Fabrication and Characterization of One Dimensional ZnO Nanostructures Title pages BY Chun Cheng A Thesis Presented to The Hong Kong University of Science and Technology In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Nano Science and Technology June 2009, Hong Kong Copyright © by Chun Cheng, 2009 i Authorization I hereby declare that I am the sole author of the thesis. I authorize the Hong Kong University of Science and Technology to lend this thesis to other institutions or individuals for the purpose of scholarly research. I further authorize the Hong Kong University of Science and Technology to reproduce the thesis by photocopying or by other means, in total or in part, at the request of other institutions or individuals for the purpose of scholarly research. __________________ Chun Cheng June, 2009 ii Fabrication and Characterization of One Dimensional ZnO Nanostructures Signature Pages BY Chun Cheng This is to certify that I have examined the above PhD thesis and have found that it is complete and satisfactory in all respects, and that any and all revisions required by the thesis examination committees have been made. APPROVED: ___________________________________________ PROF. NING WANG, SUPERVISOR ___________________________________________ PROF. ZIKANG. TANG, PROGRAM DIRECTOR Nano Science and Technology The Hong Kong University of Science and Technology June, 2009 iii Acknowledgment First of all, I would like to thank my supervisor, Prof. Ning Wang, for his support, guidance, encouragement and help in my research work and my life. His insight into nanomaterials science and enthusiasm for research has been a constant source of inspiration for me. His discipline has ensured I concentrate on my research and achieve the best possible results and his encouragement has helped me face problems with great confidence. Through his rich experience and deep understanding of crystallography, he taught me to think and conduct experiments independently and to experience the joy of scientific discovery. His inimitable passion for research and rapid and accurate handling of any experiment always impressed me. Undoubtedly, Prof. Wang will be an honorable model for me throughout my life. I would like to express my sincere gratitude to him for his valuable guidance, inspiration and great support throughout my entire research. I would also like to acknowledge Prof. K. K. Fung, Prof. I. K. Sou, Prof. G. H. Chen, Prof. J. Y. Dai and Prof. Y. D Wu for serving as my thesis examination committee members and providing practical comments. A special thank goes to Mr. Frankie Chan and Cai Yuan, my colleagues, who have given me the utmost help in the training of TEM and many other techniques. Special thanks are also given to Mr. T. K. Cheung for his skillful assistance in the TEM and Miss W.Y. Law for her generous help. The conscientious attitude they have always shown toward their work made a great impression on me. I would like to express my gratitude to my research group members and my research collaborators. We had many valuable discussions and they gave me much assistance. Special thanks are directed to Mr. Roy Ho, Mr. Gordon Suen and Mr. Tai Lun Wong, who provided much technical support in my experiments. Thanks to Dr. Yu Kaifeng, Dr. Xin Renlong, Mr. Xiao Zhizhao, Miss Feng Lin, Dr. Lei Ming, and Dr. Gu Changdong et al for working together with me to advance our research. I learned much from them and improved my teamworking skills by cooperating with them. I would like to thank my classmates, Mr. Xu Zhuli, Dr. Xie Hang, Dr. Ding Lu, Dr. Zhang Xieqiu, Mr. Lu Weixin, Mr. Li Baikui, Mr. Zhang Bei, Mr. Shi Wu, Dr. Liu Liyu, Dr. Zai Jianpan, et al. for their kind help and support. The help I received from other colleagues and research groups in HKUST is also greatly appreciated. I would also like to thank all my friends for their encouragement and concern. iv Last, but not the least, I would like to express my sincere thanks to my wife, my elder sister and my parents for their understanding and support. v Table of Contents Title pages..............................................................................................................................i Authorization........................................................................................................................ii Signature Pages .................................................................................................................. iii Acknowledgment .................................................................................................................iv Table of Contents.................................................................................................................vi List of Figures......................................................................................................................ix List of Tables....................................................................................................................xviii Abstract .............................................................................................................................xix Chapter 1 Research Background.........................................................................................1 1.1 Introduction ................................................................................................................1 1.2 General properties of ZnO .........................................................................................3 1.2.1 Crystal structures of ZnO .......................................................................................3 1.2.2 Growth habits of ZnO crystals ...............................................................................4 1.3 1D ZnO Nanostructures .............................................................................................6 1.3.1 Growth morphologies of 1D ZnO nanostructures ...................................................6 1.3.2 Photoluminescence (PL) of 1D ZnO nanostructures ...............................................8 1.3.3 Applications of 1D ZnO nanostructures................................................................ 11 1.4 Fabrication of 1D ZnO nanostructures....................................................................15 1.4.1 1D ZnO nanostructures from vapor transport synthesis ........................................15 1.4.2 1D ZnO nanostructures from solution synthesis ...................................................20 1.4.3 Patterned and aligned growth of ZnO 1D nano-arrays ..........................................21 Chapter 2 Structural and Optical Characterization Techniques......................................23 2.1 Transmission electron microscopy ...........................................................................23 2.1.1 Structure and operation modes of TEM ................................................................23 vi 2.1.2 Diffraction in the TEM ........................................................................................26 2.1.3 Imaging in the TEM.............................................................................................27 2.2 PL spectroscopy ........................................................................................................29 2.2.1 Introduction of luminescence ...............................................................................29 2.2.2 Photoluminescence ..............................................................................................31 Chapter 3 1D ZnO Nanostructures Fabricated with Vapor Transported Process...........33 3.1 Method I: Direct oxidation of Zn metal at high temperature in air .......................33 3.2 Method II: Carbon thermal method ........................................................................42 3.3 Method III: Direct evaporation of ZnO at high temperature and vacuum condition .........................................................................................................................................46 3.3.1 Controllable growth of ZnO NW arrays on carbon-based materials ......................48 3.3.1.1 Vertical growth of ZnO NW arrays on PR ......................................................48 3.3.1.2 Horizontal growth of ZnO NWs on PR...........................................................61 3.3.1.3 ZnO NW arrays grown on other carbon-based materials ...............................62 3.3.1.5 Summary .......................................................................................................64 3.3.2 Non c-axis growth of 1D ZnO nanostructures: substrate and temperature dependent morphologies................................................................................................................65 3.3.3 Defect related 1D ZnO nanostructures..................................................................71 3.3.3.1 Twin induced growth of Y-shaped ZnO nanobelts...........................................71 3.3.3.2 Screw induced growth of ZnO flowers ...........................................................77 3.4 Summary...................................................................................................................85 Chapter 4 Ambient Stability of ZnO NWs: Structural Degradation and Related PL .....86 4.1 Introduction ..............................................................................................................86 4.2 Experimental section ................................................................................................86 4.3 Results and discussion ..............................................................................................87 4.3.1 Structure and PL studies of ZnO nanostructures in air ..........................................87 4.3.2 Structure and PL studies of ZnO nanostructures in aggressive atmosphere ...........90 4.4 Summary...................................................................................................................93 Chapter 5 Chemical Stability and Biocompatibility of 1D ZnO Nanostructures.............95 vii 5.1 Introduction ..............................................................................................................95 5.2 Experimental section ................................................................................................96 5.2.1 Preparation and characterization of 1D ZnO nanostructures .................................96 5.2.2 In vitro experiments in SPS solution ....................................................................96 5.3 Results and discussion ..............................................................................................97 5.4 Summary................................................................................................................. 104 Chapter 6 1D TiO2-ZnO Nanohybrids ............................................................................ 106 6.1 Introduction ............................................................................................................ 106 6.2 Synthesis and characterization of TZO nanohybrids............................................ 107 6.2.1 Experimental section.......................................................................................... 107 6.2.2 Morphology and structure characterization......................................................... 108 6.3 Growth processes and mechanism of TZO nanohybrids ...................................... 111 6.4 Phase transition of TiO2 in TZO nanohybrids ....................................................... 114 6.5 Enhanced photocatalytic performance of TZO nanohybrids ............................... 117 6.6 Assemble TiO2 nanoparticles on other ZnO nanostructures................................. 123 6.7 Zn2TiO4/ZnO NW heterostructures ....................................................................... 124 Chapter 7 Conclusions & Future work ........................................................................... 130 Publication........................................................................................................................ 133 Reference .......................................................................................................................... 136 viii List of Figures Figure 1.1 Publication statistics on one-dimensional nanostructures for (a) ZnO and (b) Si and their corresponding citation report. The data were compiled on May 7, 2009 through the database from Institute of Scientific Information using the following keywords that appear in the topic: ZnO (or zinc oxide) vs. Si (or silicon) together with NR, NW, nanobelt, nanoribbon, nanotip, nanoring, nanofiber, nanospring, nanohelix, or nanobrush. ................................................ 2 Figure 1.2 Stick and ball representation of ZnO crystal structures: (a) cubic rock salt, (b) cubic zinc blende, and (c) hexagonal WZ. The shaped gray and black spheres denote Zn and O atoms, respectively [1]. ................................................................................................................ 3 Figure 1.3 (a) WZ structure ZnO, (b) important planes of WZ ZnO, (c) polar facets of ZnO nanostructures................................................................................................................... 4 Figure 1.4 (a) and (b) the idealized growth habits of the ZnO crystal; (c)-(d) and (e)-(f) are the practice growth habits in neutral and alkali mediums respectively [7]. ............................................ 5 Figure 1.5 Crystallographic axes and planes of ZnO ........................................................................... 7 Figure 1.6 Typical growth morphologies of 1D ZnO nanostructures and their corresponding facets .... 8 Figure 1.7 (a) Low magnification TEM image showing the size uniformity of ZnO nanobelts. (b) PL spectra acquired from the width= 200 nm wide ZnO nanobelts and the width= 6 nm wide ZnO nanobelts [10]. (c) Illustration of the calculated defect energy levels in ZnO from different literature sources. (d) Room-temperature PL spectra of different nanostructures: 1) Tetrapods, 2) needles, 3) NRs, 4) shells, 5) highly faceted rods, 6) ribbons/combs [9]...... 10 Figure 1.8 Photoconduction in NW photodetectors [39]. (a) Schematic illustration of a NW surface absorbed with oxygen molecules and corresponding energy band diagram. (b) Trapping and photoconduction mechanism in NW. Excited holes are captured by O2- ion, and release O2 by losing an electron. ........................................................................................................... 12 Figure 1.9 Schematic representation of an XSC applying ZnO NWs as the electron transport material. In a DSC the ZnO NWs are loaded with an adsorbed layer of a light harvesting Dye and the hole conductor is typically a liquid electrolyte with a I-/I3- redox couple [46]. ......................... 14 Figure 1.10 (a) Schematic diagrams showing the piezoelectric effect in a tetrahedrally coordinated cation-anion unit. (b) The experimentally measured piezoelectric coefficient d33 for ZnO and its comparison to that of the bulk [49]. ............................................................................ 15 Figure 1.11 (a) NRs formed due to anisotropic growth of ZnO crystals. (b) Unidirectional growth of ZnO single crystals due to screw dislocation. (c) Growth induced by twining. (d) Self-catalytic ix growth of ZnO NWs by Zn droplets. (e) ZnO crystals contain no catalysts and defects). (f) ZnO whiskers growth due to dislocations. (g) ZnO bi-crystal growth due to twining. (h) Zn or Zn-rich phase observed on the tips of ZnO NWs ............................................................ 17 Figure 1.12 Schematics of (a) and (b) the typical self-catalytic growth based on the VS process; (c) The catalyst assisted VLS process. ......................................................................................... 19 Figure 1.13 Schematic diagrams depicting the patterned growth of NWs by EBL through (a) vapor transport growth and (b) solution based growth. .............................................................. 22 Figure 1.14 The Scanning Electron Microscopy (SEM) image of aligned and partnered ZnO NWs from patterned growth sites fabricated by EBL: (a) and (b) by Au-assistant VLS growth [65] and (c)-(d) by solution synthesis . .......................................................................................... 22 Figure 2.1 Schematics of TEM consisting of five systems: illumination, specimen stage, imaging, magnification and data recording systems. The enlarged parts on the right are EDX and EELS for chemical composition analysis......................................................................... 25 Figure 2.2 Ray paths in TEM (a) diffraction mode and (b) imaging mode. ........................................ 26 Figure 2.3 Ray diagrams showing (a) SAED and (b) CBED pattern formation respectively. ............. 26 Figure 2.4 Ray diagram of CBED for determining the polarization of ZnO ....................................... 27 Figure 2.5 Comparison of the use of an objective aperture in TEM to select (a) the direct or (b) the scattered electrons forming BF and DF images, respectively. .......................................... 28 Figure 2.6 (a) Band diagram of semiconductor. (b) Electrons are excited from VB to CB (c) Electron transition from CB to VB. ............................................................................................... 30 Figure 2.7 The experimental set-up for PL measurements ................................................................. 32 Figure 3.1 (a) A sketch map of reaction apparatus and the deposition areas for tetrapods and NWs. (b) and (c) the SEM images of the ZnO structures formed by oxidation of Zn at different positions of the tube. X-ray diffraction (XRD) patterns (d) and (e) corresponding to the tetrapods and uniform NWs in b and c, respectively......................................................... 34 Figure 3.2 (a) The TEM image of ZnO NWs. (b) The HRTEM image recorded on one single ZnO NW with zone axis[1 100] . The inset is the Fast Fourier Transform (FFT) pattern of the HRTEM image.............................................................................................................................. 35 Figure 3.3 (a) and (b) Low magnification TEM images of a ZnO tetrapod. (c) A high resolution TEM image of the core region. (d) A CBED pattern from a leg of the tetrapod.......................... 37 Figure 3.4 TEM images of complexes consisting of ZnO tetrapods ................................................... 37 Figure 3.5 (a) BF TEM image and (b) DF TEM image of one single complex built by two tetrapods. The x insets in (b) are CBED patterns and simulated ones (below). (c) The enlarged image of the circled part. (d) HRTEM image of the IDB...................................................................... 38 Figure 3.6 The PL spectra of (a) the NW ensemble and (b) one single NW ....................................... 39 Figure 3.7 Normalized PL spectra at different positions, A, B and C, of a tetrapod. The dotted line corresponds to point A, the dashed line to point B and the solid line to point C. The inset shows a SEM image of the chosen tetrapod and the three positions A, B and C on it........ 40 Figure 3.8 SEM images showing the three typical morphologies of the as prepared products: (a) micro rods; (b) micro brushes; (c)-(e) micro & nano pyramids; (f) micro particle film and (g) their corresponding growing site temperatures......................................................................... 44 Figure 3.9 (a) The TEM image of the tip of one single nano pyramid and the EDS spectrum of the circled area. (b) and (c) The HRTEM image and its corresponding FFT pattern. .............. 45 Figure 3.10 PL of the ZnO micro & nano pyramid ensemble and one single ZnO pyramid................ 46 Figure 3.11 (a) The schematic diagram of the experimental setup for synthesis of ZnO nanomaterials. (b) The distribution of temperature in the stove from the center. The temperatures listed in the top-right box are the ones measured by thermal couple. ............................................. 47 Figure 3.12 Zn partial pressure over ZnO and vapor pressure of Zn solid, Zn liquid, and ZnO .......... 48 Figure 3.13 Fabrication process of ZnO nanostructure arrays directly from PR. (a) and (b) are Si substrates coated by PR patterns. (c) and (d) are the resulting NW arrays......................... 50 Figure 3.14 The optical (a) and SEM (b) images of ZnO NW arrays grown on a PR-coated silicon substrate. (c) ZnO NWs formed on an Au-coated silicon substrate. (d) XRD spectra recorded from the samples shown in (b) and (c). The upper one and the button one are corresponding to the sample in (b) and (c), respectively.......................................................................... 