by ED or conventional TEM. To confirm the presence of all components within CuxO
TiO2
Au materials and shed light on their spatial distribution, sample # N°10 was characterized by advanced transmission electron microscopy techniques. Figure 8a is an HAADF-STEM (Z-contrast) image, showing the presence of finely dispersed gold nanoparticles into a CuO deposit approximately 350 nm thick. As the Z-contrast image is mass-thickness sensitive, the “heavy” gold nanoparticles appear as bright white spots in the image. A careful micrograph inspection enabled to corroborate the in-depth dispersion of Au nanoaggregates, as already revealed by XPS depth profiling (Figure 5). In order to confirm that the observed NPs were indeed composed by gold, HR-TEM measurements were carried out and a representative image is shown in Figure 8b. Most of the Au NPs are defect-free and almost spherical in shape; however, some 6415

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir defective particles, showing twinning or stacking faults, are present. The low gold nanoaggregate density, demonstrated by TEM performed at different positions, further evidenced the remarkable degree of metal NP dispersion, in good agreement with the absence of Au reflections in GIXRD and ED patterns. The NPs correspond to crystalline gold with a face-centered cubic (fcc) structure. The pertaining particle size distribution is displayed in Figure 8b, inset. An average diameter of 6.7 nm with a distribution width of 1.6 nm was obtained, in line with FE-SEM analysis.

’ CONCLUSIONS This study was devoted to the development of a novel synthetic approach for the preparation of multifunctional CuxO
TiO2 and CuxO
TiO2
Au (x = 1, 2) nanosystems. The proposed strategy consisted in the initial CVD of copper (I, II) oxides on Si(100), followed by the deposition of titanium oxide with different loadings. Optimization of the preparation process enabled the synthesis of single-phase CuxO nanostructures, whose porosity, in synergy with the CVD infiltration power, led to a tailored dispersion of titania nanoparticles into the pristine matrix. The use of mild conditions enabled to avoid the formation of Cu
Ti
O ternary phases. Key parameters to control the system characteristics were the TiO2 deposition time and the CuxO matrix nature, which also influenced CuxO
TiO2 interfacial interactions. The functionalization with Au nanoparticles via RF-sputtering led to nanomaterials with a controllable dispersion of metal nanoaggregates and an intimate contact between the three components. In summary, these characteristics are extremely promising in view of possible applications exploiting the characteristics of p
n junctions at the CuxO
TiO2 interface and the peculiar reactivity of gold nanoparticles. In particular, the most interesting perspectives for advancement of the present research activities will concern the functional validation of the obtained system in photoactivated H2 generation by reforming processes, as well as in gas sensing of toxic/flammable pollutants. Their use as innovative electrodes in Li-ion batteries exploiting the high contact area with the electrolyte will also be the subject of future investigation. ’ ASSOCIATED CONTENT

bS

Supporting Information. Representative XPS, GIXRD, and AFM data for the obtained specimens. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT This work was supported by funding from the European Community’s Seventh Framework Programme (FP7/20072013) under grant agreement number ENHANCE-238409. PRIN-COFIN 2008 “Organization of molecular architectures on functional surfaces and nanosystems”, Padova University PRAT 2008 “Nano-organization of functional molecular architectures on inorganic surfaces for eco-sustainable processes” programs, and the Fund for Scientific Research Flanders

ARTICLE

(FWO) are also acknowledged for financial assistance. Thanks are due to Mr. Antonio Ravazzolo for skillful technical support.

