DOI: 10.1021/cg1002012
Highly Oriented ZnO Nanorod Arrays by a Novel Plasma Chemical Vapor Deposition Process
2010, Vol. 10 2011–2018
)
)
Daniela Bekermann,† Alberto Gasparotto,*,‡ Davide Barreca,§ Laura Bovo,‡ Anjana Devi,† Roland A. Fischer,† Oleg I. Lebedev,#, Chiara Maccato,‡ Eugenio Tondello,‡ and Gustaaf Van Tendeloo
)
† Inorganic Materials Chemistry Group, Lehrstuhl f€ ur Anorganische Chemie II, Ruhr-University Bochum, 44780 Bochum, Germany, ‡Department of Chemistry, Padova University and INSTM, 35131 § Padova, Italy, CNR-ISTM and INSTM, Department of Chemistry, Padova University, 35131 Padova, Italy, #Laboratoire CRISMAT, ENSICAEN, UMR 6508 CNRS, 14050 Caen Cedex, France, and Electron Microscopy for Materials Science (EMAT), University of Antwerp, 2020 Antwerpen, Belgium
Received February 8, 2010; Revised Manuscript Received March 8, 2010
ABSTRACT: Strongly c-axis oriented ZnO nanorod arrays were grown on Si(100) by plasma enhanced-chemical vapor deposition (PE-CVD) starting from two volatile bis(ketoiminato) zinc(II) compounds Zn[(R0 )NC(CH3)dC(H)C(CH3)dO]2, with R0 = -(CH2)xOCH3 (x = 2, 3). A systematic investigation of process parameters enabled us to obtain the selective formation of ZnO nanorods with tailored features, and provided an important insight into their growth mechanism. The morphology, structure, and composition of the synthesized ZnO nanosystems were thoroughly analyzed by field emission-scanning electron microscopy (FESEM), energy dispersive X-ray spectroscopy (EDXS), glancing incidence X-ray diffraction (GIXRD), X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy (TEM). Photoluminescence (PL) measurements were carried out to gain information on the optical properties. Specifically, one-dimensional (1D) ZnO architectures could be grown on Si(100) substrates at temperatures as low as 200-300 C and radio frequency (RF)-power values of 20 W, provided that a sufficiently high mass supply to the growth surface was maintained. To the best of our knowledge, the present work reports the mildest preparation conditions ever appeared in the literature for the PE-CVD of ZnO nanorods, a key result in view of potential large-scale technological applications.
*Corresponding author: Tel: þ39-0498275192. Fax: þ39-049-8275161. E-mail:
[email protected].
fundamental phenomena in low-dimensional structures, but also for the implementation of innovative multifunctional materials.6,13,19-22 To date, various preparation routes for the synthesis of ZnObased systems have been reported, including hydrothermal methods, sol-gel, electrochemical processes, spray pyrolysis, sputtering, pulsed laser deposition, evaporation, molecular beam epitaxy, and chemical vapor deposition (CVD).4-6,12,21,23-26 Among CVD methodologies, plasma enhanced-chemical vapor deposition (PE-CVD) is an extremely promising strategy for the integration of supported micro- and nanosystems into functional devices. Besides the possibility to achieve a conformal deposition even on complex-shaped surfaces,27,28 PE-CVD enables the tailored synthesis of nanosystems even at low temperatures on thermally labile substrates, maintaining appreciable growth rates.4,21,23,29-32 In addition, a major advantage is the possibility of avoiding the use of catalytic nanoparticles as promoters of the 1D growth, resulting in simplified processes and preventing the incorporation of metal impurities.20,21,33,34 Finally, it is worthwhile highlighting that electric field effects and plasma bombardment during PE-CVD processes can be beneficial for both system crystallinity and the oriented growth of anisotropic zinc oxide nanostructures.20,24,27,35,36 In this context, the present work is focused on the synthesis of highly oriented ZnO nanorod arrays by PE-CVD. Even though several works have been published on the CVD of ZnO, the majority of them have been focused on thermalCVD. Conversely, to the best of our knowledge, only very few papers on the PE-CVD of ZnO nanorods are available in the literature.20,34,36,37 The chemical control of CVD and PE-CVD processes relies on the design of novel molecular precursors with improved
r 2010 American Chemical Society
Published on Web 03/18/2010
1. Introduction ZnO, one of the most investigated inorganic materials, has been the subject of over 4000 published papers in the 2008 scientific literature.1 The main reasons fuelling this widespread interest can be traced back to the versatile and promising properties of this semiconducting oxide, featuring a wide direct band gap (Eg = 3.4 eV) and a large exciton binding energy (60 meV), together with excellent optical and electrical characteristics. In addition, ZnO is chemically and thermally stable, as well as biosafe and environment friendly.2-11 Thanks to its hexagonal wurtzite crystal structure, zinc oxide can be grown in an extremely wide range of morphologies, such as nanowires and nanotubes, nanoribbons, nanocombs, nanoplatelets, nanohelices, nanosprings, tetrapods,2,4-14 and in the form of complex hierarchical architectures.3,6,15-17 In addition, chemical modifications of ZnO, for example, by anion/cation doping, molecular grafting, or the development of composites, enable zinc oxide functionalities to be enlarged over a virtually infinite range of possibilities.2,7,18 The synthetic control of zinc oxide chemical and physical properties has indeed enabled the development of several ZnO-based systems for a variety of advanced applications, including lasers, displays, gas- and biosensors, photovoltaics, photocatalysis, and field-emission, piezoelectric, and surface acoustic wave devices.2,8,15 Despite the high number of publications and patents on ZnO, the academic and applicative interest in this material is still far from being completely satisfied. In particular, the rational design of one-dimensional (1D) ZnO nanosystems plays a strategic role not only for the understanding of
pubs.acs.org/crystal
2012
Crystal Growth & Design, Vol. 10, No. 4, 2010
Bekermann et al.