51 Figure 3.15 (a) The TEM image of an as-prepared ZnO NW and (b) the corresponding CBED patterns viewed along the [1 100] direction (the left one is the experimental result and the right one simulated by JEMS software). (c) An HRTEM image of a ZnO NW. The inset is the corresponding Fourier Transform Pattern. (d) The EDX spectrum recorded from the NW shown in (c). The copper peaks come from the sample supporting the grid. (e) PL spectra from A: a ZnO NW with a diameter of 30 nm; B: a ZnO NW with a diameter of 300 nm and C: ZnO NW arrays. (f) Reflectance spectra of ZnO NW arrays grown on (I) PR (Figure 3.14b), (II) Au-coated silicon substrate (Figure 3.14c), (III) Si substrate with the remaining carbonized PR after removing the ZnO NWs with a 10 % HNO3 solution, (IV) naked Si substrate and (V) random piled ZnO NWs....................................................................... 53 Figure 3.16 Schematic illustration showing the fabrication process of multilayer ZnO NW arrays..... 54 xi Figure 3.17 The SEM images of (a) PR/ZnO matrix and (b) multilayered structures with ZnO NW arrays as building blocks. .......................................................................................................... 54 Figure 3.18 Various ZnO nanostructure arrays from PR patterns: (a) square dot arrays, (b) hexagonal dot arrays, (c) line arrays, and (d) hexagonal networks. (e) ZnO nanopin arrays. On the right side are the corresponding enlarged images. ........................................................................... 56 Figure 3.19 (a) ZnO NWs grown on different sizes of PR patterns. Insets are enlarged pictures of the ZnO NWs formed on the patterns. (b) One single ZnO NW nucleated and grown at the corner of each small PR pattern.................................................................................................. 57 Figure 3.20 (a) Raman spectra of the photoresists before (the bottom curve) and after annealing (the top curve). (b) Nucleation and growth mechanisms of ZnO NWs on the photoresist patterns. 58 Figure 3.21 TEM images of tips of ZnO NWs. (b)-(d) are HRTEM images of tips ............................ 59 Figure 3.22 (a) The cross-section TEM image of sample of ZnO on PR in initial growth stage. The PR layer is cleaved for stress. (b) The ending of ZnO NWs on PR. (c) The interface between the root of ZnO NWs and the PR. (d) The corresponding FFT pattern shows the ZnO structure. ....................................................................................................................................... 59 Figure 3.23 The SEM images of horizontal growth of ZnO NWs. ..................................................... 61 Figure 3.24 The schematic illustration shows the horizontal growth process of ZnO NWs. ............... 62 Figure 3.25 The SEM images of ZnO NW arrays synthesized with carbon-based materials: (a) grease left on Si substrate by fingerprint, (b) HOPG, (c) graphite strip and (d) amorphous carbon film on Si substrate by a carbon coater (Denton, Bench-Top Turbo)................................. 63 Figure 3.26 A 2 inch silicon wafer with ZnO NW arrays................................................................... 63 Figure 3.27 The schematic diagram of main growth temperature zones and the corresponding typical morphologies of nanostructured products. ....................................................................... 65 Figure 3.28 (a) SEM images of typical morphologies of ZnO nanochains. (b) The TEM images of nanochains. (c) The HRTEM image of one single nanochain with its FFTs inset. (d) The sketch maps of ZnO nanochains. See text for details........................................................ 66 Figure 3.29 (a) The SEM images of nanobrush products................................................................... 67 Figure 3.30 (a) and (b) are the TEM images of ZnO [1120] /{1 100} and [1 100] /{1120} nanobrushes, respectively. Insets are SAED patterns. (c) and (d) are the corresponding CBED patterns (experimental and simulated patterns). Scale bar 2 µm for (a) and (b).............................. 68 Figure 3.31 (a) and (b) The SEM and TEM images of uniform nanobelts. (c) and (d) are the HRTEM images recorded on the [000 1] and [0001] sides of one nanobelt shown in the inset in panel xii (c). The polar directions are identified by CBED. (e) A single nanobelt showing its development from thin NWs. .......................................................................................... 69 Figure 3.32 The schematic illustration shows the relationship between typical ZnO morphologies, nanochains, nanobrushes, and nanobelts with thin NWs/nanobelts with polar (0001) side surfaces........................................................................................................................... 70 Figure 3.33 A SEM image of the as-synthesized twinned ZnO nanobelts. The inset is an enlarged image of the twinned ZnO nanobelts. Scale bar 10 μm and 1μm for the inset. ............................ 72 Figure 3.34 (a) Bright-field TEM image of a single twinned ZnO nanobelt. (b) and (c) Dark-field TEM images of twinned ZnO nanobelts recorded by a center objective aperture in the positions of (0001) and (0001’) of the SAED pattern taken from the whole twinned ZnO nanobelts respectively. The insets in (a) are SAED patterns recorded from the place as marked by arrows. Scale bar 1μm for all........................................................................................... 72 Figure 3.35 (a) Bright-field TEM images of a single twinned ZnO nanobelt for HRTEM. (b)-(f) are HRTEM images recorded from the position 1-5. The insets pointed by arrows in (f) are CBED patterns recorded at the two sides respectively and the simulated ones that are marked with stars. The insets at the right bottom in (f) are the corresponding FFT of f). Scale bar 1μm for (a); 5μm for (b)- (f).................................................................................................... 74 Figure 3.36 (a) High magnified optical images of single Y-shaped ZnO nanobelts. (b) The large area PL spectrum of Y-shaped ZnO nanobelts. ............................................................................. 76 Figure 3.37 SEM images of the ZnO hierarchical structures as-synthesized. (a), (c) and (e) are typical morphologies from low temperature region to high temperature region. (b), (d) and (f) are corresponding enlarged ones of (a), (c) and (e). Number 1, 2 and 3 mark single-layered, multilayered and multifid flowers, respectively. Scale bar, 100 μm for (a), (c) and (e); 20 μm for (b) and (d); 10μm for (f). ........................................................................................... 78 Figure 3.38 (a) TEM image of a nanoflower with the beam direction along the open direction of it. The inset is the SAED pattern of the nanoflower, which shows the nanoflower open toward [0001] and be single crystal. b) A TEM image of a nanoflower with the beam direction perpendicular to the stem. c) The CBED pattern of e) taken along [1 100] showing that the nanoflower grows from Zn-terminated polar (0001) site. Scale bar, 2 μm for (a); 5 μm for (b). ....................................................................................................................................... 78 Figure 3.39 (a) and (b) are the projection maps for nanoflowers that consist of different (1 10 x) (x=1, 2, 3, 4, 5, 6) viewed along [1 100] and[1120] . (c) and (d) are two typical nanoflowers with large opening angles. (e) is a typical nanoflower with small opening angles. (f)-(h) are the xiii corresponding three dimensional profile maps of (c)-(e). (f) and (g) consist of {1 104}planes and (h) consists of {1 103} planes. Nanoflowers with left handedness (i), right handedness (j) and two screws that possess same-handedness (k). Scale bar, 10 μm for (c) and (d); 5 μm for (e) and (i)-(k). ................................................................................................................. 80 Figure 3.40 SEM images of NWs with flowers being trilled through (a) and at their tip endings (b); Belt-like hierarchical structures along [1 100] (c) and [1120] with flowers open toward [0001] (d); Some particles are observed at the ridge of the [000 1] side of the belt-like hierarchical structure(e); Developing morphologies of belt-like hierarchical structures (f)-(h); TEM image and SAED of (g) showing the belt is a single crystal with the spine along [1120] and its projection plane perpendicular to [0001]; (j)- (l) are SEM images of flowers growing on thick rectangular belt along [1120] , [1 101] and [1 102] direction. Scale bar, 10 μm for (a)-(d), (f) and (j)-(l); 5 μm for (g); 2 μm for (e) and (h); 200 nm for (i)........... 82 Figure 3.41 SEM images of developing morphologies of dendrites with flowers growing at the tips (a). Balls with six symmetry form and then flowers grow on the ball (b)-(d). Scale bar, 10 μm for (a); 2 μm for (b)-(d). ....................................................................................................... 83 Figure 3.42 (a)-(c) SEM images of flowers growing on a thick rectangular belt along[1120] , [1123] and[2243] . (d) Flower growth initiates from the screw dislocations. Scale bar, 10 μm for (a)-(c) and 5 μm for (d). .................................................................................................. 83 Figure 4.1 (a) XRD and (b) SEM image of ZnO NW product. (c) HRTEM image of a single ZnO NW. The inset is the corresponding FFT pattern. (d) Normalized PL spectrum taken from the large-area ZnO NWs. ...................................................................................................... 88 Figure 4.2 (a) the SEM image of a fresh tapered NW. (b) PL spectra collected at the thinner part, A, (dash line) and at the thicker part, B, (solid line) of a tapered fresh NW. (c) HRTEM of the single NW aged in air for seven weeks. ..................................................................................... 89 Figure 4.3 (a) SEM image of a tapered NW aged in the complex environment of CO2 + H2O for different time: “1”, “2” and “3” denote sample aged for zero, one and three weeks, respectively. (b) The PL spectra of an individual NW before (“1”) and after aging in the CO2 + H2O environment for one week (“2”) and for three weeks (“3”), respectively. (c) TEM image of the NW aged for one week. (d) and (e) HRTEM image of area A and B marked in Figure 4.3c, respectively. (f) TEM image of the NW aged for three weeks. (g) HRTEM image of the area C marked in Figure 4.3f. ................................................................................................. 92 Figure 4.4 (a) PL spectra recorded from an individual ZnO NW before and after annealing in H2 at 400ºC for 30min. (b) Another results from a ZnO NW before and after annealing in O2 at 400ºC for xiv 30min. ............................................................................................................................ 93 Figure 5.1 (a) XRD patterns of ZnO NWs fabricated by the hydrothermal method (bottom) and NWs (middle) /nanobelts (top) by thermal vapor deposition. (b) PL spectra of ZnO NWs fabricated by the hydrothermal method (bottom) and NWs (middle) /nanobelts (top) by the thermal vapor deposition.............................................................................................................. 98 Figure 5.2 (a) ZnO NWs fabricated by the hydrothermal method. (b) ZnO NWs fabricated by thermal evaporation. (c) Amorphous thin calcium phosphate shells formed on the surfaces of the ZnO NWs synthesized by the hydrothermal method. These ZnO NWs acted as the templates. (d) Calcium phosphate shells coated on the ZnO NWs synthesized by thermal evaporation. (e) HRTEM image of calcium phosphate shell-ZnO NW structures and the EDS spectrum of the calcium phosphate shell. Scale bars 200 nm for (a) and (b), 100 nm for (c) and (d), 5 nm for (e)................................................................................................................................... 99 Figure 5.3 TEM images of etched ZnO nanostructures: (a) NWs viewed along[1 100] ; (b) nanobelts with a (0001) dominant plane; (c) and (e) with a (1 100) dominant plane; (d) with a (1120) dominant plane. Scale bar 50 nm for (a), (b); 200 nm for (c)-(e)......................... 101 Figure 5.4 CBED experimental patterns (left) and simulated patterns (right) for the samples shown in (a) Figure 5.3a, (b) Figure 5.3c, (c) Figure 5.3d, (d) Figure 5.3e.......................................... 102 Figure 5.5 HRTEM images showing the typical morphologies of the voids viewed along (a) [1 100] and (b) [1120] directions. Insets are corresponding Fourier transform patterns. (c) and (d) are the profile graphs of the outlined areas A and B shown in (a) and (b). Scale bar 5nm for a) and b). ..................................................................................................................................... 103 Figure 5.6 (a) Schematic model of an etched void. (b) and (c) 3D projections of the voids viewed along the [1 100] and [1120] directions. d) and e) are profile images of cross-sections of (b) and (c) cut perpendicular to the [0001] direction. ...................................................................... 104 Figure 6.1 SEM images of (a) ZnO NRs, (b) ZnO NRs capped with TiO2 particles. ........................ 108 Figure 6.2 (a) XRD pattern of the as-prepared nanohybrid product. The reflections of ZnO crystal are marked by the indices. The Al peaks come from the aluminum holder in XRD measurement. (b) TEM image showing the morphology of the fabricated ZnO/TiO2 nanohybrid product. (c) TiO2 nanoparticles assembled at one end of the ZnO NR only. (d) SAED pattern of the nanohybrid product. ...................................................................................................... 109 Figure 6.3 (a) HRTEM image of an individual ZnO/TiO2 structure. (b) EDX spectra recorded by focusing the electron beam on the NR and the TiO2 cap, respectively. The C and Cu signals came from a carbon-supporting film that was prepared on a copper grid. (c) The amorphous xv caps (the hole of each of which is indicated by the arrows) that may have become detached from the ZnO/TiO2 nanohybrids during TEM sample preparation. (d) The CBED pattern taken along the [1120] direction. (e) The corresponding simulated CBED pattern..............111 Figure 6.4 Typical TEM images of the products after (a) 1 hour and (b) 3 hours of reaction. The inset in (b) is an enlarged TEM image to show the big difference of the contrasts for TiO2 particle and ZnO NR. Amorphous particles are marked with arrows. (c) DSC and TGA results for the nanohybrid product. The inset in (c) is the enlarged DSC curve for the temperature range of 250 to 500oC. The scale bar is 100 nm............................................................................113 Figure 6.5 Schematic illustration showing the growth mechanism of TZO nanohybrids. ..................113 Figure 6.6 XRD patterns of TZO, TZO300 and TZO600 products. The apparent sharp peaks are all attributed to WZ ZnO.....................................................................................................114 Figure 6.7 HRTEM images showing the phase transformation by annealing at (a) 300°C and (b) 600°C for 2 hours. The insets show the low magnification TEM images and the FFT patterns of the areas that are marked by the dashed lines in the main picture. The images were taken from the zone axes of [1120] for ZnO, [201] for anatase TiO2 (Figure 6.7a), and [100] for rutile TiO2 (Figure 6.7b). c) Anatase crystallites formed at the outermost shell of the cap after illumination with a convergent electron beam. The insets show images of the TiO2 caps before and after electron beam illumination. The scale bar is 5 nm. ................................115 Figure 6.8 UV-Vis absorption spectra of amorphous TiO2 NPs, ZnO NRs, TZO nanohybrids and the annealing products TZO300 and TZO600.......................................................................117 Figure 6.9 PL spectra of ZnO NRs, TZO nanohybrids, TZO300 and TZO600 ..................................118 Figure 6.10 The home-made photocatalytic reactor system ..............................................................119 Figure 6.11 (a) Photodegradation of MB by TZO products amorphous, ZnO NRs, amorphous TiO2 NPs and the blank experiment. (b) A comparison of photocatalytic activity of ZnO NRs, TZO, TZO300 and TZO600. .................................................................................................. 120 Figure 6.12 Illustration of photoinduced charge transfer and separation in the interface of TZO heterostructures............................................................................................................. 121 Figure 6.13 FTIR spectra of the as-synthesized TZO, TZO300 and TZO600 samples in the wavenumber ranges of 4000-400 cm-1. The broad absorptions at about 3352 and 1639 cm-1 are assigned to the hydroxyl groups of chemisorbed and/or physisorbed H2O molecules on the samples. A strong absorption band near 540 cm-1 reveals the vibration properties of ZnO NRs. Other unsigned peaks are attributed to remnant organic species in the samples. ....................... 121 Figure 6.14 The SEM images of ZnO nanostructures with TiO2 nanoparticles assembled on their 0001 xvi ends. ............................................................................................................................. 123 Figure 6.15 The SEM images of (a) ZnO NW arrays, (b) TZO NW arrays, (c) Zn2TiO4/ZnO NW arrays, and (d) Schematic illustration of the formation process of Zn2TiO4/ZnO heterostructures. ..................................................................................................................................... 124 Figure 6.16 (a) The low magnification TEM image of a single hydrogen TiO2 hydrate/ZnO NW heterostructure. (b) SAED pattern recorded at the nanoparticle. (c) The HRTEM image of nanoparticle illuminated by electron beam and its FFT. ................................................. 125 Figure 6.17 (a) and (e) the low magnification TEM images of two single Zn2TiO4/ZnO NW heterostructures with different oriented relationships. (c) and (g) are the HRTEM images near the interface of the heterostructures. (b) and (d) are FFT patterns corresponding to area A and B respectively; (f) and (h) are FFT patterns corresponding to area C and D respectively. ..................................................................................................................................... 127 Figure 6.18 Schematic illustration of two series of oriented relationships between the spinel Zn2TiO4 and WZ ZnO. ...................................................................................................................... 128 Figure 6.19 TEM of one single TZO and corresponding EDS elemental mapping of oxygen, titanium and zinc of (a) single TZO nanohybrid and (b) Zn2TiO4/ZnO heterostructures ............... 129 xvii List of Tables Table 1.1 Calculated cleavage energies of different surfaces of WZ ZnO ............................................ 7 Table 3.1 A summary of properties of carbon-based materials for the growth of ZnO NWs ............... 64 Table 4.1 Aging conditions and results of structure and PL property ................................................. 91 Table 5.1 Ion concentrations of SPS ................................................................................................. 97 Table 5.2 A comparison of methods for identifying the polarity of ZnO nano crystals ..................... 