’ REFERENCES (1) Mane, S. R.; Lee, W. J.; Pathan, H. M.; Han, S. H. J. Phys. Chem. B 2005, 109, 24254. (2) Limmer, S. J.; Chou, T. P.; Cao, G. J. Phys. Chem. B 2003, 107, 13313. (3) Nie, Z.; Petukhova, A.; Kumacheva, E. Nat. Nanotechnol. 2010, 5, 15. (4) Shifu, C.; Sujuan, Z.; Wei, L.; Wei, Z. J. Nanosci. Nanotechnol. 2009, 9, 4397. (5) Armelao, L.; Barreca, D.; Bertapelle, M.; Bottaro, G.; Sada, C.; Tondello, E. Thin Solid Films 2003, 442, 48. (6) Barreca, D.; Gasparotto, A.; Maccato, C.; Tondello, E.; Lebedev, O. I.; Van Tendeloo, G. Cryst. Growth Des. 2009, 9, 2470. (7) Zhang, Y. G.; Ma, L. L.; Li, J. L.; Yu, Y. Environ. Sci. Technol. 2007, 41, 6264. (8) Jun, J.; Jin, C.; Kim, H.; Park, S.; Lee, C. Appl. Surf. Sci. 2009, 255, 8544. (9) Su, X.; Zhao, J.; Li, Y.; Zhu, Y.; Ma, X.; Sun, F.; Wang, Z. Colloids Surf., A 2009, 349, 151. (10) Vigil, E.; Gonzalez, B.; Zumeta, I.; Domingo, C.; Dom
enech, X.; Ayll
on, J. A. Thin Solid Films 2005, 489, 50. (11) Tung, H. -T.; Song, J. -M.; Dong, T. -Y.; Hwang, W. -S.; Chen, I. -G. Cryst. Growth Des. 2008, 8, 3415. (12) Li, G.; Dimitrijevic, N. M.; Chen, L.; Rajh, T.; Gray, K. A. J. Phys. Chem. C 2008, 112, 19040. (13) Bessekhoud, Y.; Robert, D.; Weber, J. V. Catal. Today 2005, 101, 315. (14) Bandara, J.; Udawatta, C. P. K.; Rajapakse, C. S. K. Photochem. Photobiol. Sci. 2005, 4, 857. (15) Teleki, A.; Bjelobrk, N.; Pratsinis, S. E. Sens. Actuators, B 2008, 130, 449. (16) Patil, S. A.; Patil, L. A.; Patil, D. R.; Jain, G. H.; Wagh, M. S. Sens. Actuators, B 2007, 123, 233. (17) Ruiz, A. M.; Cornet, A.; Shimanoe, K.; Morante, J. R.; Yamazoe, N. Sens. Actuators, B 2005, 109, 7. (18) Yates, H. M.; Brook, L. A.; Ditta, I. B.; Evans, P.; Foster, H. A.; Sheel, D. W.; Steele, A. J. Photochem. Photobiol., A 2008, 197, 197. (19) Xu, S.; Sun, D. D. Int. J. Hydrogen Energy 2009, 34, 6096. (20) Wu, J. C. S.; Tseng, I. H.; Chang, W. -C. J. Nanopart. Res. 2001, 3, 113. (21) Kim, K. -H.; Ihm, S. -K. J. Hazard. Mater. 2007, 146, 610. (22) Col
on, G.; Maicu, M.; Hidalgo, M. C.; Navío, J. A. Appl. Catal., B 2006, 67, 41. (23) Xin, B.; Wang, P.; Ding, D.; Liu, J.; Ren, Z.; Fu, H. Appl. Surf. Sci. 2008, 254, 2569. (24) L
opez, R.; G
omez, R.; Llanos, M. E. Catal. Today 2009, 148, 103. (25) Ryu, S. W.; Kim, E. J.; Ko, S. K.; Hahn, S. H. Mater. Lett. 2004, 58, 582. (26) Hsiung, T. L.; Wang, H. P.; Lu, Y. -M.; Hsiao, M. C. Radiat. Phys. Chem. 2006, 75, 2054. (27) Ditta, I. B.; Steele, A.; Liptrot, C.; Tobin, J.; Tyler, H.; Yates, H. M.; Sheel, D. W.; Foster, H. A. Appl. Microbiol. Biotechnol. 2008, 79, 127. (28) Mungkalasiri, J.; Bedel, L.; Emieux, F.; Dor
e, J.; Renaud, F. N. R.; Maury, F. Surf. Coat. Technol. 2009, 204, 887. (29) Barreca, D.; Battiston, G. A.; Casellato, U.; Gerbasi, R.; Tondello, E. J. Phys. IV 2001, 11, Pr11–253. (30) Tripathi, M.; Pandey, K.; Kumar, S. D. Sol. Energy Mater. Sol. Cells 2007, 91, 1663. (31) Lu, F. -H.; Fang, F. -X.; Chen, Y. -S. J. Eur. Ceram. Soc. 2001, 21, 1093. 6416