Table 1. Main Process Parameters Adopted for ZnO PE-CVD from Precursor 1 and Corresponding Sample Thickness (uncertainty = ( 10 nm)a sample
deposition temperature (C)
RF-power (W)
thickness (nm)
ZnO1 ZnO2 ZnO3 ZnO4
100 200 300 400
40
65 90 105 150
ZnO5 ZnO6 ZnO7 ZnO8
100 200 300 400
20
70 250 405 330
ZnO9
300
10
130
a
In all cases, the O2, the auxiliary Ar, and the carrier Ar flow rates were set at 20, 15, and 60 sccm, respectively. Total pressure = 1.0 mbar; precursor vaporization temperature = 140 C.
mass-transport properties and clean decomposition pathways, enabling the formation of the target systems under soft and controlled conditions. Even though some novel zinc precursors have recently been reported,4,14,22,25,38 most of CVD routes to ZnO have involved so far the use of zinc dialkyls [e.g., diethyl(DEZ) or dimethyl-zinc (DMZ)] and conventional zinc β-diketonates [such as Zn(acac)2, with acac = 2,4-pentanedionate or Zn(dpm)2, with dpm = 2,2,6,6-tetramethyl-3,5-heptanedionate].19,28,29,33,35 In particular, PE-CVD of ZnO nanorods has been usually performed starting from DEZ.20,34,36,37 Nevertheless, these compounds are pyrophoric and extremely reactive and/or possess a narrow temperature window between vaporization and decomposition.22,25,26 As a consequence, CVD and PE-CVD of ZnO from newly designed precursors still deserve further investigation. To this regard, a key point of the present work is the use of the recently reported bis(ketoiminato) zinc(II) compounds Zn[(R0 )NC(CH3)dC(H)C(CH3)dO]2, with R0 = -(CH2)2OCH3 (1) or -(CH2)3OCH3 (2),39 adopted for the first time in the PE-CVD of ZnO nanorod arrays. In this context, the interest has been devoted to softening the PE-CVD conditions, generally harsh in terms of deposition temperature and power density.20,34,36,37 Special attention has been paid to the interplay between the adopted process parameters and the structural and morphological properties of the resulting zinc oxide nanosystems. 2. Experimental Section ZnO depositions were performed by a two-electrode custombuilt PE-CVD apparatus.40 p-Type Si(100) substrates (MEMC, Merano, Italy) were cleaned, without removing the native oxide layer, by iterative dipping in sulfonic detergent and distilled water, distilled water, acetone, and isopropylic alcohol.41 Electronic grade argon and oxygen were used as plasma sources. The precursor was placed in an external vessel heated by an oil bath and transported toward the deposition zone by an Ar flow. Two further auxiliary gas-lines were used to introduce Ar and O2 gases directly into the reactor. To avoid undesired condensation phenomena, the gas lines connecting the precursor vessel and reaction chamber were heated at 160 C. For all experiments, the deposition time and interelectrode distance were set at 1 h and 6 cm, respectively. Further details concerning the synthesis conditions are summarized in Tables 1 and 2. Field emission-scanning electron microscopy (FE-SEM) measurements were performed at primary beam acceleration voltages between 10 and 20 kV by a Zeiss SUPRA 40 VP instrument, equipped with an Oxford INCA x-sight X-ray detector for energy dispersive X-ray spectroscopy (EDXS) analyses.
Glancing incidence X-ray diffraction (GIXRD) patterns were recorded by means of a Bruker D8 Advance diffractometer equipped with a G€ obel mirror and a CuKR source (40 kV, 40 mA), at a fixed incidence angle of 1.0. X-ray photoelectron spectroscopy (XPS) measurements were run on a Perkin-Elmer Φ 5600ci spectrometer, using a nonmonochromatized MgKR excitation source (1253.6 eV), at working pressures lower than 10-9 mbar. Binding energy (BE) correction was performed assigning to the C1s signal of adventitious carbon a value of 284.8 eV. Arþ sputtering treatments were carried out at 3.0 kV, with an argon partial pressure of 5 10-8 mbar. Peak fitting was performed by a least-squares procedure, adopting GaussianLorentzian peak shapes. Transmission electron microscopy (TEM) analyses were carried out using a JEOL 40000EX microscope operated at 400 kV and having a 0.17 nm point resolution. Samples for cross-section observations were prepared by standard focused ion beam (FIB) techniques,42 using a FEI Nova 200 Nanolab DualBeam SEM/FIB system. A dual-cap C-Pt protection layer was used to protect ZnO nanorod arrays from ion milling damage.43 Room temperature photoluminescence (PL) measurements were performed by using a Xenon UV-lamp (450 W, Edimburgh Instruments Ltd.) with an excitation wavelength of 300 nm. Spectra were recorded using an Horiba Jobin Yvon T64000 triple monochromator coupled with an intensified charge-coupled device (CCD; Andor iStar).