105 xviii Fabrication and Characterization of One Dimensional ZnO Nanostructures By Chun Cheng Nano Science and Technology The Hong Kong University of Science and Technology Abstract In this thesis, one dimensional (1D) ZnO nanostructures with controlled morphologies, defects and alignment have been fabricated by a simple vapor transfer method. The crystal structures, interfaces, growth mechanisms and optical properties of ZnO nanostructures have been investigated by scanning electron microscopy (SEM), transmission electron microscopy (TEM) and photoluminescence (PL) spectroscopy. Great efforts have been devoted to the patterned growth and assembly of ZnO nanostructures as well as the stability of ZnO nanowires (NWs). Using carbonized photoresists, a simple and very effective method has been developed for fabricating and patterning high-quality ZnO NW arrays. ZnO NWs from this method show excellent alignment, crystal quality, and optical properties that are independent of the substrates. The carbonized photoresists provide perfect nucleation sites for the growth of aligned ZnO NWs and also perfectly connect to the NWs to form ideal electrodes. This approach is further extended to realize large area growth of different forms of ZnO NW arrays (e.g., the horizontal growth and multilayered ZnO NW arrays) on other kinds of carbon-based materials. In addition, the as-synthesized vertically aligned ZnO NW arrays show a low weighted reflectance (Rw) and can be used as antireflection coatings. Moreover, non c-axis growth of 1D ZnO nanostructures xix (e.g., nanochains, nanobrushes and nanobelts) and defect related 1D ZnO nanostructures (e.g., Y-shaped twinned nanobelts and hierarchical nanostructures decorated by flowers induced by screw dislocations) is also present. Using direct oxidization of pure Zn at high temperatures in air, uniformed ZnO NWs and tetrapods have been fabricated. The spatially-resolved PL study on these two kinds of nanostructures suggests that the defects leading to the green luminescence (GL) should originate from the structural changes along the legs of the tetrapods. Surface defects in these ZnO nanostructures play an unimportant role for the GL emission. On the other hand, those ZnO tapered structures fabricated by a modified carbon thermal method with the assistance of Au catalysts display strong UV emission, indicating a good crystallization quality. The stability, structural degradation and related PL property of ZnO NWs under different environments of surface treatments have been investigated by high-resolution transmission electron microscopy (HRTEM) and near field optical microscopy (NSOM). For high-quality ZnO NWs, the UV emission shows no change and no DL emission was generated during the structural degradation. For those ZnO NWs showing GL emission, the commonly used treatment methods e.g., post-annealing can not effectively eliminate the GL emission. The chemical stability and biocompatibility of ZnO nanostructures in simulated physiological solution (SPS) are studied by electron diffraction and HRTEM. ZnO nanostructures fabricated by the thermal evaporation method were found to survive much longer in SPS than those fabricated using a hydrothermal solution method. Calcium hydrogen phosphate amorphous layers structures have been observed to have excellent interfacial contacts with ZnO NWs. The shapes of the voids formed in the ZnO NWs are due to the interesting anisotropic etching behaviors in SPS which can be used to identify the polar directions of ZnO nanocrystals. Using hydrothermal reaction, TiO2/ZnO (TZO) nanohybrid structures have been found to form through the site-specific deposition of TiO2 on ZnO nanorods (NRs). TEM studies have revealed each ZnO NR to be assembled with one TiO2 cap at the Zn terminated (0001) surface. The polarity of the ZnO (0001) surface plays an important role in the formation of the TZO nanohybrid structures. The TZO nanohybrids contain uniform and atomically flat interfaces between ZnO and TiO2 with tunable crystal phases, which can be amorphous, anatase and rutile through annealing treatments. These nanohybrid structures demonstrate an enhanced xx photocatalytic activity due to the improved interface structures for a better interfacial charge-transfer/spatially separation process of photogenerated charge carriers. The site-specific deposition method has also been applied to assemble TiO2 on the (0001) surfaces of other ZnO nanostructures such as tetrapods, nanofilms, nanoflowers and NW arrays produced by different synthesis techniques. Through high temperature annealing, the TZO nanohybrid structures can be further converted into Zn2TiO4/ZnO nanostructures with certain orientation relationships. These nanohybrid structures may synergize the properties of both components and lead to many promising applications. xxi Chapter 1 Research Background 1.1 Introduction Zinc oxide is an important and promising material with great potential for lots of practical applications [1], such as optical waveguides, chemical and gas sensors, spin functional devices, piezoelectric transducers, surface acoustic wave devices, varistors, phosphors, transparent conductive oxides, and ultraviolet (UV) light emitters. The lack of a centre of symmetry in wurtzite (WZ), combined with large electromechanical coupling, has resulted in strong piezoelectric and pyroelectric properties and the consequent use of ZnO in mechanical actuators and piezoelectric sensors [2]. Its wide bandgap (3.37 eV at room temperature) makes ZnO a promising material for photonic applications in the UV or blue spectral range, while the high exciton-binding energy (60 meV) allows an efficient excitonic emission even at room temperature. In addition, ZnO doped with transition metals shows great promise for spintronic applications. It has also been suggested that ZnO exhibits sensitivity to various gas species [3], namely ethanol, acetylene, and carbon monoxide, which makes it suitable for sensing applications. ZnO is transparent to visible light and can be made highly conductive by doping. Also, ZnO is biocompatible which enables it suitable for biomedical applications. Moreover, ZnO is chemically stable and environmentally friendly. Consequently, ZnO is a versatile functional material that has a diverse group of growth morphologies, in the form of powders, single crystals, thin films, and nano-sized structures [4], such as nanoparticles, nanorods (NRs), nanowires (NWs), nanobelts, nanobrushes, nanocombs, nanorings, nanohelixes/nanosprings and nanocages. These novel nano-structural ZnO materials have received broad attention due to their distinguished performance in electronics, optics and photonics. With reduction in size, novel electrical, mechanical, chemical and optical properties are introduced, which are believed to be the result of surface and quantum confinement effects [5]. With large surface to volume ratio, various deep level emissions of intrinsic or extrinsic defects in the visible range contribute to the optical transitions more significantly in addition to the UV excitonic emission. 1 One dimensional (1D) ZnO nanostructures are an attractive and ideal system for studying the electronic transport process in one dimensionally confined objects, which are beneficial not only for understanding the fundamental phenomena in low dimensional systems, but also for developing new generation nanodevices of high performance. Due to the recent interest in this field the number of researchers working on the 1D ZnO nanostructure topics has been increasing rapidly. As for 1D nanostructures, ZnO has equal importance to silicon-based 1D nanostructures, according to the literature (Figure 1.1), and is playing an increasingly key role in developing nanoscience and nanotechnology. Figure 1.1 Publication statistics on one-dimensional nanostructures for (a) ZnO and (b) Si and their corresponding citation report. The data were compiled on May 7, 2009 through the database from Institute of Scientific Information using the following keywords that appear in the topic: ZnO (or zinc oxide) vs. Si (or silicon) together with NR, NW, nanobelt, nanoribbon, nanotip, nanoring, nanofiber, nanospring, nanohelix, or nanobrush. The application and development of 1D ZnO nanostructures largely relies on improvements in both growth and characterization techniques. In this thesis, we develop new 2 methods for 1D ZnO nanostructure growth, in particular the rational growth of ZnO NW arrays. Basic optical properties, chemical stability and biocompatibility properties of 1D ZnO nanostructures are investigated for the promising application of nanodevices based on 1D ZnO nanostructures. Furthermore, ZnO NW nanohybrids are developed to synergize the properties of different components as an active exploration to broaden the applications of 1D ZnO nanostructures. 1.2 General properties of ZnO 1.2.1 Crystal structures of ZnO Zinc oxide crystallizes into three forms [1]: cubic rock salt, cubic zinc blende, and hexagonal WZ, as shown in Figure 1.2. The rock salt NaCl-type structure is only observed at relatively high pressure ~10 GPa. The zinc blende form can be stabilized by growing ZnO on substrates with a cubic lattice structure. The WZ structure is most stable and thus most common in ambient conditions. Figure 1.2 Stick and ball representation of ZnO crystal structures: (a) cubic rock salt, (b) cubic zinc blende, and (c) hexagonal WZ. The shaped gray and black spheres denote Zn and O atoms, respectively [1]. ZnO crystallizes in the WZ structure belongs to the space group P63mc. The hexagonal structure of ZnO can be simply described as a number of alternating planes composed of tetrahedrally coordinated O2− and Zn2+ ions, stacked alternately along the c-axis (Figure 1.3a). 3 The structure lacks inversion symmetry and cutting the crystal perpendicularly to the c-axis results in two structurally different surfaces. Hence the two opposite sides of the c-oriented wafer are terminated with one type of ion only. These basal surfaces (bases of the prism shown in Figure 1.3b) are usually referred to as (0001)-Zn and (000 1) -O terminated surfaces. The (1 100) and (1120) surfaces are the prism faces and the (1121) surface is the pyramid face of the crystal. Polar surfaces are an important characteristic of ZnO. The most common polar surface is the basal plane. Oppositely charged ions produce positively charged (0001)-Zn and negatively charged (000 1) -O surfaces, resulting in a normal dipole moment and spontaneous polarization along the c-axis as well as a divergence in surface energy. To maintain a stable structure, the polar surfaces generally have facets or exhibit massive surface reconstructions, but ZnO ± (0001) are exceptions: they are atomically flat, stable and without reconstruction. Efforts to understand the superior stability of the ZnO ± (0001) polar surfaces are at the forefront of research in today’s surface physics [6]. In addition to the most typical ± (0001) polar surfaces, ± (10 1 1) and ± (10 11) are also polar surfaces as shown in Figure 1.3c. Figure 1.3 (a) WZ structure ZnO, (b) important planes of WZ ZnO, (c) polar facets of ZnO nanostructures. 1.2.2 Growth habits of ZnO crystals The growth habits of crystals are mainly determined by their internal structure and are 4 simultaneously affected by external growth conditions. The study of crystal growth reveals the growth mechanism of the crystal and vice versa. ZnO growth habits have been investigated by the hydrothermal method because in this method the supersaturation of the solution is low, so that the as-prepared crystals tend to have a regular polyhedral crystal face. It is observed that the relationship between growth speeds along different directions is V[0001] > Vnr ⊥{10 1 1} > Vnr ⊥{10 11} > V[000 1 ] . Figure 1.4 shows the idealized growth habits of the ZnO crystals according to this relationship and their growth habits in neutral and alkali mediums respectively. W. J. Li et al. [7] correlate the growth rate at various crystal faces with their interface structure, and consider the effect of external conditions on crystal growth by the rule that the crystal face with the corner of the coordination polyhedron present at the interface has the fastest growth rate; the crystal face with the edge of the coordination polyhedron present at the interface has the second fastest growth rate; the crystal face with the face of the coordination polyhedron present at the interface has the slowest growth rate. In terms of this rule, the growth habits of ZnO crystal particles and the effect of a reaction medium on it are successfully explained. The actual synthesized ZnO crystal shape depends on the temperature, the precursor, the solution basicity and the shapes of the seed crystals, and so on. Figure 1.4 (a) and (b) the idealized growth habits of the ZnO crystal; (c)-(d) and (e)-(f) are the practice growth habits in neutral and alkali mediums respectively [7]. 5 1.3 1D ZnO Nanostructures 1.3.1 Growth morphologies of 1D ZnO nanostructures Structurally, ZnO has three types of fast growth directions: < 1120 > (± [1120] , ± [1210] , ± [2110] ); < 1 100 > (± [1 100] , ± [1010] , ± [01 10] ); and ± [0001] (Figure 1.5). Together with the polar surfaces due to atomic terminations and abundant structure symmetries, that is, two major structural symmetries 6-, 2-fold and corresponding subsymmetries, ZnO can exhibit a wide range of morphologies. Furthermore, from the kinetic process aspect, crystal structures tend to minimize the total energy. For nanostructures, the surface energy is the dominant part of the total energy because of their high aspect ratio property. Up to now, no experimental measurements of cleavage energies (or surface energies, half of the cleavage energies) of ZnO surfaces are to be found. Calculated cleavage energies of ZnO surfaces taken from the current literature are compiled in Table 1.1 Though the absolute values of these cleavage energies vary using different calculation methods, the relative relation between them are the same: γ {1 100} <γ {1120} <γ {0001} . According to this relationship, {1 100} planes should be the dominant planes that are frequently observed in ZnO crystals. However, the relative surface activities of various growth facets under given conditions can be tuned, resulting in diversified morphologies of ZnO nanostructures. Figure 1.6 shows a few typical growth morphologies of 1D ZnO nanostructures. These structures tend to maximize the areas of (1 100) and (1120) facets because of the lower total energy. The morphology shown in Figure 1.6d is dominated by the polar surfaces. Planar defects and twins are observed occasionally parallel to the (0001) plane, but dislocations are rarely seen. 6 Figure 1.5 Crystallographic axes and planes of ZnO Table 1.1 Calculated cleavage energies of different surfaces of WZ ZnO [8] Surface orientation (1 100) Cleavage energy (J/m2) 2.3a 2.3b 1.8c (1120) 4.1a 2.5b,1.7 1.9c O- (000 1) , Zn(0001) (1×1) 4.0a 4.3a 3.4 3.5c Zn- (0001) triangles 2.5c These total-energy calculations may act as a rough guideline for predicting the stability of the various ZnO surfaces. a B3LYP functional. b Local density approximation (LDA). c Generalized gradient approximation (GGA-PBE) 7 Figure 1.6 Typical growth morphologies of 1D ZnO nanostructures and their corresponding facets 1.3.2 Photoluminescence (PL) of 1D ZnO nanostructures The majority of the reported PL spectra of ZnO nanostructures have been performed at room temperature, though variable temperature PL studies have been measured on some of the samples [9]. Room temperature PL spectra of ZnO typically consist of a UV emission and possibly one or more visible bands owing to defects and/or impurities. In room temperature PL spectra, some variations in the position of the PL peak have been observed for ZnO nanostructures of different sizes as shown in Figure 1.7a and 1.7b [10]. These differences in the peak positions of individual nanobelts, which are large enough so that there could be no quantum confinement effects, indicate that there is likely a different explanation for the variation in the band-edge (BE) emission in ZnO nanostructures reported in different studies. Even though quantum confinement has been proposed as a reason for the blue shift in the band-edge emission with decreasing size [11], any shift due to quantum confinement in nanocrystals with diameters of 24, 38, and 57nm is not likely considering the fact that the Bohr radius of ZnO is 2.34 nm [12]. One possible reason for the variations in the position of the BE emission in various ZnO nanostructures with relatively large dimensions is the different concentrations of native defects. Since the defect density on the surface is higher than in the bulk [13], spectral shifts owing to different defect concentrations are expected to occur in nanostructures of different sizes due to different surface-to-volume ratios. The fact that the decay times in time-resolved PL from ZnO NRs are size dependent [14] is consistent with the 8 assumption of different defect levels/concentrations for structures with different surface-to-volume ratios. Thus, the defects could affect the position of the BE emission as well as the shape of the PL spectrum. Although there have been several reports with weak defect and strong UV emission in ZnO nanostructures [15], in some cases only defect emission [16] is observed or the UV emission is much weaker compared to the defect emission [17]. Therefore, clarifying the origins of different defect emissions is an important issue. However, it should be noted that the ratio of the intensity of UV and defect emission is dependent on the excitation density [18], as well as the excitation area [19]. Thus, neither of the ratios of these two emissions can be used as an absolute determining factor of the crystalline quality of ZnO, although they are useful in comparing the quality of different samples when the measurements are performed under identical excitation conditions. Room temperature PL spectra from ZnO can exhibit several different peaks in the visible spectral region, which have been attributed to the defect emission. Emission lines at 405, 420, 446, 466, 485, 510, 544, 583, and 640 nm have been reported [18]. Several calculations of the native defect levels in ZnO have been reported, as summarized in Figure 1.7c [9]. An example of defect emissions (normalized PL spectra) from different ZnO nanostructures is shown in Figure 1.7d. Green emission is the most commonly observed defect emission in ZnO nanostructures, similar to other forms of ZnO. The intensity of the blue–green defect emission was found to be dependent on the diameter of NWs, but both increased [20] and decreased [21] defect emission intensity with decreased wire diameter were observed. Several different hypotheses have been proposed: Green emission is often attributed to singly ionized oxygen vacancies, although this assignment is highly controversial. Other hypotheses include antisite oxygen [22], which was proposed by Lin et al. [23] based on the band structure calculations. Green emission was also attributed to oxygen vacancies and zinc interstitials [ 24 ]. Cu impurities have been proposed as the origin of the green emission in ZnO [25]. Blue-green defect emission was also reported in Cu doped ZnO NWs [26]. However, although Cu was identified as a possible reason of green emission in ZnO, this cannot explain the defect emission in all ZnO nanostructure samples, especially where defect emission exhibits strong dependence on annealing temperature and atmosphere which would be more consistent with an 9 intrinsic defect rather than Cu impurity. Other hypotheses include various transitions related to intrinsic defects, such as donor–acceptor transitions [27], recombination at Vo** centers [28], zinc vacancy [29], and surface defects. Although the singly ionized oxygen vacancy [30] is a commonly proposed hypothesis, which is supported by the observation of the enhancement of the green defect by annealing it at temperatures above 600ºC [31], this assignment recently has been challenged [25]. The donor–acceptor transition hypothesis used to explain the green and yellow emissions has also been questioned [32]. In addition, while the Zn vacancy hypothesis is supported by the investigation of the effect of O and Zn implantation [33], a blue rather than green emission would be expected based purely on the theoretically predicted energy levels for Zn vacancy [34]. Therefore, the origin of the green emission is still an open and controversial issue and the identification of the exact origin of this emission requires further study. Figure 1.7 (a) Low magnification TEM image showing the size uniformity of ZnO nanobelts. (b) PL spectra acquired from the width= 200 nm wide ZnO nanobelts and the width= 6 nm wide ZnO nanobelts [10]. (c) Illustration of the calculated defect energy levels in ZnO from different literature sources. (d) Room-temperature PL spectra of different nanostructures: 1) Tetrapods, 2) 10 needles, 3) NRs, 4) shells, 5) highly faceted rods, 6) ribbons/combs [9]. 