Catalytic growth of ZnO nanostructures by r.f. magnetron sputtering

  • María Arroyo-Hernández1Email author,

    Affiliated with

    • Raquel Álvaro1,

      Affiliated with

      • Sheila Serrano1 and

        Affiliated with

        • José Luis Costa-Krämer1

          Affiliated with

          Nanoscale Research Letters20116:437

          DOI: 10.1186/1556-276X-6-437

          Received: 4 November 2010

          Accepted: 24 June 2011

          Published: 24 June 2011

          Abstract

          The catalytic effect of gold seed particles deposited on a substrate prior to zinc oxide (ZnO) thin film growth by magnetron sputtering was investigated. For this purpose, selected ultra thin gold layers, with thicknesses close to the percolation threshold, are deposited by thermal evaporation in ultra high vacuum (UHV) conditions and subsequently annealed to form gold nanodroplets. The ZnO structures are subsequently deposited by r.f. magnetron sputtering in a UHV chamber, and possible morphological differences between the ZnO grown on top of the substrate and on the gold are investigated. The results indicate a moderate catalytic effect for a deposited gold underlayer of 4 nm, quite close to the gold thin film percolation thickness.

          Introduction

          Single crystalline zinc oxide (ZnO) nanowires are usually grown by wet chemical and vapour transport methods. The latter are performed at temperatures in the 850 to 1400°C range [1, 2]. Lower temperature (400°C) metalorganic vapour-phase epitaxial growth of vertically well-aligned ZnO nanorods has been also reported in [3]. Another kind of nanowires, Si and GaAs, are grown by vapour-liquid-solid deposition (VLS) using gold nanoparticle catalysts [4, 5]. Notably, III to V nano-whiskers have been grown on III to V substrates by metalorganic chemical vapour deposition (MOCVD) [6, 7]. This approach relies on annealing a thin Au film to form the seed particles [8]. In this way, a homogeneous whisker width distribution is obtained, the mean size of which could be controlled by the thickness of the Au layer and the way this layer transforms to nanoparticles. A similar approach to form ZnO nanostructures is reported herein, but using r.f. magnetron sputtering in ultra high vacuum (UHV) conditions, as a first step towards size-, shape- and position-controlled nanowires, similarly to what Samuelson and coworkers [9] started in GaAs in 2001. Our approach aims at obtaining nanostructures with low level of impurities for future studies on the correlation between defects and transport and photonic properties.

          Experimental

          ZnO films were grown on both silicon (100) and sapphire (Al2O3) substrates by a ZnO target magnetron sputtering. The stoichiometry of the films was checked under different growth conditions by non-RBS spectroscopy. Prior to the ZnO growth, a gold ultra thin underlayer was deposited by thermal deposition at 0.2 Å/s deposition rate. The base pressure is 10-8 mbar and increases slightly to approx. 10-7 mbar during the deposition process. For comparison purposes, a gold pattern was predefined on the substrate. This gold pattern allowed a straightforward comparison of possible ZnO morphology differences on a subsequent scanning electron microscopy inspection. The pattern was defined by electron lithography: a 200-nm PMMA-A4 resin was deposited by spinning for 1 min at 5000 revolutions per minute. Subsequently, they were cured on a hot plate for 4 min at 180°C. For the lithography, a high-resolution LEO 1455 scanning electron microscopy was used. Finally, the developing process was performed by immersing the samples in 4-methyl-2-pentanone + isopropyl alcohol (1:3) for 1.5 min and a subsequent rinse in isopropyl alcohol for 30 s to stop the process. After the development, the patterns were coated with desired gold thickness and subsequently lifted off in acetone.

          The gold films were thermally annealed using a tungsten wire heater placed below the holder substrate inside the UHV sputtering system. The annealing was performed for 20 min at 450°C in 10-2 mbar Ar pressure. The ZnO structures were grown by r.f. magnetron sputtering using a ZnO target. The base pressure is 10-8 mbar to ensure a low level of impurities. The growing conditions are: 100 W r.f. power, 500°C, 10-2 mbar Ar pressure, to ensure good crystallographic and conducting properties [10].

          The atomic force microscopy (AFM) analysis was performed using a commercial AFM (Nanotec, Madrid Spain) microscope, measuring in contact mode. Commercial tips (Nanosensors, Neuchatel, Switzerland) were used with K = 36 to 58 N/m and resonant frequencies in the 328 to 359 KHz range.