3. Results and Discussion The structure of the two adopted bis(ketoiminato) zinc(II) compounds is reported in Figure 1. For precursor 1, an initial optimization of the instrumental setup and process parameters was accomplished to ensure an efficient mass transport from the vapor phase toward the deposition zone and to obtain a homogeneous substrate coverage. The resulting optimized operational conditions are summarized in Table 1. One of the main aims of the present work was the synthesis of ZnO nanorods with tailored morphological and structural properties under soft PE-CVD conditions. Accordingly, in the case of precursor 1, the combined influence exerted by the growth temperature and the applied RF-power, both related to the energy supply during the deposition process, was investigated. In this regard, Figure 2 provides a comparative FE-SEM morphology overview for samples synthesized from precursor 1 (see Table 1). As can be observed, the use of high RF-power values (40 W samples, ZnO1-ZnO4) resulted in the obtainment of a relatively compact system morphology. In particular, variations of the substrate temperature from 100 to 400 C produced a thickness increase from ≈60 nm to ≈150 nm. While the ZnO1 sample showed the presence of pseudoglobular particles typical for an isotropic growth, the development of a columnar morphology was evidenced for specimens ZnO2 and ZnO3. In particular, the 300 C sample (ZnO3) presented the most pronounced anisotropy, with the formation of some faceted tips on the column surface. Conversely, at 400 C (ZnO4), a compact ZnO film characterized by a rather smooth surface was obtained. Even though at 20 W (samples ZnO5-ZnO8) a qualitatively similar evolution was observed as a function of the substrate temperature, the system morphology was systematically characterized by a more pronounced 1D growth with respect to specimens synthesized at 40 W (ZnO1-ZnO4). This difference could be traced back to an enhanced energy supply to the substrate surface at 40 W, leading to more marked coalescence phenomena between adjacent nanoaggregates and to an increased lateral growth.24,34 Regarding samples synthesized at 20 W, the most interesting results were obtained at intermediate temperatures, that is,
Article
Crystal Growth & Design, Vol. 10, No. 4, 2010
2013
Table 2. Main Process Parameters Adopted for ZnO PE-CVD from Precursor 2 (RF-power = 20 W)a
sample
deposition temperature (C)
total pressure (mbar)
O2 flow rate (sccm)
auxiliary Ar flow rate (sccm)
carrier Ar flow rate (sccm)
precursor vaporization temperature (C)
ZnO10 ZnO11 ZnO12 ZnO13
100 200 300 400
1.0
20
15
60
140
110 355 505 410
ZnO14 ZnO15
300
0.5
8
-
30
140 150
100 220
a
thickness (nm)
In the last column, the overall deposit thickness (uncertainty = ( 10 nm) is reported.
Figure 1. Molecular structure of the bis(ketoiminato) zinc(II) precursors adopted in the present work.
for ZnO6 (200 C) and ZnO7 (300 C). In fact, under these conditions ordered arrays of nanorods aligned perpendicularly to the substrate surface were grown, with an aspect ratio (length/diameter) increasing from 9.5 at 200 C to 12.4 at 300 C. In the case of sample ZnO7, having the more pronounced 1D morphology, the density of these nanostructures per unit area, estimated by plane-view images, was ≈950 nanorods/μm2. An increase of the growth temperature up to 400 C (sample ZnO8) resulted in the formation of a relatively compact cross-section texture, although the surface appeared rougher than in the case of ZnO4. This phenomenon suggested that, at 400 C, a reduction of the applied RF-power produced a progressive transformation from a compact film to a columnar one,24 accompanied by an enhancement of the overall system thickness, that is, of the growth rate. A similar effect indicated that the use of more drastic RF-power conditions (40 W) favored etching phenomena over the competitive deposition ones,35 whereas the use of a lower RF-power (20 W) had a beneficial influence on the growth rate and also induced an improvement of the anisotropic 1D morphology. In order to further ascertain the latter effect, PE-CVD of ZnO from precursor 1 was also undertaken at 300 C and 10 W (ZnO9). Under these conditions columnar structures were still obtained, but the deposit thickness (130 nm) and aspect ratio (4.9) were appreciably lower than in the case of sample ZnO7 obtained at the same temperature, but at RF-power = 20 W. The above results, in line with previous literature works,23,33,35,37 highlight the need for intermediate growth temperatures and RF-power values to obtain anisotropic ZnO nanorod arrays. GIXRD analyses on specimens ZnO1-ZnO9 (Figure 3) evidenced the sole reflections expected for the hexagonal ZnO wurtzite structure (zincite; JCPDS card no. 36-1451). Nevertheless, both the relative peak amplitudes and the total diffracted intensity were significantly influenced by the synthesis conditions. At 100 C, the recorded patterns were characterized by four well-evident reflections at 2ϑ = 31.8, 34.4, 36.3, and 47.5,