1.3.3 Applications of 1D ZnO nanostructures One dimensional ZnO nanostructures have been intensively investigated for understanding the fundamental phenomena in low dimensional systems and for developing new generation nanodevices with high performance. The following is a review of several typical application examples 1D ZnO nanostructures. UV laser, waveguide, and photodetector: PL spectra show that ZnO NW is a promising material for UV emission, while its UV lasing property is of more importance and interest. The well-facetted ZnO NWs form promising optical resonance cavities which greatly generate highly directional lasing at room temperature in well aligned ZnO NWs as reported by M. H. Huang et al [35]. The additional advantages of ZnO NW lasers are that the excitonic recombination lowers the threshold of lasing, and the quantum confinement yields a substantial density of states at the BEs and enhances radiative efficiency. ZnO NWs are also natural candidates for optical waveguide due to its near-cylindrical geometry and large refractive index (about 2.0). ZnO NWs have been reported as sub-wavelength optical waveguides [ 36 ]. Optically pumped light emission was guided by a ZnO NW and coupled into a SnO2 nanoribbon. These findings show that ZnO nanostructures can be potential building blocks for integrated optoelectronic circuits. Besides UV emitting and lasing, study on utilizing ZnO NWs for UV photodetection and optical switching has been reported by Kind et al [37]. Defect state related visible wavelength detection and polarized photodetection of ZnO NWs were also observed [38]. Photocurrent reaches its maximum when the electric field component of the incident light is polarized parallel to the NW’s long axis. This behavior, one of the characteristics of an 1D system, renders it a promising application in a high contrast polarizer. From the photoconductivity measurements of ZnO NWs, it is observed that O2 has an important effect on the photoresponse, e.g., O2 adsorption on the NW surface could significantly expedite the photocurrent relaxation rate. It is reported that the photocurrent relaxation time is around 8s when exposed to air but hours when placed in a vacuum [39]. Figure 1.8 demonstrates how the desorption–adsorption 11 process of O2 affects the photo response of a ZnO NW. Upon illumination, photogenerated holes discharge surface chemisorbed O2 by surface electron–hole recombination, while the photogenerated electrons significantly increase the conductivity (Figure 1.8c). When illumination is cut off, O2 molecules re-adsorb onto the NW surface and reduce the conductivity (Figure 1.8b). Figure 1.8 Photoconduction in NW photodetectors [39]. (a) Schematic illustration of a NW surface absorbed with oxygen molecules and corresponding energy band diagram. (b) Trapping and photoconduction mechanism in NW. Excited holes are captured by O2- ion, and release O2 by losing an electron. Light emitting diode (LED) and solar cell: ZnO are reported to show n-type semiconductor behavior due to native defects such as oxygen vacancies and zinc interstitials. The difficulty of p-type doping is the major impediment of ZnO for wide ranging applications in electronics and photonics. Using P2O5 as a dopant [40], p-type conduction was reported in ZnO NW arrays grown. However, the p-type conduction was unstable and changed to n-type after two months of storage in ambient air. The larger atomic size of phosphorus than oxygen maybe causes the instability. On the contrary, nitrogen, with an atomic radius similar to oxygen, 0.75 vs. 0.73 Å, is a better choice than phosphorus as a p-type dopant in ZnO NWs as reported by Yuan et al [41]. Successful p-type doping for ZnO nanostructures will greatly improve their future applications in nanoscale electronics and optoelectronics. P-type and n-type ZnO NWs can 12 serve as p-n junction diodes and LED [42]. As an alternative approach, the n-ZnO/p-type inorganic/organic-semiconductor heterojunctions have been investigated for device applications [43,44]. N-ZnO/p-GaN are the mostly widely studied since these materials have similar fundamental band gap energy (~3.4 eV), the same WZ crystal structure, and a low lattice constant misfit of 1.9%. In general, however, p-n heterojunction devices show a lower efficiency than homojunction devices. This is because an energy barrier formed at the junction interface decreases carrier injection efficiency for heterojunction devices with a large band offset. This problem can be solved by making nano sized junctions, where the carrier injection rate significantly increases for nanocontacts [45]. In addition, ZnO is being intensely investigated in excitonic solar cells (XSCs) research recently due to its similarities to TiO2 which is the most commonly used semiconductor oxide in XSCs [46]. ZnO has a conduction band edge located at approximately the same level as TiO2. Compared to TiO2, ZnO has higher electron mobility and longer electron lifetimes than TiO2 which are beneficial for solar cell performance. Another important advantage of ZnO over TiO2 is that a great variety of different morphologies and nanostructured electrodes, especially vertically-aligned NWs can be fabricated by a wide range of synthesis techniques. These ZnO nanowires can provide a higher interfacial area between the donor and the acceptor material with highly-efficient electron transport pathways. In the case of Dye Sensitized Solar Cells (DSCs), it is a possible way to obtain faster electron transport thus improving solar cell efficiency by replacing the nanoparticle electrodes with vertically-aligned nanostructures. Moreover, the application of faster electron transport materials like the vertically-aligned nanostructures is also impelled by the requirement to replace problematic liquid electrolytes in DSC by solid holeconductors with slower kinetics. Figure 1.9 shows a schematic representation of an XSC applying vertically-aligned ZnO nanostructured electrode. The highest efficiency XSCs applying ZnO have only reached 6–7% [47], which is less than the 11.3% obtained with the best DSC applying TiO2 [48]. Thus it is important to have carefully controlled parameters, such as dye or polymer concentration, or pH or sensitization with time, in order to improve solar cell efficiency. 13 Figure 1.9 Schematic representation of an XSC applying ZnO NWs as the electron transport material. In a DSC the ZnO NWs are loaded with an adsorbed layer of a light harvesting Dye and the hole conductor is typically a liquid electrolyte with a I-/I3- redox couple [46]. Piezoelectricity and other applications: Piezoelectricity is one of the most important properties of ZnO. It has been extensively studied for various applications in force sensing, acoustic wave resonator, acousto-optic modulator, and energy harvest, and so on. The origin of the piezoelectricity lies in ZnO special crystal structure, in which the oxygen atoms and zinc atoms are tetrahedrally bonded. In such a non-centrosymmetric structure, the center of positive and negative charges can be displaced due to external pressure induced lattice distortion (Figure 1.10a). This displacement causes local dipole moments, thus macroscopic dipole moments appears over the whole crystal. Actually, among the tetrahedrally bonded semiconductors, ZnO has the highest piezoelectric tensor which provides a large electro-mechanical coupling. The piezoelectric property of ZnO nanostructures was also studied for their potential applications in nonelectric mechanical systems. The piezoelectric coefficient of ZnO nanobelts was measured by AFM with conductive tips [49]. The effective piezo coefficient of nanobelt is found to be frequency dependent and much larger than that of the bulk (0001) surface (Figure1.10b). Recently, utilizing the semiconductor and piezoelectric property possessed by ZnO NWs, novel energy generators-nanogenerators were developed by Z. L. Wang’ group [ 50 ]. These nanogenerators can convert mechanical energy from muscle stretching, body movement or water flow into electricity. These “nanogenerators” could make up a possible new class of self-powered implantable medical devices, sensors and portable electronics. 14 Figure 1.10 (a) Schematic diagrams showing the piezoelectric effect in a tetrahedrally coordinated cation-anion unit. (b) The experimentally measured piezoelectric coefficient d33 for ZnO and its comparison to that of the bulk [49]. In addition, 1D ZnO nanostructures are intensively studied as antireflection layers [51], field emission devices (FED) [52], field effect transistors (FET) [53], nanosensors [54], and the like. More details about the application of 1D ZnO nanostructures can be found in recent reviews [1, 4,55]. 1.4 Fabrication of 1D ZnO nanostructures Various morphologies of 1D ZnO nanostructures have been synthesized successfully by various experimental methods. The most commonly used methods are vapor transport synthesis and solution synthesis. 1.4.1 1D ZnO nanostructures from vapor transport synthesis The vapor transport process is mostly utilized to synthesize ZnO nanostructures. During the process, Zn and oxygen or oxygen mixture vapor are transported and react with each other, forming ZnO nanostructures. According to the difference on nanostructure formation mechanisms, the vapor transport process can be categorized into the catalyst free vapor-solid 15 (VS) process and catalyst assisted vapor-liquid-solid (VLS) process. Vapor-solid process: Synthesis utilizing VS process is usually capable of producing a rich variety of nanostructures [4, 55], including NRs, NWs, nanobelts, and other complex structures. Without the aid of metal catalysts, the VS growth has been mainly used to synthesize metal oxide and some semiconductor nanomaterials. Plausible growth mechanisms such as the anisotropic growth, defect-induced growth (e.g., through a screw dislocation), and self-catalytic growth have been suggested based on electron microscopy studies. According to the classical theories of crystal growth from liquid or vapor phases, the growth fronts play a crucial role for the deposition of atoms. There are two kinds of microscopic surfaces: (1) rough surfaces on which atoms of about several layers are not well arranged. Deposition of atoms is relatively easy compared to a flat surface and crystal growth can continue if enough source atoms are continuously provided; (2) atomically flat surfaces on which atoms are well arranged. Atoms from the source have a weak bonding with flat surfaces and can easily return to the liquid/vapor phase. Atoms deposition occurs only on the atomic steps. There are three ways to generate atomic steps on a flat surface: (1) nucleation of new two-dimensional islands which is difficult because the nucleation barrier is high, and there is almost no super-cooling. The islands will be exhausted eventually (see Figure 1.11a and 1.11e); (2) screw dislocations which generate atomic steps to help atoms to deposit continuously (Figure 1.11b and 1.11f); and (3) twining structures which contain ditches at the cross of two grain surfaces. Atoms deposit at the ditches resulting in atomic steps along twining surfaces. The resulting growth can be continuous along the direction of the twining plane (Figure 1.11c and 1.11g). Followings are important factors for the nanocrystal growth in the VS process. 16 Figure 1.11 (a) NRs formed due to anisotropic growth of ZnO crystals. (b) Unidirectional growth of ZnO single crystals due to screw dislocation. (c) Growth induced by twining. (d) Self-catalytic growth of ZnO NWs by Zn droplets. (e) ZnO crystals contain no catalysts and defects). (f) ZnO whiskers growth due to dislocations. (g) ZnO bi-crystal growth due to twining. (h) Zn or Zn-rich phase observed on the tips of ZnO NWs [56]. 1) Internal anisotropic surfaces Because of anisotropic properties of different surfaces in a ZnO crystal, such as the preferential reactivity and binding of gas reactants on specific surfaces and all crystals tend to minimize their total surface energy, rod- or wire-like shapes are frequently resulted. However, the degree of the anisotropic properties of crystals is not significant large, highly anisotropic growth (i.e., the length-to-diameter ratio >100) of nanocrystals at or near the thermal equilibrium state is not expected. 2) Crystal defects Screw dislocations (the well known Burton–Cabrera–Frank theory) are known to significantly enhance the crystal growth of metals and some molecular materials [57]. This classical mechanism is based on the fact that the growth of a crystal proceeds by adding atoms at the kink sites of a surface step. Kink sites always exist on the steps even at the thermal equilibrium state. Due to the advance of the kink along the surface by the addition of atoms, the 17 crystal grows perpendicularly to the surface. In thermal equilibrium state, a perfect crystal should eventually contain no surface steps. Then, the growth of a perfect crystal depends on the nucleation of surface steps. For the growth of a real crystal, however, the growth rate is much faster than that predicted for a perfect crystal because real crystals contain defects, e.g., dislocations and twins. A dislocation cannot terminate inside a perfect crystal. They can terminate on a defect inside the crystal or on a surface. If a dislocation ends on a surface and its Burgers vector has a component normal to the surface (the screw component), a step forms starting from the emerging point of the dislocation. Leading by the dislocation, steps can winds into a spiral, and the growth of the crystal is largely enhanced without the need of nucleation for fresh surface steps. There are many reasons for the formation of a dislocation in a crystal. For Si NWs, oxygen atoms may cause the nucleation of a dislocation [58]. It has been frequently observed that screw dislocations are associated with growth of crystal in the dendrite or whisker geometries. In ultra-thin NWs, so far no screw dislocations have been evidenced. However, in thick wires, for example ZnO NWs (diameters > 200 nm), unidirectional growth induced by dislocations in VS growth mode has been observed (Figure 1.11f). The spiral feature at each whisker tip is obviously due to the steps generated by a screw dislocation. In thin ZnO NWs grown by the VS growth, however, no screw dislocations existing at the core of the NWs have been found. 3) Self-catalytic growth Self-catalytic growth of ZnO NWs has been proposed based on the fact that Zn vapor can be extracted from the ZnO vapor phase by heating ZnO powder under vacuum conditions or by a carbon thermal reaction (heating the mixture of ZnO powder and carbon powder). Figure 1.12a and 1.12b show the schematics of a typical self-catalytic growth process. The formation of a Zn droplet occurs in the gas flow or on the substrates, followed by the nucleation and growth of a solid ZnO NW due to the supersaturate ion of the liquid droplet. Incremental growth of the NW taking place at the droplet interface constantly pushes the Zn droplet upwards. However, the zinc nanoparticles, as evidence of catalytic growth, are rarely observed at the ends of ZnO NWs. It is proposed that the Zn droplets as catalysts are consumed or evaporate because of high deposition temperature and the highly saturated vapor pressure of Zn. 18 Though the self-catalytic growth mechanisms of the VS growth are complicated and unclear, many interesting morphologies of ZnO nanostructures have been produced by this method. However, this approach obviously provides less control on the geometry, alignment and precise location of ZnO nanostructures because of the poor control of the ZnO nucleation of current synthesis techniques. Vapor-liquid-solid deposition Controlled growth of 1D ZnO nanostructures has been achieved by the catalyst assisted VLS process. In this process, various nanoparticles or nanoclusters have been used as catalysts [59], such as Au, Cu, Ge, and Sn, and so on. Figure 1.12c shows a schema of a typical VLS process. The formation of an eutectic alloy droplet occurs at each catalyst site. The alloy droplets absorb Zn and O from the vapor phase resulting in the supersaturation of ZnO. As a consequence, crystal growth occurs at the liquid-solid interface by precipitation and NW growth commences. Thus, such a growth method inherently provides site-specific nucleation at each catalytic site. Based on the VLS mechanism, the diameter of NWs can be tuned by using different sizes of nanoparticles or nanocluster catalysts. In addition, the control of NW growth location and alignment has been realized by using patterning techniques and choosing proper epitaxy substrates. Figure 1.12 Schematics of (a) and (b) the typical self-catalytic growth based on the VS process; 19 (c) The catalyst assisted VLS process. 1.4.2 1D ZnO nanostructures from solution synthesis The major advantages of the solution-based technique (in aqueous or non-hydrolytic media) for synthesizing nanomaterials are high yield, low cost and easy fabrication. The solution-based technique has been demonstrated as a promising alternative approach for mass production of metal, semiconductor and oxide nanomaterials with excellent controls of the shape and composition with high reproducibility. The ZnO nanocrystals synthesized in aqueous media may often suffer from poor crystallinity, but those synthesized under nonhydrolytic conditions at a high temperature, in general, show much better crystal quality [60]. For the formation of 1D ZnO nanostructures from solution, several growth mechanisms have been developed, such as self-assembly attachment growth [61], anisotropic growth [62] of crystals by thermodynamic or kinetic control and structural directed growth by templates [63]. The anisotropic growth of ZnO crystals induced by different surface energies can lead to the formation of elongated nanocrystals. However, the differences in the surface energies of most materials are not large enough to cause highly anisotropic growth of long NWs. By adding surfactants as soft templates [64] to the reaction solution, some surfaces of nanocrystals can be modulated, i.e., the surfactant molecules selectively adsorb and bind onto certain surfaces of the nanocrystals and thus reduce the growth of these surfaces. This selective capping effect induces the nanocrystal elongation along a specific direction to form NWs. Through DC or AC electrochemical deposition, ZnO can be introduced into the nanochannels of the hard template [63], such as anodized alumina membranes containing nanosized channels, track-etched polymer porous membranes, and some special crystals containing nanochannels. As a matter of fact, in many cases, the structural directors may not exist or the growth process can provide self-constitutive templates. The formation mechanism of ZnO NWs in solution is complicated and the selection and function of the structural directors require further and systematic investigation. 20 1.4.3 Patterned and aligned growth of ZnO 1D nano-arrays The growth of patterned and aligned 1D ZnO nanostructures shows great promise for applications in sensors [3], antireflection [5], room-temperature UV lasers [35], nanogenerators [50] and FED [52]. For realizing further nanodevice applications in numerous biotechnology and optoelectronics, it is essential that the periodicity and patterns could be controlled and designed with deliberate control over interfeature distance, positions, shape, and orientation controlled ZnO NR/NW arrays. Traditionally, aligned 1D ZnO nanostructures are achieved on lattice matching substrates using a catalyzing metal particle, often gold, in vapor transport processes, as growth sites to initiate or guide the growth. It is therefore crucial to be able to control the position of the growth sites for the growth of periodic 1D nanostructure arrays. There have been efforts to obtain nanoscale-patterned metal catalysts [65] by focused ion beam patterning, electron-beam lithography (EBL), dip-pen nanolithography, nanoimprint lithography, scanning tunneling microscope lithography, mask lithography via porous alumina, and self-assembled micro- or nanospheres. Figure 1.13a and Figure 1.14a-b show the general growth processes for and examples of the patterned and aligned ZnO NWs by the EBL technique through the vapor transport growth, respectively. Though excellent controllable vertical growth of ZnO 1D nano-arrays have been obtained by these patterned techniques, there are still some apparent shortcomings: 1) time and cost-consuming preparation processes; 2) special requirements for lattice matching substrates, such as Al2O3, GaN, and SiC, and the like [65], which are all expensive; 3) low availability for large scale synthesis limited by the small area operation properties of these patterned techniques; 4) the contamination resulting from the introduction of metal catalysts. It seems that some solution processes using [0001] textured ZnO nanoparticle seeds/ZnO thin film can overcome these shortcomings. Figure 1.13b and Figure 1.14c-d show the general processes and examples for the patterned and aligned 1D ZnO NWs by EBL techniques through solution based growth. However the products have poor crystallization due to the low growth temperature with the risk of introducing impure ions. Therefore the mass production of high-quality patterned and highly aligned ZnO NWs at low cost is still a challenge for nanotechnologists and an important research issue in this thesis. 21 Figure 1.13 Schematic diagrams depicting the patterned growth of NWs by EBL through (a) vapor transport growth and (b) solution based growth. Figure 1.14 The Scanning Electron Microscopy (SEM) image of aligned and partnered ZnO NWs from patterned growth sites fabricated by EBL: (a) and (b) by Au-assistant VLS growth [65] and (c)-(d) by solution synthesis [66]. 22 Chapter 2 Structural and Optical Characterization Techniques Scanning electron microscopy and transmission electron microscopy (TEM) are two useful characterization techniques to understand the morphology, structure and composition of as-synthesized nanomaterials. Optical characterization is an important problem as ZnO nanostructures are excellent emitters. PL the most often used technique to investigate their optical quality. In this chapter, the discussion is, however, limited to techniques which are frequently used in the following studies such as TEM and PL. 2.1 Transmission electron microscopy Transmission electron microscope is powerful equipment for the characterization of materials. TEM provides diffraction pattern, diffraction contrast image and phase contrast (high-resolution) image. If equipped with attachments, such as electron energy loss spectroscopy (EELS) and energy dispersive X-ray spectroscopy (EDX), it can provide elemental information of the materials. As one of the most powerful tools in nanotechnology, TEM has played an important role in characterizing 1D nanostructures, not only in determining crystal and surface structure, including growth direction, side/top surfaces, surface polar direction, surface reconstruction, point defects, but also chemical structure. 