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417

Langmuir

ARTICLE

(32) Bandoli, G.; Barreca, D.; Gasparotto, A.; Seraglia, R.; Tondello, E.; Devi, A.; Fischer, R. A.; Winter, M.; Fois, E.; Gamba, A.; Tabacchi, G. Phys. Chem. Chem. Phys. 2009, 11, 5998. (33) Fois, E.; Tabacchi, G.; Barreca, D.; Gasparotto, A.; Tondello, E. Angew. Chem., Int. Ed. 2010, 49, 1944. (34) Gasparotto, A.; Barreca, D.; Devi, A.; Fischer, R. A.; Fois, E.; Gamba, A.; Maccato, C.; Seraglia, R.; Tabacchi, G.; Tondello, E. ECS Trans. 2009, 25, 549. (35) Barreca, D.; Comini, E.; Ferrucci, A.; Gasparotto, A.; Maccato, C.; Maragno, C.; Sberveglieri, G.; Tondello, E. Chem. Mater. 2007, 19, 5642 and references therein.. (36) Chiarello, G. L.; Selli, E.; Forni, L. Appl. Catal., B 2008, 84, 332. (37) Sakthivel, S.; Shankar, M. V.; Cherepy, N. J.; Smestad, G. P.; Graetzel, M.; Zhang, J. Z. J. Phys. Chem. B 1997, 101, 9342. (38) Gou, X.; Wang, G.; Yang, J.; Park, J.; Wexler, D. J. Mater. Chem. 2008, 18, 965. (39) Sangeetha, P.; Zhao, B.; Chen, Y. -W. Ind. Eng. Chem. Res. 2010, 49, 2096. (40) Barreca, D.; Gasparotto, A.; Maragno, C.; Tondello, E.; Sada, C. Chem. Vap. Deposition 2004, 10, 229. (41) Barreca, D.; Gasparotto, A.; Maragno, C.; Tondello, E.; Bontempi, E.; Depero, L. E.; Sada, C. Chem. Vap. Deposition 2005, 11, 426. (42) Barreca, D.; Gasparotto, A.; Lebedev, O. I.; Maccato, C.; Pozza, A.; Tondello, E.; Turner, S.; Van Tendeloo, G. CrystEngComm 2010, 12, 2185. (43) Barreca, D.; Gasparotto, A.; Tondello, E.; Sada, C.; Polizzi, S.; Benedetti, A. Chem. Vap. Deposition 2003, 9, 199. (44) Briggs, D.; Seah, M. P. Practical Surface Analysis, 2nd ed.; John Wiley and Sons: New York, 1990. (45) Wu, J. C. S.; Lin, H. M.; Lai, C. L. Appl. Catal., A 2005, 296, 194. (46) Zhang, Y. C.; Tang, J. Y.; Wang, G. L.; Zhang, M.; Hu., X. Y. J. Cryst. Growth 2006, 294, 278. (47) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer: Eden Praire, MN, 1992. (48) Pattern No. 5-667. Joint Committee on Powder Diffraction Standards, 2000. (49) Zhu, B.; Zhang, X.; Wang, S.; Zhang, S.; Wu, S.; Huang, W. Microporous Mesoporous Mater. 2007, 102, 333. (50) Pattern No. 45-937. Joint Committee on Powder Diffraction Standards, 2000. (51) Armelao, L.; Pascolini, M.; Bottaro, G.; Bruno, G.; Giangregorio, M. M.; Losurdo, M.; Malandrino, G.; Lo Nigro, R.; Fragala, M. E.; Tondello, E. J. Phys. Chem. C 2009, 113, 2911. (52) Losurdo, M.; Giangregorio, M. M.; Sacchetti, A.; Capezzuto, P.; Bruno, G.; Malandrino, G.; Fragala, I. L. J. Mater. Res. 2006, 21, 1632.

6417

dx.doi.org/10.1021/la200698t |Langmuir 2011, 27, 6409–6417