attributed respectively to ZnO (100), (002), (101), and (102) reflections (JCPDS card no. 36-1451). The relative intensities of the observed peaks suggested the absence of marked preferential orientation effects. This result, in line with the observed globular morphology of both ZnO1 and ZnO5 samples (compare Figure 2), confirmed the occurrence of an isotropic crystallite growth.7,26 In a different way, for ZnO nanodeposits synthesized in the 200-400 C range (Figure 3), a strong c-axis preferential orientation was always observed, as indicated by the remarkable intensity of the (002) peak with respect to the other reflections. The total diffracted intensity for samples synthesized at the same temperature was always higher at 20 W, in agreement with the increased deposit thickness at the lowest adopted RF-power (see also Figure 9). It is worthwhile observing that for specimen ZnO7, characterized by the most anisotropic 1D architecture, the GIXRD pattern was dominated by the (002) reflection. This result, highlighting the occurrence of a direct relationship between the sample structure and morphology, indicated that the observed ZnO nanorods were characterized by a Æ001æ growth direction perpendicular to the substrate surface.3,5,8,20,34,44-46 The chemical composition of ZnO samples was investigated by the combined use of XPS and EDXS. Specifically, XPS surface analyses evidenced the expected zinc and oxygen signals (Figure 4a), whose spectral features were in agreement with ZnO formation, irrespective of the adopted processing parameters.14,22,47 In addition, despite carbon peaks were detected, their rapid disappearance upon Arþ erosion (4 min) indicated that C presence was merely related to adventitious surface contamination.22,28,48 It is also worthwhile observing that N1s signal was never revealed by XPS investigations. As a whole, these results indicated a clean ketoiminate precursor decomposition under the adopted experimental conditions. It is worth noting that both XPS and EDXS analyses confirmed the presence of a slight oxygen excess on the sample outermost layers, that could be mainly related to the presence of -OH groups (≈30% of the total XPS O1s peak) arising from atmospheric exposure (Figure 4b,c).14,22,47,49 Such a phenomenon, frequently observed in the case of oxide-based nanosystems, was reasonably enhanced in the present case due to the high surface-to-volume ratio of the obtained ZnO nanorod arrays. In order to achieve a deeper insight into the structure and morphology of the synthesized ZnO nanosystems, selected specimens were subjected to TEM and high-resolution TEM (HRTEM) analyses. Figure 5a,b display the low magnification cross-section structure of samples ZnO6 and ZnO7. The observed topological features were in excellent agreement with results provided by FE-SEM investigation (compare Figure 2) and confirmed the formation of uniform nanorod arrays
2014
Crystal Growth & Design, Vol. 10, No. 4, 2010
Bekermann et al.
Figure 2. Cross-section FE-SEM micrographs of Si(100)-supported ZnO nanodeposits synthesized from precursor 1 as a function of the adopted RF-power and growth temperature (see Table 1). For sample ZnO7, representative plane-view and high magnification cross-section images are also reported.
Figure 3. GIXRD spectra for samples ZnO1-ZnO8 deposited on Si(100) from precursor 1.
oriented perpendicularly to the substrate surface, with aspect ratio values directly dependent on the synthesis conditions. The obtainment of ZnO 1D growth under optimized synthetic conditions is usually related to the non-centrosymmetric hexagonal wurtzite structure, and, in particular, to the divergence of surface energies, resulting in diversified growth rates as a function of crystallographic directions.5,6,14,19,46 Nevertheless, the sole occurrence of an anisotropic growth regime is not sufficient to ensure the alignment of the resulting nanorods perpendicular to the substrate surface. To this regard, an epitaxial relationship with the Si(100) support cannot explain the obtained results, since a native SiO2 layer ≈ 2 nm thick was always present at the ZnO/substrate interface (see Figure 5c).20,22,44 In the present case, a reason-
Figure 4. Representative XPS survey spectrum (a) and O1s surface photopeak (b) for sample ZnO7. (c) Corresponding EDXS profiles for OKR1 (red) and ZnKR1 (blue) X-ray signals along the crosssection of the same specimen.
able justification for the highly oriented growth of ZnO nanorods relies on the intrinsic polarization along the c-axis of the wurtzite structure, favoring the rod alignment in the direction of the applied electric field, that is, normal to the substrate surface.6,17,20,46
Article
Figure 5. Cross-section low magnification TEM images of samples (a) ZnO6 and (b) ZnO7. (c) Representative HRTEM image of the ZnO/substrate interface for specimen ZnO6.
Figure 6. Cross-section HRTEM images of single ZnO nanorods in samples (a) ZnO6 and (b) ZnO7, obtained from precursor 1. White arrows mark stacking faults (SF) along the Æ001æ growth direction. The areas highlighted by white rectangles are enlarged at the bottom of the figure to show different SF types in the two samples. In (b), two adjacent nanorods grown with a similar Æ001æ orientation are imaged.