2.1.1 Structure and operation modes of TEM TEM operates on the same basic principles as the optical microscope but uses electrons instead of light as luminescence source. The resolution of a light microscope is limited by the wavelength of light while TEM using much lower wavelength electron (e.g., λ = 0.00251 nm for an accelerating voltage of 200 keV) as source makes it possible to get much smaller resolution in atomic scale. Accordingly, electromagnetic lenses rather than optical lens are used in the TEM to focus the electrons by creating an electron-magnetic field. TEM consists of illumination system, specimen stage, imaging system, magnification system and data recording system. The structure of TEM is illustrated in Figure 2.1. The illumination system comprises 23 the electron gun and condenser lenses. Our Philips CM120 and JEOL 2011 TEM use LaB6 thermionic filament and JEOL 2011F TEM works with a field emission gun. The accelerated electrons are emitted from the electron gun. The divergent electron beam then be demagnified and focused on the specimen by the condensed lens below the electron gun. The imaging system, i.e., the objective lens, is most critical since it determines the resolving power of the instrument and performs the first stage of imaging. The specimen stage is inserted in pole piece gap of objective lens. In this critical region the incident electron beam interacts with specimen. The objective lens forms an inverted initial image in the image plane and also a diffraction pattern in the back focal plane. The magnification system consists of the intermediate and projection lenses that magnify the image or diffraction pattern further. The data recording system includes a fluorescent screen, photographic films and a CCD camera. The magnified image or diffraction pattern can be observed directly in the screen or photographed in the film or recorded with CCD camera. In addition, our JEOL 2010F TEM is equipped with an Oxford Links EDS analysis system and a Gatan ENFINA EELS system for chemical composition analysis. The working principle of EDS and EELS can be found in the book written by Daisuke Shindo & Tetsuo Oikawa [67]. 24 Figure 2.1 Schematics of TEM consisting of five systems: illumination, specimen stage, imaging, magnification and data recording systems. The enlarged parts on the right are EDX and EELS for chemical composition analysis. There are basically two operation modes of TEM: imaging and diffraction modes, which provide information in real space and reciprocal space, respectively. The ray paths of imaging and diffraction modes are illustrated in Figure 2.2. The parallel electron beam illuminates a crystalline specimen and reacts with the specimen. If the sample is thin enough, some of the electrons are transmitted, some are scattered. The objective lens forms a diffraction pattern on the back focal plane with the electrons scattered and combines them to generate an image on the image plane. This is the first image present in the ray path named as the 1st intermediate image. The image or diffraction pattern formed is then further magnified by the intermediate and project lenses. The former is called the imaging mode (see Figure 2.2b), and the latter the diffraction mode (see Figure 2.2a). Switching from the imaging mode to diffraction mode is easily achieved by changing the strength of the intermediate lens. 25 Figure 2.2 Ray paths in TEM (a) diffraction mode and (b) imaging mode. 2.1.2 Diffraction in the TEM There are two ways to obtain a diffraction pattern from a specific area in the specimen, as illustrated by the ray diagrams in Figure 2.3. We can either select the area using a selective aperture which encloses only the area of interest or focus the electron beam onto the area of interest. The former is known as selected area electron diffraction (SAED) and the latter convergent beam electron diffraction (CBED). For SAED, the aperture is put in the image plane that creates a virtual aperture in the specimen (see Figure 2.3a for illustration). Figure 2.3 Ray diagrams showing (a) SAED and (b) CBED pattern formation respectively. In this thesis, CBED is frequently used for the polarity determination of 1D ZnO nanostructures. The differently charged surfaces show distinct reactivity and thus display different growth behaviors. Experimentally determining if the surface terminates with Zinc or oxygen is important for understanding the polar surface dominated growth phenomena of 1D ZnO nanostructures. The CBED with a strong dynamic diffraction effect is a unique tool used to answer this question. CBED patterns are formed with a converged electron probe illuminating the sample area, which can be as small as a few nanometers. The convergence angle is equivalent to imaging the crystal through a range of directions, so is sensitive to the 3D crystal structure of the sample. The beam convergence determines the size of the diffraction 26 disks. The cross point of the converged beam can be at or below the specimen, depending on the quality of the specimen and the required convergence angle. The incident probe consists of many plane-wave components propagating along different directions, thus forming a converged conical electron probe (Figure 2.4). For an incident beam P, diffraction results in a complete point diffraction pattern consisting of P’s as ruled by the Bragg reflection law. A similar point diffraction pattern set is formed for another plane-wave component Q. Therefore, for cases where there are no disk overlaps, a perfect registration is retained between each incident beam direction and the diffracted beams. The intensity distribution of the disks in the CBED pattern contains the local symmetry and point group information of the electron beam illustrated area. More details and procedures have been given by Spence [68] and Buxton [69] and Tanaka [70] et al. Figure 2.4 Ray diagram of CBED for determining the polarization of ZnO 2.1.3 Imaging in the TEM Diffraction contrast imaging Diffraction contrast is simply a special form of amplitude contrast because the scattering occurs at special (Bragg) angles. Crystalline specimens usually give a single-crystal diffraction pattern. Diffraction contrast can be achieved by placing an objective aperture at the back focal 27 plane. Bright-filed (BF) image is formed by placing the objective aperture round the direct beam (Figure 2.8a) and dark-filed (DF) image comes from any of the diffracted beams (Figure 2.8b). Diffraction contrast is useful in identifying large structures and crystallographic features. Figure 2.5 Comparison of the use of an objective aperture in TEM to select (a) the direct or (b) the scattered electrons forming BF and DF images, respectively. Imaging in the High-resolution TEM High-Resolution TEM (HRTEM) allows the imaging of the crystallographic structure of a sample at an atomic scale. Because of its high resolution, it is an invaluable tool to study nanoscale properties of crystalline material such as semiconductors and metals. Contrast in HRTEM arises from the interference in the image plane of the electron wave with itself. The phase of the electron wave carries the information about the sample and generates contrast in the image, thus the name phase-contrast imaging. This however is true only if the sample is thin enough so that amplitude variations do not contribute to the image (the so-called weak phase object approximation, WPOA). The interaction of the electron wave with the crystallographic structure of the sample is not entirely understood yet, but a qualitative idea of the interaction can readily be obtained. Above the sample, the electron wave can be approximated as a plane wave incident on the sample surface. As a result of the interaction with the sample, the electron exit wave right below the 28 sample ψ e (u) as a function of a multitude of diffracted beams with different in plane spatial frequencies u. The exit wave firstly propagates to the front surface of the object lens, secondly passes through the lens to reach the back surface of the lens, thirdly reach the back focal plane of the lens to form the diffracted wave and finally reach the image plane to form the image waveψ i (u) . The image formed in the image plane of the objective lens is further enlarged by the intermediate lens and projected lens. 2.2 PL spectroscopy 2.2.1 Introduction of luminescence Bombarding the surface of a material with some incident radiations or particles may result in the emission of electromagnetic radiation beyond that produced by thermal black body radiation. This emission can be in the visible range (400-700 nm), UV (<400 nm) and infrared (IR; >700 nm). This general phenomenon is known as luminescence. Solid state band theory provides a way to explain the luminescence phenomenon. Semiconductor materials can be visualized as having a valence band and a conduction band with an intervening band gap (forbidden gap) (Figure 2.6a). If a crystal is bombarded by incident radiation or particle with sufficient energy, electrons from the lower-energy valence band are promoted to the higher energy conduction band (Figure 2.6b). When the energetic electrons attempt to return to the ground state valence band, they may be temporarily trapped (on the scale of microseconds) by intrinsic (structural defects) and/or extrinsic (impurities) traps (Figure 2.6c). If the energy lost when the electrons vacate the traps is emitted is in the appropriate energy/wavelength range, luminescence will result. Most of the photons fall in the visible portion of the electromagnetic spectrum (wavelengths of 400-700 nm) with some falling in the UV and IR portions of the electromagnetic spectrum. 29 Figure 2.6 (a) Band diagram of semiconductor. (b) Electrons are excited from VB to CB (c) Electron transition from CB to VB. There are several possible ways in which the traps can interact to produce luminescence (Figure 2.6c). Once the electrons are excited to the conduction band they may not encounter a trap and fall to the valence band or they move randomly through the crystal structure until a trap is encountered. From that trap, the electron might return to the ground state or it may encounter multiple traps emitting photons with wavelengths dependent on the energy differences. The intensity of the luminescence is generally a function of the density of the traps. Possible paths of electrons as they fall back to ground state of the valence band include as shown in Figure 2.6c: (left) electron falling directly back to the valence band, generally emitting UV radiation, (middle) electrons encountering a single trap, emitting luminescence proportional to the energy released from the temporary occupancy of the trap by the electron as it falls to the valence band and (right) electrons encountering multiple traps, emitting luminescence proportional to the energy released from the temporary occupancy of the traps by 30 the electron as it falls to the next trap or valence band. Intrinsic luminescence is characteristic of the host lattice. It can be due to non-stoichiometry (vacancies), structural imperfections (poor ordering in the crystal, radiation damage, shock damage, etc.) and impurities (non-activators that distort the lattice). Extrinsic luminescence results from impurities in the structure. The impurities generate luminescent centers and are most commonly transition elements, rare earth elements and actinide elements (due to the occurrence of valence electrons in either "d" or "f" orbitals). 2.2.2 Photoluminescence Photoluminescence is the spontaneous emission of light from a material under optical excitation. The excitation energy and intensity are chosen to probe different regions and excitation concentrations in the sample. PL investigations can be used to characterize a variety of material parameters. PL spectroscopy provides electrical (as opposed to mechanical) characterization, and it is a selective and extremely sensitive probe of discrete electronic states. PL analysis is nondestructive. Indeed, the technique requires very little sample manipulation or environmental control. Because the sample is excited optically, electrical contacts and junctions are unnecessary and high-resistivity materials pose no practical difficulty. In addition, time-resolved PL can be very fast, making it useful for characterizing the most rapid processes in a material. The fundamental limitation of PL analysis is its reliance on radiative events. Materials with poor radiative efficiency, such as low-quality indirect bandgap semiconductors, are difficult to study via ordinary PL. Similarly, identification of impurity and defect states depends on their optical activity. Although PL is a very sensitive probe of radiative levels, one must rely on secondary evidence to study states that couple weakly with light. Our PL set-up is shown in Figure 2.7. Lens L1 is used to focus the laser beam onto the sample and Lens L2 is use to collect the excited PL. Filter F1, a band-pass filter, is used to chop off the unwanted light except the laser beam with the desired wavelength. Filter F2 is a low-pass filter, which is placed in front of the entrance slit of the spectrometer to filter out the original laser beam scattered by the sample surface. Because the measurement does not rely on electrical excitation or detection, sample preparation is minimal. This feature makes PL 31 particularly attractive for material systems having poor conductivity or undeveloped contact/junction technology. Measuring the continuous wave PL intensity and spectrum is quick and straightforward. Figure 2.7 The experimental set-up for PL measurements For spatially resolved PL studies of nanostructures, near field optical microscope (NSOM) [71] is powerful for it breaks the far field resolution limit by exploiting the properties of evanescent waves. This is done by placing the detector very close (<< λ) to the specimen surface. This allows for the surface inspection with high spatial, spectral and temporal resolving power. With this technique, the resolution of the image is limited by the size of the detector aperture and not by the wavelength of the illuminating light. In particular, lateral resolution of 20 nm and vertical resolution of 2-5 nm has been demonstrated 32 Chapter 3 1D ZnO Nanostructures Fabricated with Vapor Transported Process Vapor phase synthesis is probably the most extensively explored approach to the formation of 1D nanostructures such as whiskers, NRs and NWs. Although the exact mechanism for 1D growth in the vapor phase is still not clear, this route has been used by many research groups to fabricate 1D ZnO nanostructures. In this chapter, we report our efforts in synthesizing ZnO nanostructures based on vapor transported process. We realized control on morphology, alignment, assembly and quality of 1D ZnO nanostructures by this approach. The growth results depended largely on the fabrication methods. The three methods introduced below all supply Zn vapor and O2 as reactants in different ways. The supply speed of reactants, growth atmosphere, temperatures, pressures, substrates and catalysts are all key factors for the final growth results. 3.1 Method I: Direct oxidation of Zn metal at high temperature in air Most methods for 1D ZnO nanostructures are relatively complicated with low yield and not suitable for commercial production. Considering that electrical and optical properties of nanomaterials depend sensitively on both shape and size, it is important to obtain the expected shape and size in a controllable way with high yield. Herein, we report the fabrication of ZnO NWs and tetrapods in air by directly oxidizing Zn metal at a high temperature [72]. This process is fast, simple, has a high yield and is of interest for industrial-scale applications. We take spatially-resolved PL measurements on single NWs and ZnO tetrapods, grown simultaneously in the same apparatus. The very different PL observed in these two structures strongly suggest that the defects leading to green luminescence (GL) originates from structural changes in the tetrapods, and not surface related as previously believed. Experiment section ZnO tetrapods and NWs were fabricated by thermal evaporation of pure Zn powder without using any catalyst. Five gram of Zn metal powder was placed at the center of a quartz tube (50cm long) which has been heated up to about 1200°C. The ends of the quartz tube were open to air so that Zn vapor reacted with oxygen forming ZnO and deposited at low temperature 33 regions (see Figure 3.1a). The reaction time is less than fifteen minutes. As shown in Figure 3.1b-c, tetrapods and NWs were formed exclusively in regions at temperatures ~1000°C and 800°C, respectively. ZnO tetrapods and NWs were dispersed on Si plates and spatially-resolved PL spectra were recorded through a NSOM system (Nanonics Inc., Israel) at room temperature. Excitation radiation at 325 nm from a He-Cd laser via a tapered optical fiber with a 200 nm aperture illuminated the sample from the top. The excitation area on the sample is about 1μm×1μm. To determine the positions accurately, topological image with this same tip was first obtained and then used as spatial reference for spatially resolved PL measurements. We chose a ZnO tetrapod with two legs broken off for the PL measurement. The resultant V-shaped sample could lie flat, and in addition emission from the core could be easily detected. It has an obvious broken surface at the core region and the angle between the two branches of the V-shaped ZnO is about 109°, consistent with the angle for a normal tetrapod. Figure 3.1 (a) A sketch map of reaction apparatus and the deposition areas for tetrapods and NWs. (b) and (c) the SEM images of the ZnO structures formed by oxidation of Zn at different 34 positions of the tube. X-ray diffraction (XRD) patterns (d) and (e) corresponding to the tetrapods and uniform NWs in b and c, respectively. Morphology and structure characterization Figure 3.1d-3.1e show the XRD spectra of as-synthesized ZnO nanostructures, which can be indexed as WZ ZnO. The diffraction peaks are sharp and no other phases can be found. This indicates that these nanostructures have very good crystallinity. The XRD spectrum of the as-prepared ZnO tetrapods is similar to the standard spectrum of ZnO powders for no preferred alignment, while the as-prepared NWs grow in the [0001] direction and they tend to lay down with the c-axis perpendicular to the substrate when taking an XRD measurement, resulting in a rather weak (0002) peak. Figure 3.2a shows the TEM image of as-synthesized ZnO NWs that are well dispersed on carbon film. The NWs are uniform in diameter and very clean. The HRTEM study in Figure 3.2b indicates the NW is a WZ structure growing in the 0001 direction and their surface is quite smooth. No stacking faults are found, indicating the quality of these NWs is quite high. Figure 3.2 (a) The TEM image of ZnO NWs. (b) The HRTEM image recorded on one single ZnO NW with zone axis [1 100] . The inset is the Fast Fourier Transform (FFT) pattern of the HRTEM image. ZnO tetrapods, however, consisted of a cubic zinc blende (ZB) core and four hexagonal WZ legs [73] as shown in Figure 3.3. The single crystalline cubic cores were responsible for the nucleation and growth of the tetrapod structures. This is because a cubic crystallite core has 35 totally eight {111} facets, and four of these {111} facets are Zn-terminated. These four facets are highly reactive and can develop into hexagonal legs with their {0001} planes parallel to the {111} planes of the cubic core. This conclusion is made according to our CBED study, which is conventionally used in TEM to determine the polarity of semiconductor compounds. The CBED patterns are obtained with a converged electron probe focusing at the interested area of the sample, the details of which can be found in Ref. [74]. The right side picture in Figure 3.3d illustrates the CBED pattern taken along the [1 100] zone axis of one WZ leg as marked by the arrow. This experimental CBED pattern matches fairly well with the simulated pattern (produced using JEMS simulation software) in Figure 3.3d. The best match is found for the sample thickness near 116 nm with electron beam energy of 120 kV. We have investigated a number of legs by using CBED and obtained the same result, namely, that all legs of the tetrapods elongate with Zn-terminated {0001} polar surfaces as the fast growth fronts. Both ZB and WZ are closed packed structures. The only difference between these two structures is the stacking sequence of the same atomic layer along the {111} planes. Stacking faults frequently occur as the main defects in both ZB and WZ crystals. We have observed a high density of stacking faults near the ZB to WZ transition region as shown in Figure 3.3c. 36 Figure 3.3 (a) and (b) Low magnification TEM images of a ZnO tetrapod. (c) A high resolution TEM image of the core region. (d) A CBED pattern from a leg of the tetrapod. In addition, self-assembly of ZnO tetrapods has been observed as shown in Figure 3.4, in which the branches of different tetrapods connect along c-axis. The complexes consisting of two, three and even four tetrapods are found. From the Figure 3.4a and 3.4b, it seems that the complexes result from the occasionally meeting and fusion of branch ends of tetrapods. The connection part was studied by DF technique to confirm our speculation. Figure 3.5a and 3.5b give the BF and DF TEM images of one single complex built by two tetrapods, respectively. The former do not provide any useful information while the latter show the place of joint by discontinuous contrast profile (Figure 3.5c). The symmetrical CBED patterns recorded on the parts on the both sides of the joint suggest the formation of an inversion domain boundary (IDB) [75] with Zn-terminated (0001) polar planes face to face between these two tetrapods. Further HRTEM study of the IDB in Figure 3.5d reveals large amounts of stacking faults at the joint and the two crystals between the joint have the same orientations. Hence we believe the complexes result from the following formation process: ZnO tetrapods grow and float in air flow; some branches of ZnO tetrapods encounter by chance and fuse together; since these tetrapods float in air flow, they freely