          The X-ray diffraction (XRD) experiments were performed using a Philips X-PERT four-cycle diffractometer with a Cu Ka radiation source in Bragg-Brentano geometry. The 2theta-omega range scanned was 30° to 95°. The crystal gold grain size was calculated from the diffractogram peak shape using Scherrer equation:
          http://static-content.springer.com/image/art%3A10.1186%2F1556-276X-6-437/MediaObjects/11671_2010_Article_550_Equ1_HTML.gif
          (1)

          where the shape factor k = 0.9, λ is 1.54Å, β the full width half maximum (FWHM) and θ the Bragg angle.

          Results

          Our approach to obtain gold seed particles relies on annealing ultra thin gold films. This annealing enhances atomic mobility and produces morphological changes that proceed towards island formation [8] or 3D growth. To check the effect of the thermal annealing on the gold film properties, AFM and XRD were measured for a selected 3-nm thickness gold film. This thickness was chosen because film percolation growth takes place around this nominal thickness value [11]. Figure 1 shows the AFM images obtained before (left pictures) and after (right pictures) the annealing. There are shown two fields of view, corresponding to 200-μm scale bar (top) and to 100-μm scale bar (bottom). The profiles correspond to the lines marked on the pictures. The morphological grain size increases from 27 to 100 nm and the surface roughness decreases from 0.72 to 0.54 nm, which confirms the mobility of gold atoms on the surface.
          http://static-content.springer.com/image/art%3A10.1186%2F1556-276X-6-437/MediaObjects/11671_2010_Article_550_Fig1_HTML.jpg
          Figure 1

          AFM pictures showing the morphology variations-grain size and roughness-due to thermal annealing for a 3-nm gold thickness film. Left images are before and right ones are after the annealing.

          The XRD spectra show that the evaporated gold is textured (111) with a crystallographic grain size of 27.5 nm (calculated from Equation 1), the same as the morphological value obtained by AFM. After the annealing process, the XRD grain size slightly increases up to 28.2 nm, representative of a moderate effect of the thermal treatment in the crystal quality (Figure 2). This marked difference with the AFM results can be understood in terms of 3D changes produced by the atomic mobility.
          http://static-content.springer.com/image/art%3A10.1186%2F1556-276X-6-437/MediaObjects/11671_2010_Article_550_Fig2_HTML.jpg
          Figure 2

          XRD pattern of a 3-nm nominal thickness gold film before (orange) and after (black) thermal annealing.

          To check a possible gold catalytic effect of the ZnO growth on gold, this was carried out for 2, 4 and 10 nm thin gold thicknesses. The morphology of the ZnO structures grown was inspected by SEM (Figure 3). We deliberately went above and below the gold percolation threshold to avoid irreproducible and drastic morphological and temperature dependencies inherent to percolation behaviour. To evaluate this possible catalytic effect of gold, the morphology of ZnO directly growth on sapphire (Figure 3, left) and on top of the annealed gold (Figure 3, right) were compared. As seen, there is no difference between both for 10 nm gold films. On the other hand, ZnO grown on 2 nm gold thickness shows a different structure depending on the material below, but only for 4 nm gold films marked differences are observed. In Figure 4, detailed images illustrating this effect are shown. The cross-sectional views clearly show a randomly oriented ZnO nanostructure when grown on 4 nm gold films, with lengths ranging from 80 to 220 nm. This confirms the moderate effect of gold for catalytic ZnO nanostructure formation.
          http://static-content.springer.com/image/art%3A10.1186%2F1556-276X-6-437/MediaObjects/11671_2010_Article_550_Fig3_HTML.jpg
          Figure 3

          SEM top views comparing the ZnO structures grown directly onto the substrate and onto an annealed ultra thin gold film of different thicknesses.

          http://static-content.springer.com/image/art%3A10.1186%2F1556-276X-6-437/MediaObjects/11671_2010_Article_550_Fig4_HTML.jpg
          Figure 4

          SEM tilted images and cross-sectional views of ZnO structures grown on 2 and 4 nm gold film nominal thickness.