HRTEM analyses confirmed that ZnO nanorods always grew along the Æ001æ direction, that is, with the c-axis mainly perpendicular to the substrate surface (see Figure 6). Although this effect was common to all 1D ZnO nanodeposits, other structural features were found to be directly dependent on the specific synthesis conditions. In this regard, Figure 6 shows some important differences between samples ZnO6 (200 C) and ZnO7 (300 C) obtained at 20 W RF-power. As can be observed, stacking faults (SF) were the main type of defects in both cases, but their density was appreciably higher for the former sample, that is, at 200 C. This phenomenon suggested that an increase of only 100 C in the growth temperature resulted in a more efficient ZnO crystallization. Such an effect, also confirmed for samples synthesized from precursor 2, was responsible for the different optical properties of the corresponding ZnO nanorod arrays (see PL results below). The most common type of SF for all samples was found to be the one shown in Figure 6b. As the stacking sequence of 2H-ZnO structure can be represented as ABABAB along the c direction, this SF can be described as ABABACAC (I1 type).49 The displacement vector R can be determined as 1/3[100]. Only in sample ZnO6, besides the above quoted I1 SF, some ZnO nanorods exhibit a different type of SF sequence
Crystal Growth & Design, Vol. 10, No. 4, 2010
2015
- ABABCBACA (Figure 6a). This alternative stacking structure can be described either as a 4H-type structure inserted in 2H-ZnO (AB/ABCBA/CA), or as an I1-type SF with an additional two-layer insertion, that is, ABAB/CB/ACAC. In this case, the displacement vector R is 2/3[100]. On the basis of the results presented so far, the attention was devoted to the use of compound 2 as the PE-CVD precursor and to a comparative analysis of the resulting ZnO nanosystem properties. In this case, samples were synthesized adopting an applied RF-power of 20 W (Table 2), corresponding to the best conditions for the obtainment of ZnO nanorod arrays in the case of compound 1. In particular, the influence of the growth temperature on the system morphological evolution was investigated for samples ZnO10-ZnO13 (Figure 7). Once again, ZnO nanorods with the highest aspect ratio (13.0) and the mostly pronounced anisotropic 1D morphology were obtained at 300 C. Nevertheless, two important differences are worth being highlighted with respect to specimens ZnO5-ZnO8, obtained from precursor 1 at the same RF-power (20 W) and growth temperatures. First, in this case, even the sample obtained at 400 C (ZnO13) presented a well-evident 1D morphology, at variance with the homologous ZnO8 grown at the same temperature. In addition, under the same processing conditions, the growth rate from precursor 2 was appreciably higher than in the case of compound 1 (see also Figure 9), due to the higher volatility of the former.39 The obtainment of 1D architectures was favored not only by a proper choice of the deposition temperature and RFpower, but also by a high mass supply to the growth surface. In order to confirm this hypothesis, two more experiments were performed at 300 C and 20 W. First, the total pressure was lowered to 0.5 mbar by reducing the O2 and Ar flow rates, maintaining constant their ratio (sample ZnO14). Subsequently, under the same pressure and gas flow conditions of ZnO14, the precursor vaporization temperature was increased to 150 C in order to obtain a more efficient precursor supply and, hence, an enhanced growth rate (sample ZnO15). As can be observed, specimen ZnO14 was characterized by an appreciably lower thickness than ZnO12 (100 vs 505 nm), mainly due to a less efficient precursor delivery,29 and by a less evident 1D growth habit. In line with the above considerations on mass transport phenomena, the increase of the precursor vaporization temperature from 140 to 150 C (sample ZnO15) resulted not only in a thickness increase (220 nm), but also in the development of a more pronounced 1D morphology (aspect ratio ≈ 8.4). For samples ZnO12 and ZnO15 (Figure 7), the density of nanorods per unit area resulted ≈620 and ≈770 nanorods/μm2, respectively. In spite of these differences, it is worth highlighting that for samples ZnO12, ZnO14, and ZnO15 a very strong Æ001æ preferential orientation was always obtained (see Supporting Information), indicating that, besides the applied RF-power, the adopted growth temperature played an important role in the morphogenesis of 1D nanosystems. Even in the case of precursor 2, ZnO nanosystems possessing the most pronounced 1D morphology were investigated by TEM. Once again, the obtained results, that agreed to a good extent with FE-SEM ones (Figure 7), revealed an appreciable density of stacking faults and a nanorod faceting directly dependent on the synthesis conditions. Figure 8 shows HRTEM images of ZnO nanorods for two most representative nanodeposits obtained from precursor 2 (ZnO11, 200 C and ZnO12, 300 C). Similar to the
2016
Crystal Growth & Design, Vol. 10, No. 4, 2010
Bekermann et al.
Figure 7. Representative cross-section FE-SEM micrographs of Si(100)-supported ZnO specimens synthesized from precursor 2 as a function of the adopted total pressure and growth temperature (see Table 2). For samples ZnO12 and ZnO15, representative plane-view images are also reported.