rotate with each other during the fusing process. Apparently, when the two branch crystals have the same orientation, the complex is stable with the lowest energy. Figure 3.4 TEM images of complexes consisting of ZnO tetrapods 37 Figure 3.5 (a) BF TEM image and (b) DF TEM image of one single complex built by two tetrapods. The insets in (b) are CBED patterns and simulated ones (below). (c) The enlarged image of the circled part. (d) HRTEM image of the IDB. Structure related PL properties Figure 3.6 shows the PL spectra of the NW ensemble and one single NW. There is only a strong and sharp BE emission and almost no visible GL emission. Taking logarithm of the PL spectrum of the NW ensemble to the base 10, a weak green emission at 525 nm is observed while there is still no apparent visible emission by the same treatment of PL spectrum of a single ZnO NW. Therefore, the visible emission from the NW ensemble comes from purities, not from single ZnO NWs. Furthermore, the same PL spectra are obtained at different positions of one single NW, indicating high uniformity of the PL property of the NW, which we believe results from the uniformity in the size and structure of ZnO NWs. 38 Figure 3.6 The PL spectra of (a) the NW ensemble and (b) one single NW For the tetrapod samples, the situation is different. They show size and structure changes in space. The diameter of the leg of the tetrapod shown in the inset of Figure 3.7b is about 1μm at the core and gradually decreases to 400nm at the end of the leg. Figure 3.7a shows the PL spectrum of the tetrapod ensemble, which is similar to those taken on one single tetrapod. Figure 3.7b is the PL spectrum taken at the end of the leg, the middle of the leg, and at the core 39 of one single tetrapod (marked as A, B and C respectively in the inset.) These spectra are normalized to the intensity at the BE peak to focus on the change in the intensity ratio GL/BE along the legs of the tetrapod. We have studied a number of such samples, and all of them (and all the legs) show the same trend depicted in Figure 3.7a-b, namely: (i) at the (broken) base, a number of distinct peaks are seen, and (ii) the GL/BE ratio decreases when moving towards the end of the leg. Finally, as shown in Figure 3.6b, the PL from a single uniform NW grown in the same apparatus shows no detectable GL at all. There is also a blue shift of the BE as the diameter of the leg decreases: the peak positions of BE are at 382.85nm, 382.48nm and 377.79nm at C, B and A, respectively, similar to that reported earlier [76]. Given the very small size of the ZB core (less than 10 nm, see Figure 3.3), we attribute the observed PL in all cases to originate from the WZ legs. Figure 3.7 Normalized PL spectra at different positions, A, B and C, of a tetrapod. The dotted line corresponds to point A, the dashed line to point B and the solid line to point C. The inset shows a SEM image of the chosen tetrapod and the three positions A, B and C on it. The origins of GL and yellow luminescence (YL) in ZnO are generally attributed to various types of point defects. These defects (for Zn and for O respectively) include: vacancy (VZn, VO), interstitial (Zni, Oi), and antisite substitution (ZnO, OZn) [77]. The three peaks observed at position C, centered at about 490nm, 530nm and 585nm, can be tentatively 40 accounted for as follows. For GL, the peak at 490nm (2.53eV) can be associated with oxygen vacancy defects VO, and the peak at 530nm (2.34eV) can be associated with antisite defects OZn level (2.38eV) or complex defects VO: Zni (2.4eV) [78]. The YL peak near 585nm (2.12eV) can be attributed to oxygen interstitials Oi as in bulk ZnO [77]. We note that the peak wavelengths vary slightly from sample to sample, probably due to the varying distributions of the defects. We now discuss the implications of the very different PL spectra for our tetrapod and NW samples, grown in different parts of the same apparatus. We first stress that with our growth method, both the tetrapods and the NWs were formed in a zinc-rich, oxygen-deficient environment which is more conducive to formation of defect types Zni, and VO as compared to types Oi, OZn and VZn. We further note that NWs formed in such an environment (at ~800C) appear to have few of these defects. In contrast, PL from tetrapods strongly indicate the presence of most or even all of the defects types among which Zni, VO, and OZn give rise to GL, and Oi gives rise to YL. In our apparatus, these tetrapods are formed at a higher temperature (~1000ºC) and, being closer to the Zn source, in even more zinc–rich environment compared to the NWs formed further downstream. The question then arises as to why in such a zinc-rich environment, there should be a significant amount of defect types OZn and Oi. It is also apparent from spatially-resolved PL measurements on the tetrapod that there are more defects at the core, and the amount decreases going towards the tip of the leg. These observations become understandable once we take into account the way the tetrapods are formed: a ZB cubic core is formed first, from which the four WZ structured legs grow, leading to the tetrapod morphology. As mentioned above, TEM studies of these tetrapods indicate the presence of stacking faults, particularly in the transition regions from the ZB to WZ structure. It is therefore most reasonable that point defects of all types will also be present in the same region. Such point defects, invisible to TEM, manifest themselves in the GL and YL. These point defects are formed primarily due to the change in structure, rather than the relative availabilities of Zn and O atoms; thus, point defects of all types can equally likely form, including Zni, VO, OZn and Oi responsible for the observed GL and YL at the core. Further down towards the tip, away from the core where the source of the stacking faults is, one would expect decreasing densities of defects of all types resulting in weaker GL and YL. For NWs, there was no structural change and consequently no detectable defects of any type under this growth condition. We note that multiple defect emission peaks have previously been reported for tetrapod ensembles [79] and ZnO bulk [80]. Our data for a single tetrapod show that the multiple peaks are not from different tetrapods but rather from a single piece of nanostructure, indicating that the defects vary from the core to the tip of the legs. Further, this trend is the same for all the 41 tetrapods. However, our observations are opposite to that observed using CL on single tetrapods of similar dimensions [ 81 ]. In their case, the GL/BE ratio is proportional to the surface-to-volume ratio, increasing as the diameter of the tetrapod legs tapers to their ends, consistent with the notion of a constant surface defect density. The PL spectra from our tetrapod samples show the opposite trend: the GL/BE ratio decreases as the tetrapod legs taper down. Further, as shown in Figure 3.6, the much thinner NWs grown further downstream (at lower temperatures) in the same apparatus show no detectable GL. Clearly larger surface-to-volume ratio does not necessarily bring in more intense GL, and the notion of surface defects leading to GL is not supported in our case and some others [82]. We instead believe the point defects might be more likely due to the change in growth orientation from the core (ZB) to the legs (WZ) and exist mainly in the bulk. While the cause for the differences between our results and that of Ref.81 is not clear, it is likely due to the different growth conditions. As for the difference between the CL and PL techniques, we note that the 3.815eV UV photons used in PL is much less energetic than the keV electrons used in CL, and therefore much less likely to induce defects in the sample under observation. 3.2 Method II: Carbon thermal method The carbon thermal method is another simple and effective technique of synthesizing 1D nanostructures, especially single crystalline ZnO NWs and nanobelts. This method supplies Zn gas by the reaction between carbon and ZnO, which greatly lowers the reaction temperature compared to the thermal evaporation method (described in Chapter 3.3). Our early group work using this method does not introduce catalyst and carrying gas [83]. The nanostructure products grow on the inner walls of the quartz tube. The morphologies of the products include NWs, nanobelts and nano urchins. However replicating and controlling the products’ morphologies is difficult because of the reasonably fast reaction speed and uneven growth atmosphere due to the narrow reaction space [41]. Other researchers develop this method by introducing carrying gas (Ar and O2) and catalysts in the normal tubular stoves with adequate reaction space. Here, we report the fabrication of ZnO micro & nano pyramids by this method with Au as catalysts. These pyramid-like products consist of a (0001) plane and two high-indexed planes. The tapering of products varies with the growth temperature and PL spectrum of single ZnO pyramid shows strong UV emission, indicating good crystallization quality. 42 Experiment section The synthesis process was carried out in a quartz tube ~diameter 2 cm, length 20 cm. The source material was pure ZnO powder mixed with graphite (molar ratio 1:1), which was placed at the closed end of the quartz tube. The other end of the quartz tube was open to the atmosphere. Silicon substrates coated with 17 nm Au were placed one by one in the quartz tube for the growth of ZnO nanostructures. The quartz tube was inserted into a horizontal tube furnace heated to 1100 °C. The temperature gradient at the location between the source material and the open end of the quartz tube was approximately 600 °C. After 60 min evaporation, the quartz tube was drawn out from the furnace and cooled down to room temperature. White color products formed on the Si substrates in the temperature range of 800 to 500 °C. Results and discussions Figure 3.8 shows the SEM images of typical morphologies, micro particle film, micro & nano pyramids, micro brushes, and micro rods growing at different temperatures. The micro & nano pyramids are of interest since their tapered morphology makes them good candidates for the FED application and antireflection layers. These pyramid-like products are three-faceted, quite different from frequently reported six-faceted ZnO micro-pyramids that grow in the <0001> direction. Interestingly, the tapering of the as-synthesized ZnO pyramids decreases the further away the growth zone is from the source (See Figure 3.8c-f), thus making the tapering of products a tunable parameter, which is a key factor affecting the FED properties. 43 Figure 3.8 SEM images showing the three typical morphologies of the as prepared products: (a) micro rods; (b) micro brushes; (c)-(e) micro & nano pyramids; (f) micro particle film and (g) their corresponding growing site temperatures. 44 Figure 3.9 (a) The TEM image of the tip of one single nano pyramid and the EDS spectrum of the circled area. (b) and (c) The HRTEM image and its corresponding FFT pattern. Figure 3.9 shows the TEM image and EDS result of one single nano pyramid. A small nanoparticle was observed at the tip of the ZnO pyramid (Figure 3.9a) and was confirmed as Au by EDS, suggesting that the growth of micro & nano pyramids can be attributed to a catalyst-assisted VLS growth mechanism. Further HRTEM study in Figure 3.9b-c reveals ZnO pyramids grow in the [1 100] direction and are well crystallized without any stacking faults. The introduction of Au is critical for the formation of micro & nano pyramids. The nanobelts and nanobrushes are the main products in the same temperature zone for pyramid-like products when no Au is introduced. Herein, Au catalysts can tune ZnO morphologies to some extent though the mechanism is still unclear. Similar effects of Au on morphologies of ZnO nanostructures are found in Chapter 3.3.2 for tapered nanobelts. Figure 3.10 shows the PL spectra recorded on the micro & nano pyramid ensemble and one single ZnO pyramid. The PL spectrum of the ensemble has a sharp UV emission and a weak green emission while no detectable green emission appears in the single ZnO pyramid. Therefore, the visible emission from the ensemble comes from the ZnO particle film on the substrates, not from a single ZnO pyramid. Furthermore, the same PL spectra with only one UV 45 emission are obtained from different positions of a single pyramid, indicating the ZnO pyramids have a high crystallization quality and are free from defects. Figure 3.10 PL of the ZnO micro & nano pyramid ensemble and one single ZnO pyramid. 3.3 Method III: Direct evaporation of ZnO at high temperature and vacuum condition The morphology, structure, and orientation of the synthesized products rely on the following growth parameters: the supply speed of reactants, growth atmosphere, temperatures, pressure, substrates, and catalysts. Among these parameters, a stable speed and limited supply of reactants is critical for replicable and controllable results. The direct evaporation of ZnO can afford a controllable supply speed of reactants just by changing the heating temperature and pressure. Figure 3.11 shows the schematic diagram of our experimental setup and the temperature distribution. The highest temperature in the alumina tube at the evaporation area is always about 100 ºC lower than the setting temperature. The deposition area spans the interface where the alumina tube is in and out of the stove. Figure 3.12 shows the vapor pressure data of Zn and ZnO under different temperatures. The vapor pressure of Zn that comes from the 46 decomposition of ZnO is at least two orders of magnitude larger than that of ZnO at the same temperature. Large amounts of gray Zn powders were observed at about 400 ºC in the alumina tube, indicating that the reactant is mainly Zn vapor and O2 instead of ZnO vapor. Our experimental results show a vacuum lower than 0.1 Torr and a temperature higher than 1200 ºC are necessary for the growth of ZnO nanostructures with an appropriate decomposition rate of ZnO. In a typical ZnO nanostructure growth process, an alumina boat containing ZnO powder is placed in the center of a tube furnace. Substrates are placed downstream for the nucleation and growth of ZnO NWs. The furnace is heated to 1300 ºC and kept for half an hour under vacuum conditions (~10-2 Torr). White products were found to deposit on the substrates where the temperature reached about 700~900 ºC. In the following part, we study the controls on the growth of 1D ZnO nanomaterials by varying reaction parameters. Figure 3.11 (a) The schematic diagram of the experimental setup for synthesis of ZnO nanomaterials. (b) The distribution of temperature in the stove from the center. The temperatures listed in the top-right box are the ones measured by thermal couple. 47 Figure 3.12 Zn partial pressure over ZnO and vapor pressure of Zn solid, Zn liquid, and ZnO [84, 85] 3.3.1 Controllable growth of ZnO NW arrays on carbon-based materials 3.3.1.1 Vertical growth of ZnO NW arrays on PR Traditionally, patterned and aligned NWs can be fabricated by patterning metal catalytic particles, often gold, on lattice-matching substrates through VLS growth, although this growth strategy involves tedious lift-off processes for patterning metal catalysts and may lead to serious contamination in complementary metal oxide semiconductor processing. There have been some reports on the use of non-gold particles or seeding the growth of ZnO NWs by solvent methods, but the products have poor crystallization due to the low growth temperature with the risk of introducing impure ions. The substrates required for epitaxial growth of ZnO NWs are often expensive, including sapphire, GaN, and SiC, etc [65]. The mass production of 48 high-quality patterned and highly-aligned ZnO NWs at low cost is still a challenge for nanotechnologists. Here, we present a novel route to fabricating high-quality ZnO NW arrays with controlled morphology and NW density directly from carbonized photoresist (PR) nano/micro patterns followed by chemical-vapor deposition (CVD). Experimental section Preparation of PR patterns: Si substrates (1.5 cm×1.5 cm) were coated with a thin layer of PR (photoresist AZ1518, HPR204 or SPR6112) by spin coating at a speed of 4000 rpm for 30 s and then treated by hard-baking at 120 ºC for 60 s. The patterned PRs were made using an ABM Contact Aligner. The exposure time was set to 3.3 s and the developing time was 60 s. Growth of ZnO NW arrays: An alumina boat containing 3 g of ZnO powder was placed in the center of a tube furnace. Si substrates with PR patterns were placed downstream for the nucleation and growth of ZnO NWs. The furnace was heated to 1300 ºC and kept for half an hour under vacuum conditions (~10-2 Torr). Aligned ZnO NWs were found to grow on the substrates when the temperature was about 700~900 ºC. For the growth of the NWs shown in Figure 3.19b, the growth temperature was decreased to 1200 ºC to provide a relatively low concentration of Zn in order to inhibit the excessive nucleation and growth of the ZnO NWs. The as-grown NWs were characterized by a SEM (Philips, XL-30) and a HRTEM (JEOL, 2010F) equipped with EDX. The CBED measurement was carried out on a TEM (Philips, CM120) at 80 kV for optimal contrast. PL measurement of individual NWs was conducted using a NSOM (Nanonics, Cryoview2000) equipped with a He-Cd laser (325nm). The Raman spectra were measured at room temperature using a Jobin Yvon-T64000 micro-Raman spectrometer (Ar laser excitation at 514.5 nm). Characterization and optical properties On a PR-coated Si substrate, large-area and uniformly aligned ZnO NW arrays were fabricated by a normal CVD method (Figure 3.13a-b). To fabricate patterned ZnO NWs, we first prepared various sizes (from several hundreds of nanometers to several micrometers) and shapes (dot arrays, lines and networks) of PR patterns on various substrates by photolithography (Figure 3.13c), and, under a vacuum, these patterns were then annealed and converted to carbonaceous structures that acted as the sites for selective growth of ZnO NWs (Figure 3.13d). The sizes, shapes and numbers of ZnO NWs formed on one PR pattern can be 49 adjusted by changing the growth conditions. Controlled growth of an individual NW at a specific site was demonstrated. Figure 3.13 Fabrication process of ZnO nanostructure arrays directly from PR. (a) and (b) are Si substrates coated by PR patterns. (c) and (d) are the resulting NW arrays. As shown in the optical image in Figure 3.14a, ZnO NW arrays grown on a silicon substrate coated with carbonized PR (1.5 cm×1.5 cm) had strong absorption of visible light. The high density of ZnO NWs and their excellent vertical alignment are illustrated in the scanning electron microscope (SEM) image in Figure 3.14b. These images are very similar to those of high-quality carbon nanotubes [86] and NW [9] arrays from metal catalytic processes. The NWs were 20~200 nm in diameter and 5~12 μm in length under the growth conditions in this study (see Materials and Methods section) and the size increased with the growth time. In the experiment described here, the PR acted as a buffer layer between the substrate and the NW nuclei to enhance the growth of the aligned ZnO NWs. The high quality and excellent alignment of the NW arrays in a large area is clearly revealed by XRD spectra (see the upper one in Figure 3.14d). Since the c-axes of all ZnO hexagonal NWs (JCPDS 65-3411) were perpendicular to the substrate, only one strong (0002) diffraction peak was present. Figure 3.14c shows typical ZnO NWs grown with the assistance of Au catalysts on a Si substrate under the same growth conditions with diameter and length distributions similar to those with PR. In comparison with the ZnO NWs fabricated from PR, the poor alignment of the NWs is clearly reflected by the XRD spectra as shown in Figure 3.14d (the button one) in which diffraction peaks at (1-100), (1-101) and (1-102) appear in addition to the (0002) diffraction peak. With the assistance of the PR buffer layer, similar high-quality ZnO NW arrays have been obtained on 50 other substrates, such as quartz, Si3N4, polycrystalline Al2O3, SiC, etc. According to our results, the substrates suitable for the present method should be stable at 900 ºC and should not react with PR. The simplicity of this method makes it ideal for fast, low-cost and large-area fabrication of high-quality ZnO NW arrays. Figure 3.14 The optical (a) and SEM (b) images of ZnO NW arrays grown on a PR-coated silicon substrate. (c) ZnO NWs formed on an Au-coated silicon substrate. (d) XRD spectra recorded from the samples shown in (b) and (c). The upper one and the button one are corresponding to the sample in (b) and (c), respectively. Figure 3.15a is a TEM image showing the typical morphology of NWs grown along the [0001] direction. When viewed along the [1-100] direction, the thickness contrast profile (see the inset in Figure 3.15a) indicates the hexagonal shape of the NW. The corresponding CBED pattern (Figure 3.15b) confirms that the Zn-terminated (0001) planes are the fast growth front. The crystallinity of the NWs is shown in the HRTEM image in Figure 3.15c. Our TEM investigation revealed that the fabricated ZnO NWs contained few defects and were very pure 51 with no impurities detected within the limit of the EDX (Figure 3.15d). Although point defects, e.g., oxygen