          Finally, XRD was studied for both kinds of ZnO structures (Figure 5). The spectra show the diffraction peaks associated to the sapphire substrate and to the gold film. It can be observed that ZnO grown on 2-nm thickness gold has a polycrystalline structure, with two preferential orientations: (002) and (101). On the other hand, the ZnO structures grown on the 4-nm thickness gold are 'XRD amorphous'.
          http://static-content.springer.com/image/art%3A10.1186%2F1556-276X-6-437/MediaObjects/11671_2010_Article_550_Fig5_HTML.jpg
          Figure 5

          XRD diffraction spectra of ZnO structures grown on 2 nm (orange) and 4 nm (black) gold film nominal thicknesses.

          Conclusions

          In summary, experiments addressing a possible catalytic effect of gold on ZnO growth by r.f. magnetron sputtering under UHV conditions are presented. A moderate catalytic effect of gold is reported. The maximum effect is measured to happen at intermediate ultra thin gold nominal thicknesses, around 4 nm, and a subsequent thermal annealing at 450°C. This nominal thickness is slightly larger than the gold percolation one. The obtained ZnO nanostructures show a random orientation and are XRD amorphous. At this thickness range, the effect of the substrate temperature, the nominal ZnO thickness and the partial pressure composition during ZnO growth could be used to improve the catalytic effect and the nanostructure quality.

          Abbreviations

          AFM: 

          atomic force microscopy

          FWHM: 

          full width half maximum

          MOCVD: 

          metalorganic chemical vapour deposition

          UHV: 

          ultra high vacuum

          XRD: 

          X-ray diffraction

          ZnO: 

          zinc oxide.

          Declarations

          Acknowledgements

          We gratefully acknowledge M.U. González and J.M. Ripalda for suggestions, R González-Arrabal for non-RBS experiments and MAT2008-06330 for financial support

          Authors’ Affiliations

          (1)
          IMM-Instituto de Microelectrónica de Madrid (CNM-CSIC)

          References

          1. Fan Z, Lu JG: Nanostructured ZnO: building blocks for nanoscale devices. Int J High Speed Commun 2006, 16: 883. 10.1142/S0129156406004065View Article
          2. Wang ZL: Zinc oxide nanostructures: growth, properties and applications. J Phys Condens Matter 2004, 16: R829. 10.1088/0953-8984/16/25/R01View Article
          3. Park WI, Kim DH, Jung SW, Yi GC: Metalorganic vapor-phase epitaxial growth of vertically well-aligned ZnO nanorods. Appl Phys Lett 2002, 22: 4232–4234.View Article
          4. Wagner RS, Ellis WC: Vapor-liquid-solid mechanism of single crystal growth. Appl Phys Lett 1964, 4: 89–90. 10.1063/1.1753975View Article
          5. Givargizov EI: Growth of whiskers by the vapor-solid-liquid in current topics. In Material Science. Volume 1. Edited by: Kaldis K. Amsterdam: North-Holland; 1978:79–145.
          6. Hiruma K, Yazawa M, Haraguchi K, Ogawa K, Katsuyama T, Koguchi M, Kakibayashi H: GaAs free-standing quantum-size wires. J Appl Phys 1993, 74: 3162–3171. 10.1063/1.354585View Article
          7. Hiruma K, Murakoshi H, Yazawa M, Ogawa K, Fukuhara S, Shirai M, Katsuyama T: Growth and characterization of nanometer-scale GaAs, AlGaAs and GaAs-InAs wires. IEICE Trans Electron 1994, E77C: 1420–1425.
          8. Serrano A, Rodríguez de la Fuente O, García MA: Extended and localized surface plasmons in annealed Au Films on glass substrates. J Appl Phys 2010, 108: 074303. 10.1063/1.3485825View Article
          9. Ohlsson BJ, Björk MT, Magnusson MH, Deppert K, Samuelson L, Wallenberg LR: Size-, shape-, and position-controlled GaAs nano-whiskers. Appl Phys Lett 2001, 79: 3335–3337. 10.1063/1.1418446View Article
          10. Sundaram KB, Khan A: Characterization and optimization of zinc oxide films by r.f. magnetron sputtering. Thin Solid Films 1997, 295: 87–91. 10.1016/S0040-6090(96)09274-7View Article
          11. Fernández-Martínez I: Stress and nanostructure control for the development of magneto-electro-mechanical micro-devices. In PhDthesis. Universidad Autónoma de Madrid, Applied Physics Department; 2008.

          Copyright

          © Arroyo-Hernández et al; licensee Springer. 2011

          This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://​creativecommons.​org/​licenses/​by/​2.​0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.