Figure 8. Cross-section HRTEM images of single ZnO nanorods in samples (a) ZnO11 and (b) ZnO12, obtained from precursor 2. In (a), white arrows mark stacking faults along the Æ001æ growth direction. Exposed surface planes are also labeled. Figure (b) shows the intergrowth of two adjacent nanorods (A and B). The presence of a stacking fault in nanorod B, that stops at the boundary with A (marked by a black arrow), is also highlighted by the white rectangle and enlarged at the bottom of the figure. The fast Fourier transform (FFT) along the [010] zone is reported in the inset.
homologous samples obtained from precursor 1 (compare Figure 6), even in this case the SF density was appreciably higher at 200 C. As evidenced by HRTEM imaging (Figure 8a), SFs perpendicular to the growth direction were clearly related to the nanorod faceting, the main exposed planes being {101} and {100}. The common growth direction led to a perfect heteroepitaxial intergrowth of adjacent nanorods (Figure 8b), with an atomically sharp boundary free from amorphous or intermediate layers. On the basis of the above results, a general growth mechanism can be proposed as a function of the adopted synthetic parameters. In this regard, Figure 9 summarizes the dependence of the growth rate on the deposition temperature for samples synthesized from compounds 1 and 2. Interestingly, irrespective of the used precursor, four growth regimes, corresponding to different morphological features, can be
identified in the graph. In particular, specimens grown at 100 C are characterized by an isotropic growth, that is, by the formation of pseudo-globular particles (compare Figures 2 and 7). Reasonably, at the lowest deposition temperature (100 C), the limited thermal energy supply was the main factor preventing the 1D growth, irrespective of the adopted RFpower (20 or 40 W). Upon increasing the deposition temperature to 200-300 C, two growth regimes can be observed, namely, a columnar (“quasi-1D”) and a 1D one, corresponding to the formation of ZnO nanorods, occurring at higher growth rates. Similar results have also been recently reported in the PE-CVD of TiO2 from Ti(IV) isopropoxide.50 As a whole, the present findings provide a useful guideline for the control of the ZnO system morphogenesis from a phenomenological point of view. In the absence of epitaxial relationships with the substrate, as in the present case, a possible explanation for the obtainment of highly oriented nanorods is as follows. At the beginning of the deposition process, the formation of small ZnO seeds takes place. At intermediate temperatures (200 or 300 C) and in the presence of plasma activation, such crystallites mainly grow with {001} planes parallel to the substrate surface since this orientation, characterized by a spontaneous polarization, is favored by the applied electric field. As the growth process proceeds, these ZnO nanoparticles act as self-catalysts for the formation of ZnO anisotropic systems5,6,20 and preferentially grow along the Æ001æ direction when a high mass supply to the substrate surface is provided, giving rise to the observed nanorod arrays.5,19,24 An increase of the deposition temperature up to 400 C led to the formation of ZnO films maintaining a strong Æ001æ preferential orientation but showing a progressive evolution from compact deposits to columnar ones upon increasing the growth rate (see Figure 9). Although higher growth rates still favored the formation of more anisotropic nanostructures, the increased thermal energy supply at 400 C resulted in an enhanced lateral growth, explaining thus the formation of relatively compact films rather than 1D (or “quasi-1D”) systems. Finally, the optical properties of two selected ZnO nanorod arrays obtained from precursor 2 and characterized by the most anisotropic 1D morphology (ZnO11 and ZnO12) were preliminarily investigated by PL measurements (Figure 10).
Article
Crystal Growth & Design, Vol. 10, No. 4, 2010
2017
Figure 9. Synoptic diagram showing the dependence of the ZnO morphology on PE-CVD growth parameters. Red and blue markers correspond to ZnO samples synthesized from precursors 1 and 2, respectively.
4. Conclusions
Figure 10. PL spectra of sample ZnO11 (200 C) and ZnO12 (300 C).
In the case of sample ZnO11, the PL spectrum presented a strong UV peak centered at 384 nm, displaying an appreciable tailing toward higher wavelengths. This signal was ascribed to the near band-edge emission of ZnO, originated from freeexciton recombination.2,5,8,20,45 On the other hand, the emission in the visible region could be traced back to the combined contribution of various kinds of defects for nanorods synthesized at 200 C, even though their exact nature can be hardly deduced from the present spectrum.22,48 In a different way, the PL spectrum of sample ZnO12 presented a narrower excitonic band centered at 374 nm. The blue-shift of the UV band with respect to specimen ZnO11 could be related to the higher surface-to-volume ratio of ZnO12 nanorods array.34 In fact, although the average nanorod diameter was ≈39 nm in both cases, their aspect ratio was appreciably different, that is, 9.1 for ZnO11 and 13.0 for ZnO12. Regarding ZnO12, an additional structured emission band in the region between 420 and 540 nm was also detected and ascribed to the presence of point defects, such as oxygen vacancies and zinc interstitials.2,5,20,21,24 The relatively welldefined visible emission of sample ZnO12 with respect to ZnO11 was reasonably related to the system structural evolution upon increasing the deposition temperature.