vacancies and impure atoms, have been long suggested to affect the PL of ZnO nanocrystals, such defects were not detected by electron microscopy in this study. No good way to determine the defect types and the density qualitatively in nanostructured materials has been established [9]. With a NSOM, we investigated the PL properties of individual ZnO NWs and found that ZnO NWs with small diameters (<30 nm) had strong UV emissions without any defect emissions (Figure 3.15eA), while NWs with large diameters (>300 nm) always had weak and broad peaks from defect emissions (Figure 3.15eB). A UV emission at about 380 nm corresponds to the near-band-edge free excitonic emission, while a green-band emission (the defect emission) from 500- 550 nm is commonly referred to as a deep-level or trap-state emission [78]. The origin of the deep-level emission is not yet clearly identified, but is generally attributed to point defects such as singly ionized oxygen vacancies and extrinsic impurities [9]. The UV emission peak positions measured from the as-grown ZnO NWs are diameter-dependent. The position of the PL peaks shifts from 375 nm (Figure 3.15eA) to 383 nm (Figure 3.15eB) as the diameter increases from 30 nm to 300 nm. The PL spectrum obtained from a large area of the NW arrays shows an intense UV emission at 381 nm with a narrow full width at half maximum (FWHM) of about 15 nm and a weak and broad green emission peak at about 520 nm. These results suggest that the as-grown ZnO NWs have high crystalline quality. We believe that the high-quality NWs are fabricated by the growth process described here since no metal catalysts are involved. In addition, we have found that the as-grown samples show excellent antireflection properties, which may have potential applications such as in dielectric antireflection coatings to enhance the efficiency of photovoltaic devices by increasing light coupling in the active region of the devices [51]. We have observed that Si substrates covered by ZnO NW arrays have lower reflectance spectra in the range of 350-1100 nm (Figure 3.15fI and II) and weighted reflectance (Rw) values [51] (11.8 % for I and 19.8 % for II) compared with random piled ZnO NWs (Figure 3.15fV, Rw 82.5 %) and polished Si substrate (Figure 3.15fVI, Rw 43.8 %). Our samples also show better antireflection response than ZnO NW arrays grown from Au catalysts (Figure 3.15fII). Strong alignment and uniform distribution of ZnO NWs can enhance the ARCs by effectively trapping light and leading to a broadband suppression in reflection [51,87]. Since the carbonized PR underneath the ZnO NWs also contributes to the antireflection properties because of its absorption of visible light, we measured the Rw value of the Si substrate coated with the carbonized PR after removing the ZnO NWs with 10 % HNO3 solution. We measured a low Rw of 26.4 %, suggesting that the good ARCs in our samples are the result of the special structure of the aligned ZnO NW arrays 52 formed on the carbonized PR. Figure 3.15 (a) The TEM image of an as-prepared ZnO NW and (b) the corresponding CBED patterns viewed along the [1 100] direction (the left one is the experimental result and the right one simulated by JEMS software). (c) An HRTEM image of a ZnO NW. The inset is the corresponding Fourier Transform Pattern. (d) The EDX spectrum recorded from the NW shown in (c). The copper peaks come from the sample supporting the grid. (e) PL spectra from A: a ZnO NW with a diameter of 30 nm; B: a ZnO NW with a diameter of 300 nm and C: ZnO NW arrays. (f) Reflectance spectra of ZnO NW arrays grown on (I) PR (Figure 3.14b), (II) Au-coated silicon substrate (Figure 3.14c), (III) Si substrate with the remaining carbonized PR after removing the ZnO NWs with a 10 % HNO3 solution, (IV) naked Si substrate and (V) random piled ZnO NWs. Multilayered ZnO NW arrays A natural idea is to repeat the above growth process for a multilayered structure (Figure 53 3.16). Figure 3.17a shows the morphologies of ZnO NW arrays coated with PR. The PR penetrates the ZnO NW arrays forming a PR/ZnO matrix layer. The surface is quite rough with caves everywhere. After the CVD growth, multilayered ZnO NW arrays can be obtained as shown in Figure 3.17b. However, the multilayered structure cracks due to the shrinkage of the PR at high temperatures. The new growth of ZnO NW arrays appears on the two sides of the curved PR leaving a three layered structure. The problems of unfairness and multilayer structure cracking will be dealt with in our future work. This multilayer structure is promising in the application of DSSC, LED and nanogenerators based on ZnO NW arrays. Figure 3.16 Schematic illustration showing the fabrication process of multilayer ZnO NW arrays. Figure 3.17 The SEM images of (a) PR/ZnO matrix and (b) multilayered structures with ZnO NW arrays as building blocks. 54 Patterned and vertical growth of ZnO NW arrays Compared to other carbon materials, carbonized PRs have remarkable advantages since they can be easily patterned by conventional photolithography. Figs. 3.18a-d illustrate several patterned ZnO NW arrays fabricated by our strategy, such as dot arrays (Figs. 3.18a and b), line arrays (Figure 3.18c), networks (Figure 3.18d). The sizes, lengths and shapes of the ZnO NWs and their densities in one PR unit can be modified by the growth conditions. As shown in Figure 3.18e, other morphologies such as ZnO nanopin arrays can be fabricated at a high temperature of 900 ºC. Most importantly, our approach has the potential capability to control the number of ZnO NWs on each PR unit. As demonstrated in Figure 3.19, we find that the number of ZnO NWs formed on one pattern decreases as the size of the PR pattern decreases. When the size of the PR pattern is about one micrometer, only a few ZnO NWs form (see the insets at the bottom of Figure 3.19a). In Figure 3.19b, we demonstrate our control of single NW growth on a small PR pattern. This growth is achieved by decreasing the evaporation temperature of the ZnO powder to provide a relatively low concentration of Zn to inhibit the excessive nucleation and growth of ZnO NWs. Notably, these single NWs are positioned at the corners of the square PR patterns (the dark contrast as marked by the arrows). The formation of one ZnO nucleus on each small PR pattern might be due to the high mobility of the initial Zn catalytic atoms deposited on the PR pattern. Since corners or edges of a patterned structure are often the preferred nucleation sites for material deposition, the catalytic atoms may diffuse preferentially to the corner to form a ZnO nucleus. The density of ZnO NW nuclei on the carbonized PR is mainly determined by the substrate temperature, vacuum condition, and source vapor concentration. Once these factors are fixed, the density of the ZnO NW nuclei (i.e., the number of ZnO NWs per unit area) is fixed. Here, when the size of the carbonized PR is small enough, only one ZnO NW is formed on each small PR square pattern (Figure 3.19b). 55 Figure 3.18 Various ZnO nanostructure arrays from PR patterns: (a) square dot arrays, (b) hexagonal dot arrays, (c) line arrays, and (d) hexagonal networks. (e) ZnO nanopin arrays. On the right side are the corresponding enlarged images. 56 Figure 3.19 (a) ZnO NWs grown on different sizes of PR patterns. Insets are enlarged pictures of the ZnO NWs formed on the patterns. (b) One single ZnO NW nucleated and grown at the corner of each small PR pattern. Mechanism of ZnO NW growth and c-axis alignment To understand the formation mechanism of ZnO NWs grown on patterned PR, we used Raman scattering to study the structural changes of the PR layers during the fabrication process. The pristine PR was composed of a photoactive compound called diazonaphthoquinone and novolak. The Raman spectrum of the pristine PR showed a uniform background (the inset curve in Figure 3.20a) indicating that the pristine PR might contain structures similar to hydrogenated amorphous carbon (a-C: H). After ZnO growth reaction, apparent D and G peaks appeared (Figure 3.20a). The position of the D and G peaks are about 1345 and 1601 cm-1, respectively, which means that carbonaceous materials similar to graphitic structure are formed [88]. The peaks of ZnO at 331 cm-1 corresponds to the second-order Raman spectrum arising from zone-boundary M point phonons 2-E2(M), 378 cm-1 corresponds to A1 symmetry with TO mode, 438 cm-1 corresponds to non-polar optical phonons high E2 mode, and 1162 cm-1 corresponds to 57 E1 symmetry with 2LO mode [1]. We believed that the carbon plays a critical role in the nucleation of ZnO NWs on carbonized PR. Zn vapor is hard to condense on naked Si surface and carbonized PR surface to form Zn liquid droplets because of their negligible mutual solubility. However, the zinc oxide vapor phase evaporated from the solid source could easily react with carbon [83] at a high temperature, and Zn could be extracted to form Zn droplets by the following reactions (Figure 3.20b): 2ZnO ( g ) + C ( s ) ↔ 2 Zn(l ) + CO2 ( g ) ……. (1) or 2ZnOx ( g ) + xC ( s ) ↔ 2 Zn(l ) + xCO2 ( g ) ……. (2) Since the melting temperature of Zn is much lower than that of ZnO, Zn droplets form preferentially on the carbonized PR layer and act as catalysts for ZnO NW growth as suggested in Refs. [89 ,90]. The reactions (1) and (2) require a high temperature and Zn vaporizes above 907 ºC mean that our growth temperature is in the range 700~900 ºC. Although the presence of impure nanoparticles or capped Zn particles at the tips of NWs has been regarded as one of the characteristics of VLS growth, our HRTEM investigation (Figure 3.21 b-d) reveals that there are no capped nanoparticles at the tips of the as-synthesized ZnO NWs. Figure 3.20 (a) Raman spectra of the photoresists before (the bottom curve) and after annealing (the top curve). (b) Nucleation and growth mechanisms of ZnO NWs on the photoresist patterns. 58 Figure 3.21 TEM images of tips of ZnO NWs. (b)-(d) are HRTEM images of tips Figure 3.22 (a) The cross-section TEM image of sample of ZnO on PR in initial growth stage. The PR layer is cleaved for stress. (b) The ending of ZnO NWs on PR. (c) The interface 59 between the root of ZnO NWs and the PR. (d) The corresponding FFT pattern shows the ZnO structure. To experimentally clarify the aligned growth of ZnO NWs, a cross-section sample of the ZnO NWs is prepared for TEM observation. Herein, it can clearly be seen that ZnO directly nucleates and grows on the surface of PR as shown in Figure 3.22. Especially, the definite interface between the root of ZnO NWs and the region of PR is shown in Figure 3.22b-c, and the corresponding FFT pattern of HRTEM image of the ZnO nuclei is showed in Figure 3.22d. The carbonized PR is amorphous structure according to our HRTEM investigation and ZnO is found nucleating with c-axis parallel to the surface of amorphous PR as shown in Figure 3.22c-d. Since there is no lattice matched relationship between amorphous PR and ZnO, the final vertical growth of ZnO NWs comes from the [0001] direction self-assembly of ZnO crystal nucleuses. Our experiments provide an important new example of c-axis alignment of ZnO NWs. The ZnO nuclei are about tens of atomic layer thick. Texturing seems to be an intrinsic thermodynamic feature of the growth of these nuclei [91]. Currently, two mechanisms that could cause the c-axis texturing of nucleating ZnO seeds despite the high energy of the {0001} surface are proposed: (1) The {0001} surface energy depends on the crystal thickness so that very thin ZnO crystals prefer a {0001} orientation, which is then kinetically locked-in as growth proceeds. Calculations on isolated ZnO slabs reveal that the {0001} cleavage energy does drop as a slab is made thinner, but probably not enough to decrease below the energies of the nonpolar faces. (2) The initial few atomic layers of ZnO form a low-energy configuration different from the bulk lattice and later convert to the (0001) orientation by a minor structural transformation. Claeyssens et al. has reported that extremely thin ZnO films may exist in a graphitic arrangement that undergoes a barrierless transition to the (0001) morphology above a threshold thickness of 10-20 Å [91]. Moreover, the nanometer-scale flatness of carbonized PR and the mutual insolubility of Zn, ZnO and carbon this will further favor the movement and rotation of ZnO nucleus resulting in c-axis texturing. Apparently, similar to solution methods using ZnO seeds, our high growth temperature will benefit the formation of c-axis textured ZnO nucleus and then the aligned ZnO NWs. This was confirmed by the observation that NWs at higher deposition temperature zone have better alignment than those in lower deposition temperature zone. 60 3.3.1.2 Horizontal growth of ZnO NWs on PR Lower down the supply speed of reactants by a lower evaporation temperature, the horizontal growth of ZnO NWs can be found as shown in Figure 3.23a-c. The vertical and horizontal growth of ZnO NWs can further afford ZnO NW networks (Figure 3.23d-f). The special interaction of Zn and ZnO with carbon can be used to explain these unexpected growth phenomena. Zn-carbon and ZnO-carbon systems are immiscible. Following the self-catalytic VLS mechanism of the ZnO NW formation on PR, liquid Zn will first be extracted and the carbon surface covered by carbon thermal reactions (Figure 3.24a). Then, thermodynamically, small Zn droplets form naturally on amorphous carbons by colliding and trapping processes. However, the interfacial action of Zn and carbon could trigger a repulsion response between Zn and carbon, due to the immiscibility of the zinc oxide-carbon system. Thus, the repulsion provides a thermodynamic driving force to impel Zn droplets to move randomly on the PR surface. Apparent line moving traces can be observed on the surface of the PR film (Figure 3.23c). The PR edge is a more favorable site for Zn droplets to accumulate from the energy aspect (Figure 3.24b). Meanwhile, the nucleation of ZnO clusters would occur during the motion of Zn droplets or when the Zn droplets stop. ZnO NWs grow perpendicularly to the surface of PR so that vertical growths appear on the top plane of the PR film, while horizontal or inclined growths appear on the sides. Similar non vertical growth phenomena of NRs are also observed near the edges of the PRs as shown in Figure 3.17-3.19. A thick PR film and a sharp edge are two critical factors for horizontal growth. More efforts are in progress for patterned and controllable horizontal growth, the aim being the direct assembly of NWs. Figure 3.23 The SEM images of horizontal growth of ZnO NWs. 61 Figure 3.24 The schematic illustration shows the horizontal growth process of ZnO NWs. 3.3.1.3 ZnO NW arrays grown on other carbon-based materials Hou T. G. et al. attributes the vertical growth of ZnO nanostructures to be the result of a good epitaxial lattice match of the c plane of ZnO with the hexagonal basal plane of HOPG [92]. However until now, details of oriented relationship between ZnO and HOPG have never been presented. According to our experimental results, crystallnity of carbon-based materials is not necessary for the aligned growth of ZnO NW arrays since crystallized carbon materials (graphite strips, HOPG), non-crystalline carbon materials (amorphous carbon film) and even grease can work well as shown in Figure 3.25. Flatness and stability at high temperatures for the carbon-based films are the only requirements for the vertical growth of ZnO NW arrays. Most importantly, the products can easily be scaled up to larger reaction chambers. A 2 inch wafer with vertical ZnO NW arrays is demonstrated in Figure 3.26. Furthermore, under the present growth conditions, these carbon-based materials not only provide perfect nucleation sites for the growth of aligned ZnO NWs, but also form excellent electrodes that connect to the NWs. These electrodes have excellent biocompatibility, chemical inertness, good thermal conductivity, thermal and mechanical stability and therefore are ideal for many nanomaterial applications [93]. Table 3.1 gives a summary of the properties of used carbon-based materials. 62 Figure 3.25 The SEM images of ZnO NW arrays synthesized with carbon-based materials: (a) grease left on Si substrate by fingerprint, (b) HOPG, (c) graphite strip and (d) amorphous carbon film on Si substrate by a carbon coater (Denton, Bench-Top Turbo). Figure 3.26 A 2 inch silicon wafer with ZnO NW arrays. 63 Table 3.1 A summary of properties of carbon-based materials for the growth of ZnO NWs HOPG Graphite strip Amorphous carbon Carbonized PR Remarks a Flatness Steps Cracks No Atomic flat Flat a, Available for NO No Yes but hard Easy to perform evaporation patterns Available for to perform NO Yes Yes -5 -5 -5 b, Yes large area Electrical thermal further annealing will be better -3 b 10 10 10 10 80–230 80–230 10-1 10-1 Good Good Good Good resistivity (Ωm) Thermal conductivity W·m−1·K−1 Chemical inertness and stability 3.3.1.5 Summary We have demonstrated a simple and effective method for the large area fabrication and patterning of high-quality ZnO NW arrays with controlled nucleation sites and densities on carbon-based materials. ZnO NWs nucleate preferentially on carbon-based materials, which are also excellent electrodes for connecting to ZnO nanostructures. We realized control of the preferential growth orientations (vertical and horizontal) and the growth of multilayered structures and network structures using ZnO NW arrays as building blocks. Further investigation will focus on the application for as-synthesized complex structures on an energy harvester. 64 3.3.2 Non c-axis growth of 1D ZnO nanostructures: substrate and temperature dependent morphologies. The crystallographic anisotropy of ZnO results in anisotropic growth. Under thermodynamic equilibrium condition, the facet with higher surface energy is usually small in area, while the lower energy facets are larger. Specifically in the ZnO growth, the highest growth rate is along the c axis and the large facets are usually {1 100} and {1120} . WZ ZnO has been shown to form structures such as NWs, nanobelts, nanorings, nanosprings, and nanohelices [55]. It is believed that the polar surfaces (Zn terminated or O terminated) play an important role in the formation of these nanostructures. The differently charged surfaces show distinct reactivity and thus display different growth behaviors. In this chapter, we show the controllable fabrication of three types of nanobelts/nanobrushes that have large polar surfaces while with different dominant planes {0001}, {1 100} and {1120} , respectively. These polar surface dominated 1D nanostructures could have potential applications as nanosensors and nanotransducers. Experiment section The growth process and parameters for the following nanostructures are similar to those for the growth of ZnO NWs in Chapter 3.3 while the substrates are changed to polycrystalline sapphire plates, which were placed one by one downstream. The main growth temperature ranges from 500ºC to 850ºC and three distinctive temperature zones were identified: zone I (500ºC～600ºC), zone II (600ºC～690ºC), and zone III (690ºC～850ºC) according to the morphologies of ZnO nanostructures as observed by SEM. Figure 3.27 schematically depicted the corresponding temperature and morphologies of nanostructured products. Figure 3.27 The schematic diagram of main growth temperature zones and the corresponding typical morphologies of nanostructured products. 65 Results and discussions Zone I (500ºC～600ºC): Figure 3.28a show the SEM image of 1D ZnO nanosized products collected at Zone I. The products show a necklace-like morphology and we name them nanochains. Figure 3.28b show the TEM images of nanochains. The image contrast of particles increases from the lateral to the middle, indicating the particles a rhombus shape. The projection angles of the rhombus shape particles at [0001] and [1120] directions are 60º and 94º, respectively. Compared with the sketch map in Figure 3.28d, the rhoumbus particles can be identified enclosed by polar planes {1 101} and {0001} . Though these particles are tens of nanometers is size, the wire part that connect them is quite thin, less than ten nanometers. Further HRTEM studies in Figure 3.28c confirm that the nanochains grow along [1 100] . The special morphologies of nanochains can be attributed to a secondary growth on the naked (0001) planes of pre-grown [1 100] / {0001} (growth direction/dominant plane) NWs/nanobelts. According to our knowledge these nanochains that are enclosed all by polar surfaces, are being observed for the first time. Figure 3.28 (a) SEM images of typical morphologies of ZnO nanochains. (b) The TEM images of nanochains. (c) The HRTEM image of one single nanochain with its FFTs inset. (d) The 66 sketch maps of ZnO nanochains. See text for details. Zone II (600ºC～690ºC): Figure 3.29 show the SEM images of as-fabricated products at zone II. Products collected near 600ºC show saw blade like morphologies (see Figure 3.29a) while products at a relatively high temperature show brush like morphologies (nanobrushes) (see Figure 3.29b). There are two types of nanobrushes, that is, [1 100] / {1120} and [1120] / {1 100} nanobrushes as shown in Figure 3.30b-c, respectively. Both types of nanobrushes are in the <0001> polar direction towards the side surface and the finger-like structures grow toward the [0001] direction as identified by CBED (Figure 3.30d-e). Combined with the SEM observations, these nanobrushes can be regarded as the results of a secondary growth on the active Zn-terminated polar planes of thin NWs/nanobelts. In addition, we found that these two types of nanobrushes showed distinct morphologies. Apparently wide belt arrays grow from nanobelts with dominant planes {1 100} (Figure 3.30b) while thin wire arrays grow from nanobelts with dominant planes {1120} (Figure 3.30c). The experimental results show that these two types of nanobrushes always coexist. Statistic result shows that the nanobrushes with dominant planes {1120} are larger than 90% though they are not energetically favorable compared with those with dominant planes {1 100} . Figure 3.29 (a) The SEM images of nanobrush products. 