The present work was focused on the PE-CVD of ZnO nanorod arrays, with particular regard to the control of their morphogenesis as a function of the process parameters under mild synthesis conditions. Starting from two versatile bis(ketoiminato) precursors 1 and 2, never adopted in PE-CVD applications to date, the selective formation of highly pure, aligned, and c-axis oriented nanorods on Si(100) substrates was obtained at temperatures as low as 200-300 C and an RFpower of 20 W, in the absence of any catalyst. A phenomenological growth mode based on the ZnO crystal structure, as well as on electric field effects in the used PE-CVD configuration, was proposed to explain the selective formation of anisotropic 1D nanosystems. The present findings pave the way to the possible growth of high-quality ZnO nanorods even on amorphous and polymeric supports, an important goal for the development of advanced nanomaterials and nanodevices. In particular, future efforts will be devoted to investigate the functional properties of the obtained ZnO systems for eventual applications in field-emission devices, as well as in the sensing of selected pollutants and in their photocatalytic decomposition into nontoxic products. Acknowledgment. The research leading to these results has received funding from the Rektorat of the Ruhr-University Bochum (support for young female researchers) and the European Community under the 6th Framework Program [contract for an Integrated Infrastructure Initiative (Reference 026019, ESTEEM)]. CNR-INSTM PROMO, CARIPARO 2006 programme “Multi-layer optical devices based on inorganic and hybrid materials by innovative synthetic strategies”, RD-IFSC and RD-Plasma have provided further financial support. Mr. A. Ravazzolo (CNR, Padova, Italy) is acknowledged for skillful technical assistance. Supporting Information Available: LRI measurements (Figure SI-1) and additional GIXRD data (Figure SI-2). This material is available free of charge via the Internet at http://pubs.acs.org.
References (1) (a) http://apps.isiknowledge.com/; (b) http://www.scopus.com/ home.url.
2018
Crystal Growth & Design, Vol. 10, No. 4, 2010
(2) Kuang, Q.; Jiang, Z.-Y.; Xie, Z.-X.; Lin, S.-C.; Lin, Z.-W.; Xie, S.-Y.; Huang, R.-B.; Zheng, L.-S. J. Am. Chem. Soc. 2005, 127, 11777. (3) Zhang, T.; Dong, W.; Keeter-Brewer, M.; Konar, S.; Njabon, R. N.; Tian, Z. R. J. Am. Chem. Soc. 2006, 128, 10960. (4) Sandana, V. E.; Rogers, D. J.; Hosseini Teherani, F.; McClintock, R.; Bayram, C.; Razeghi, M.; Drouhin, H.-J.; Clochard, M. C.; Sallet, V.; Garry, G.; Falyouni, F. J. Vac. Sci. Technol., B 2009, 27, 1678. (5) Wu, C. C.; Wuu, D. S.; Lin, P. R.; Chen, T. N.; Horng, R. H. Cryst. Growth Des. 2009, 9, 4555. (6) Wang, Z. L. ACS Nano 2008, 2, 1987. (7) Fu, Y.-S.; Du, X.-W.; Kulinich, S. A.; Qiu, J.-S.; Qin, W.-J.; Li, R.; Sun, J.; Liu, J. J. Am. Chem. Soc. 2007, 129, 16029. (8) Wang, X.; Song, J.; Li, P.; Ryou, J. H.; Dupuis, R. D.; Summers, C. J.; Wang, Z. L. J. Am. Chem. Soc. 2005, 127, 7920. € ur, U¨.; Alivov, Ya. I.; Liu, C.; Teke, A.; Reshchikov, M. A.; (9) Ozg€ Dogan, S.; Avrutin, V.; Cho, S. J.; Morkoc-, H. J. Appl. Phys. 2005, 98, 041301. (10) Kang, S. T. W.; Yuldashev, U.; Panin, G. N. In Handbook of Semiconductor Nanostructures and Nanodevices; Balandin A. A.; Wang K. L., Ed.; American Scientific Publishers: Valencia, California, 2006. (11) Zhou, J.; Xu, N. S.; Wang, Z. L. Adv. Mater. 2006, 18, 2432. (12) Pradhan, D.; Leung, K. T. Langmuir 2008, 24, 9707. (13) Gao, P. X.; Lao, C. S.; Ding, Y.; Wang, Z. L. Adv. Funct. Mater. 2006, 16, 53. (14) Barreca, D.; Ferrucci, A. P.; Gasparotto, A.; Maccato, C.; Maragno, C.; Tondello, E. Chem. Vap. Deposition 2007, 13, 618. (15) Gao, P. X.; Wang, Z. L. J. Am. Chem. Soc. 2003, 125, 11299. (16) Sounart, T. L.; Liu, J.; Voigt, J. A.; Huo, M.; Spoerke, E. D.; McKenzie, B. J. Am. Chem. Soc. 2007, 129, 15786. (17) Liu, Z.; Wen, X. D.; Wu, X. L.; Gao, Y. J.; Chen, H. T.; Zhu, J.; Chu, P. K. J. Am. Chem. Soc. 2009, 131, 9405. (18) Lindsay, R.; Michelangeli, E.; Daniels, B. G.; Ashworth, T. V.; Limb, A. J.; Thornton, G.; Gutierrez-Sosa, A.; Baraldi, A.; Larciprete, R.; Lizzit, S. J. Am. Chem. Soc. 2002, 124, 7117. (19) Wu, J.-J.; Liu, S.-C. J. Phys. Chem. B 2002, 106, 9546. (20) Liu, X.; Wu, X.; Cao, H.; Chang, R. P. H. J. Appl. Phys. 2004, 95, 3141. (21) Kim, S.-W.; Fujita, S.; Fujita, S. Appl. Phys. Lett. 2005, 86, 153119/1. (22) Scalisi, A. A.; Toro, R. G.; Malandrino, G.; Fragala, M. E.; Pezzotti, G. Chem. Vap. Deposition 2008, 14, 115. (23) Falyouni, F.; Benmamas, L.; Thiandoume, C.; Barjon, J.; Lusson, A.; Galtier, P.; Sallet, V. J. Vac. Sci. Technol., B 2009, 27, 1662. (24) Park, J.-H.; Jang, S.-J.; Kim, S.-S.; Lee, B.-T. Appl. Phys. Lett. 2006, 89, 121108/1. (25) Black, K.; Jones, A. C.; Chalker, P. R.; Gaskell, J. M.; Murray, R. T.; Joyce, T. B.; Rushworth, S. A. J. Cryst. Growth 2008, 310, 1010.
Bekermann et al. (26) Ali, M.; Friedenberger, N.; Spasova, M.; Winterer, M. Chem. Vap. Deposition 2009, 15, 192. (27) Sun, J.; Mourey, D. A.; Zhao, D.; Jackson, T. N. J. Electron. Mater. 2008, 37, 755. (28) Rowlette, P. C.; Allen, C. G.; Bromley, O. B.; Wolden, C. A. J. Vac. Sci. Technol., A 2009, 27, 761. (29) Barankin, M. D.; Gonzalez, E., II; Ladwig, A. M.; Hicks, R. F. Sol. Energy Mater. Sol. Cells 2007, 91, 924. (30) Jaeger, R. C. Introduction to Microelectronic Fabrication, 2nd ed.; Upper Saddle River, NJ: Prentice Hall, 2002; Vol. 5. (31) Konuma, M. Film Deposition by Plasma Techniques; SpringerVerlag: Berlin, 1992. (32) Konuma, M. Plasma Techniques for Film Deposition; AlphaScience International Ltd.: Harrow, UK, 2005. (33) Kim, D. C.; Kong, B. H.; Cho, H. K. J. Phys. D: Appl. Phys. 2009, 42, 065406. (34) Jeong, S. H.; Lee, C. B.; Moon, W. J.; Song, H. J. J. Nanosci. Nanotechnol. 2008, 8, 5098. (35) Li, B. S.; Liu, Y. C.; Shen, D. Z.; Zhang, J. Y.; Lu, Y. M.; Fan, X. W. J. Cryst. Growth 2003, 249, 179. (36) Yamaoka, K.; Terai, Y.; Yamaguchi, T.; Fujiwara, Y. Phys. Status Solidi C 2008, 5, 3125. (37) Jeong, S. H.; Park, K. H.; Song, H. J. J. Korean Phys. Soc. 2007, 50, 1692. (38) Matthews, J. S.; Onakoya, O. O.; Ouattara, T. S.; Butcher, R. J. Dalton Trans. 2006, 31, 3806. (39) Bekermann, D.; Rogalla, D.; Becker, H.-W.; Winter, M.; Fischer R. A.; Devi, A. Eur. J. Inorg. Chem. 2010, 1366. (40) Barreca, D.; Gasparotto, A.; Tondello, E.; Sada, C.; Polizzi, S.; Benedetti, A. Chem. Vap. Deposition 2003, 9, 199. (41) Armelao, L.; Bertoncello, R.; Coronaro, S.; Glisenti, A. Sci. Technol. Cultural Heritage 1998, 7, 41. (42) Langford, R. M.; Petfold-Long, A. K. J. Vac. Sci. Technol., A 2001, 19, 2186. (43) Thompson, K.; Lawrence, D.; Larson, D. J.; Olson, J. D.; Kelly, T. F.; Gornam, B. Ultramicrosc. 2007, 107, 131. (44) Lee, D. J.; Park, J. Y.; Yun, Y. S.; Hong, Y. S.; Moon, J. H.; Lee, B.-T.; Kim, S. S. J. Cryst. Growth 2005, 276, 458. (45) Umar, A.; Karunagaran, B.; Suh, E. K.; Hahn, Y. B. Nanotechnol. 2006, 17, 4072. (46) Li, W.-J.; Shi, E.-W.; Zhong, W.-Z.; Yin, Z.-W. J. Cryst. Growth 1999, 203, 186. (47) Barreca, D.; Gasparotto, A.; Maccato, C.; Maragno, C.; Tondello, E. Surf. Sci. Spectra 2007, 14, 19. (48) Rowlette, P. C.; Allen, C. G.; Bromley, O. B.; Dubetz, A. E.; Wolden, C. A. Chem. Vap. Deposition 2009, 15, 15. (49) Stampfl, C.; Van de Walle, C. G. Phys. Rev. B 1998, 57, R15052. (50) Borras, A.; Sanchez-Valencia, J. R.; Widmer, R.; Rico, V. J.; Justo, A.; Gonzalez-Elipe, A. R. Cryst. Growth Des. 2009, 9, 2868.