67 Figure 3.30 (a) and (b) are the TEM images of ZnO [1120] / {1 100} and [1 100] / {1120} nanobrushes, respectively. Insets are SAED patterns. (c) and (d) are the corresponding CBED patterns (experimental and simulated patterns). Scale bar 2 µm for (a) and (b). Zone III (690ºC～850ºC): Figure 3.31a shows the SEM image of typical belt-like morphologies of products (nanobelts) that were collected from zone III. These nanobelts have a uniform width of hundreds of nanometers and length of tens of micrometers. TEM studies in the Figure 3.31b reveal these nanobelts have smooth surfaces and most of them grow toward [1 100] with the dominant planes {1120} , which is identified by SAED (the inset in Figure 3.31b). These nanobelts also have {0001} polar planes as side surfaces. A low magnification TEM image of a single nanobelt is given in the inset of Figure 3.31c, which clearly displays the contrast between the [0001] and [000 1] sides. The surface at the [000 1] side is uniform while the surface at [0001] is uneven as shown in Figure 3.31c-d. The phenomenon is caused by the fact that Zn-terminated ZnO (0001) polar surface is chemically active and the oxygen terminated [000 1] polar surface is inert. We believe that the secondary growth on the [0001] sides of pre-grown [1 100] / {0001} NWs/nanobelts extends the dimension along polar directions resulting in new [1 100] / {1120} nanobelts. The observation of Figure 3.31e may provide evidence to support the inference. Similarly, all other non c-axis grown NWs/nanobelts may have their own development though they are rarely observed. 68 Figure 3.31 (a) and (b) The SEM and TEM images of uniform nanobelts. (c) and (d) are the HRTEM images recorded on the [000 1] and [0001] sides of one nanobelt shown in the inset in panel (c). The polar directions are identified by CBED. (e) A single nanobelt showing its development from thin NWs. The product morphologies are strongly temperature dependent. TEM investigations reveal that the typical ZnO morphologies, nanochains, nanobrushes, and nanobelts are structurally related as demonstrated in Figure 3.32. They can all be regarded as developing from pregrown thin NWs/nanobelts with polar {0001} side surfaces, in which active Zn-terminated surfaces have a secondary growth. Depending on the local temperatures, the deposition and dispersion conditions of liquid Zn on the (0001) sides are different since the Zn vapor pressure decreases as the temperature decreases. At high temperature zone III, few Zn concentrates and the secondary growth on (0001) sides is a slow VS growth process. At medium temperature zone II, more Zn concentrates and dispersed liquid Zn droplets can form partly covering the side surfaces and act as catalysts initialing a fast self-catalyst VLS growth. Simultaneously the uncovered side surfaces also have a slow VS growth. Hence, the nanofingers should be much longer than the width of the frame of final nanobrushes and the latter should be comparable to that of nanobelts found in zone III. These are consistent with our electron microscopy observation results (see Figure 3.30 and Figure 3.31). At low temperature zone I, overmuch liquid Zn cover the whole polar (0001) surfaces and the 69 secondary growth is inhibited [95-97]. Only small rhoumbus particles grow out of the polar (0001) planes. These liquid Zn easily oxidize and disappear in the growing and cooling process and this makes them barely observable by TEM. [94,95,96] Figure 3.32 The schematic illustration shows the relationship between typical ZnO morphologies, nanochains, nanobrushes, and nanobelts with thin NWs/nanobelts with polar (0001) side surfaces. Moreover, it is noted that the polycrystals Al2O3 substrate is critical for the formation of these non c-axis growth of 1D nanostructures. For example, under the same conditions, only NWs grown along c-axis are observed on carbon-based material coated substrates (See Figure 3.25b-d, and 3.25f) or Au-coated sapphire substrates. If considering a self-catalyst VLS growth, it is believed that the reactions between substrates, Zn droplet catalyst and ZnO nucleus determined the growth orientation of 1D ZnO nanostructures though the mechanism is not immediately clear. This will be studied as a quite important issue in the future. [97,98,99] 70 3.3.3 Defect related 1D ZnO nanostructures 3.3.3.1 Twin induced growth of Y-shaped ZnO nanobelts Bicrystal ZnO nanostructures have attracted much attention as bicrystal structures connecting building blocks can be an effective method for manually organizing a nanostructure [73, 98-100]. The growth of these nanostructures opens an option for assembling nanoscale blocks into a two-dimensional structure and the introduction of a twinned boundary into ZnO nanostructures can strongly affect the electronic, magnetic, optical and mechanical properties of ZnO [101-103]. Among these reported bicrystal structures, reflected twinned crystals of ZnO NWs/nanobelts are the most reported. According to our investigation, there are mainly two kinds of reflected twinned ZnO nanocrystals: a) the twinned planes are {1-10X} with X=0, 1, 2, 3, 4 [99-103]; b) the twinned planes are {11-2Y} with Y=2 [98]. Herein, we reported a new type of Y-shaped twinned ZnO nanobelt with twinned planes {11-21}. Due to the importance of surface polarities on the growth morphology, we performed convergent-beam electron diffraction CBED experiments and simulations, and the polarities of the nanostructures were determined. To study the optical properties of the twinned structures, we also carried out a room temperature large area PL. [100,101,102] These Y-shaped twinned ZnO nanobelts were fabricated by the thermal evaporation method. In brief, high purity ZnO powder (99.9%) was placed at the center of an evacuated (1~2 Torr) tube furnace. A polycrystalline sapphire substrate was placed downstream in a lower temperature region (400-800 ºC) of the furnace. After the furnace was heated to 1350 ºC and held for 0.5 h, the pressure was then increased to 100 Torr within 1 minute for another 0.5 h. The system temperature was raised to a designated set point at a rate of 10 °C min–1. An argon carrier gas was sent through the system at a rate of 25 sccm. After the reaction, the system cooled to room temperature and white products were found deposited in the substrate. Structural characterization of the as-synthesized materials were characterized and analyzed by a scanning electron microscopy (SEM, Philips XL30), a JOEL JEM-2010F HRTEM equipped with EDX. The CBED patterns were recorded by using a Philips transmission electron microscope (CM120), and the CBED simulation was performed by using the JEMS simulation software. Optical images were recorded by an optical microscope (Olympus BX60). A room temperature large area PL was excited by the 325 nm line of a He-Cd laser. Figure 3.33 shows the SEM image of as-synthesized products that have a Y-shaped morphology with widths of several hundred of nanometers and lengths up to 10 micrometers. 71 The high magnification SEM image in the inset of Figure 3.33 clearly reveals that each Y-shaped structure consists of two separate nanobelts separated by a distinct grain boundary with the direction parallel to the growth direction, which indicates the twinned crystal structures of the products. Unlike previously reported twinned ZnO nanostructures, the growth directions of nanobelts change at the end of the twinned structure forming two branches. Figure 3.33 A SEM image of the as-synthesized twinned ZnO nanobelts. The inset is an enlarged image of the twinned ZnO nanobelts. Scale bar 10 μm and 1μm for the inset. Figure 3.34 (a) Bright-field TEM image of a single twinned ZnO nanobelt. (b) and (c) Dark-field TEM images of twinned ZnO nanobelts recorded by a center objective aperture in 72 the positions of (0001) and (0001’) of the SAED pattern taken from the whole twinned ZnO nanobelts respectively. The insets in (a) are SAED patterns recorded from the place as marked by arrows. Scale bar 1μm for all. TEM demonstrates that the products are WZ structure twinned ZnO nanobelts. The twinned boundaries cannot be observed in the bright-field TEM image of the sample in Figure 3.34a because of quite a large sample thickness, while it is clearly revealed in Figure 3.35a when using a higher accelerated voltage of 200 kV. Selected area electron diffraction (SAED) patterns taken from the stem (inset in Figure 3.35a, as marked by an arrow) have two sets of spots with the [1-100] zone axis, which are the same as those taken from the two branches. This result reveals that the crystals on each branch follow the same orientation as the closest stem part. It was further confirmed by dark field TEM images that are recorded by central objective aperture in the (0001) and (0001’) positions of the SAED pattern taken from the whole twinned ZnO nanobelt respectively as shown in Figure 3.34b-c. The diffraction contrast from (0001) gives only the left part of the twinned nanobelts while the diffraction contrast from (0001’) gives the right part of the twinned nanobelts. The smooth feature from the stem to the branch in Figure 3.34b-c reveals no defects and coherent crystal orientations of each side of the twinned nanobelts. While the contrast is due to bending-induced strain, and the image indicates the equal projected thickness of the nanobelt. 73 Figure 3.35 (a) Bright-field TEM images of a single twinned ZnO nanobelt for HRTEM. (b)-(f) are HRTEM images recorded from the position 1-5. The insets pointed by arrows in (f) are CBED patterns recorded at the two sides respectively and the simulated ones that are marked with stars. The insets at the right bottom in (f) are the corresponding FFT of f). Scale bar 1μm for (a); 5μm for (b)- (f). Structures of the twinned nanobelts were further investigated by HRTEM and CBED. HRTEM images taken from different positions of a twinned ZnO nanobelt are shown in Figure 3.35b-f. Figure 3.35f shows two sets of clear lattice images of [1-100] zone axis with a clear twinned boundary. The angle between the (0001) planes in the two crystals is 33º. The growth direction of the twinned nanobelt is [11-26], which is perpendicular to the [1-100] zone axis and in the {11-21} twinned plane (see the inset at the right bottom of Figure 3.35f). (11-21), [11-26] is a new twinned system other than those results previously reported. It is interesting that the growth directions of the two crystals change from [11-26] (stem, Figure 3.33b and 3.33d recorded from 1 and 3) to [0001] (branch, Figure 3.35c and 3.35e recorded from 2 and 4). This special growth phenomenon can be correlated with the ZnO [0001] polar direction dominated growth. To confirm this, a CBED study is carried out. The CBED patterns are formed with a 74 converged electron probe focusing at the sample area in the nanometer range (for details, see reference). Two CBED patterns recorded from the branches of the twinned nanobelts respectively are shown in insets in Figure 3.35f (marked with arrows) along the [1-100] zone axis. The thickness of the twinned nanobelts is about 100~300 nm, appropriate for the current study of CBED. These experimental CBED patterns match fairly well with the simulated patterns (marked with stars, produced using JEMS simulation software). The best match was found for the sample thickness near 120 nm (with electron beam energy of 120 kV). We investigated a number of twinned ZnO nanobelts by using CBED and obtained the same result, namely, that the two sides of the twinned nanobelts are O-terminated toward the twinned boundary and Zn-terminated outward. Based on the growth condition and observed structure, we believe the formation of the twinned ZnO nanobelts as follows: first, the ZnO powders are vaporized and decomposed to Zn, ZnOx and O2 at high temperature and low pressure. They are then transported to the lower temperature zone, where they later oxidize to ZnO nuclei. Then twin-like ZnO nuclei might be induced by a sudden change in ambient pressure (increase from 1 Torr to 100 Torr) in the growth process as previously reported [103]. Previous work has shown that the preferred condensations of metal vapor on the grain boundaries resulting in atomic steps along twining surfaces and on the polar surfaces of ionic crystal can make these positions as the fast growth sites [104]. Hence we believe the final Y-shaped morphologies of twinned nanobelts are a competition result of the fast growth on these activated sites. The presence of these twin-like ZnO nuclei can result in the fast growth in the [11-26] direction which is parallel to the twin plane and finally in the stem formation of the twinned ZnO nanobelts. Afterwards the growth in the [0001] direction dominates because of the changes in the growth condition (the decrease in temperature and zinc vapor concentration) and finally we see the Y-shaped twinned nanobelts. The above noted growth is most likely to be controlled by a self-catalyzed VLS mechanism since no additional purities are added. Large amounts of gray Zn powders are found in the temperature zone of less than 400ºC indicating a rich zinc reactive atmosphere or an atmosphere reactive to lack of oxygen, which further supports the self-catalyzed VLS growth. Since the thickness of as-synthesized twinned ZnO nanobelts ranges from 100 nm to 300 nm, just half a wavelength of lights from ultra UV to yellow, the optical images (Figure 3.36a) recorded from the samples dispersed on silicon wafer show colorful properties caused by equal thickness interference. The different color reveals an uneven thickness distribution or bending of single twinned ZnO nanobelt. PL spectra of ZnO NWs at room temperature are shown in Figure 3.36b. A typical green defect emission at 521 nm is observed as largely reported in ZnO 75 nanostructures. Consider the oxygen-lack reaction atmosphere, this green emission can be assigned as recombination of oxygen vacancies V●O electrons with photoexcited holes in the valence band [9]. It is interesting that no UV emission can be observed. Previous work shows that only a rather weak UV emission appears when there exists large numbers of twins in ZnO nanostructures [99]. We believe the twinned structure of as-synthesized ZnO nanobelts can strongly inhibit the UV emission and the introduction of twin can be used as an effectively method to tune the optical properties of ZnO nanostructures. Figure 3.36 (a) High magnified optical images of single Y-shaped ZnO nanobelts. (b) The large area PL spectrum of Y-shaped ZnO nanobelts. 76 3.3.3.2 Screw induced growth of ZnO flowers The synthesis of these hierarchical structures was carried out by a simple vapor deposition process. Commercially available ZnO powder was placed in the center of a horizontal tube furnace. Polycrystalline alumina substrates were placed downstream at the lower temperature region of a horizontal tube furnace. The furnace is heated up to 1200 ºC and kept for 2 hours at a pressure of 2×10-2 Torr. After the growth, the furnace was cooled down to room temperature. As well as the white products collected from the substrates, gray powder was found at the 350 ºC temperature zone which was also sampled. The XRD data confirmed that the as-synthesized white sample was WZ ZnO. EDS study identified that only Zn and O with a ratio of about 1:1 existed without other impurities in the as-synthesized product. Optically, the product appeared white and covered the three alumina deposition substrates with a relatively high yield. The scanning electron microscopy (SEM) studies show three dominant morphologies from low temperature to high temperature: NWs, belts, dendrites, and brushes (we call them type I, type II, type III and type IV), which all consist of flowers grown at specific sites (Figure 3.37). Though the hierarchical morphologies shown here are complex and multifarious, they still comprise two parts: the first is the stem or the spine; the second is the branches of the flowers that grow at certain sites from the first part. 77 Figure 3.37 SEM images of the ZnO hierarchical structures as-synthesized. (a), (c) and (e) are typical morphologies from low temperature region to high temperature region. (b), (d) and (f) are corresponding enlarged ones of (a), (c) and (e). Number 1, 2 and 3 mark single-layered, multilayered and multifid flowers, respectively. Scale bar, 100 μm for (a), (c) and (e); 20 μm for (b) and (d); 10μm for (f). The structure of these flowers was analyzed with a transmission electron microscope (TEM). Figure 3.38a shows a low-magnification TEM image of a flower with its open direction parallel to the electron beam. Image contrast varies following the thickness of the petals. Corresponding electron diffraction patterns recorded from the petals without any rotation identifies that the flower is a single crystal and grows along <0001>. The arrows of the projection of the flower are along [1120] and the flat sides are along [1 100] . No additional points contributing from defects are observed. Furthermore, CBED analysis by focusing the beam on the edge of the flower has been used to study the polar surface orientation of ZnO flowers. Experimental and simulated CBED patterns indicate that the Zn-terminated polar surface points to the direction that the flower opens (see Figure 3.38b, for a detailed method to carry out CBED please refer to Ref. [74]). Figure 3.38 (a) TEM image of a nanoflower with the beam direction along the open direction of it. The inset is the SAED pattern of the nanoflower, which shows the nanoflower open toward [0001] and be single crystal. b) A TEM image of a nanoflower with the beam direction perpendicular to the stem. c) The CBED pattern of e) taken along [1 100] showing that the nanoflower grows from Zn-terminated polar (0001) site. Scale bar, 2 μm for (a); 5 μm for (b). 78 These flowers have various morphologies as demonstrated in Figure 3.37: single-layered, multilayered and multifid, and so on (marked with numbers 1, 2 and 3 in Figure 3.37b-d and 3.37f). Even so, these flowers can generally be classified into two types according to their opening angles as shown in Figure 3.39c-e. The substrate temperature affects the flower’s opening angle because at lower temperatures they have wider opening angles. The SEM observation of the developing branches of flowers indicates that both undergo helical growth. The final morphologies of flowers with wide opening angles always show distinct left or right handed characteristic (Figure 3.39i-k). After analysis of nearly one hundred separated flowers with chirality, it is found that about 90 % are right-handed and only less than 10 % are left-handed. Furthermore, small numbers of branches of flowers with two screws in same directions are also observed (see Figure 3.39k). Since the flowers presented here are single crystals, crystallogeometry analysis can be used based on the exterior morphology of a single perfect crystal that obeys the constancy law of interfacial angles. Three dimensional models were constructed to identify the “exact” structure. Figure 3.39a-b give the profile maps for nanoflower models that consist of different {1 10 x} (x=1, 2, 3, 4, 5, 6) viewed along [1 100] and [1120] . By deliberately comparing an observed flower crystal, we found that models consisting of {1 103} planes match well with wide opening flowers. (Figure 3.39c and 3.39f; Figure 3.39d and 3.39g) and models consisting of {1 104} planes match well with small opening angled flowers (Figure 3.39e and 3.39h). 79 Figure 3.39 (a) and (b) are the projection maps for nanoflowers that consist of different (1 10 x) (x=1, 2, 3, 4, 5, 6) viewed along [1 100] and [1120] . (c) and (d) are two typical nanoflowers with large opening angles. (e) is a typical nanoflower with small opening angles. (f)-(h) are the corresponding three dimensional profile maps of (c)-(e). (f) and (g) consist of {1 104} planes and (h) consists of {1 103} planes. Nanoflowers with left handedness (i), right handedness (j) and two screws that possess same-handedness (k). Scale bar, 10 μm for (c) and (d); 5 μm for (e) and (i)-(k). Figure 3.37a shows SEM images of the typical morphologies of type I and type II mixed at low temperature. Type I consists of thick and long NWs with flowers having large opening angles on their tips or drilled through (Figure 3.40a-b). The flowers are hexagonal indicating that the NWs grow along [0001]. There are two kinds of belt-like hierarchical structures of type II coexisting: Type II-A has spines of V-shape belts and type II-B has spines of rectangular belts (Figure 3.37b). Flowers of type II-A and B assemble in a line and open on one side of the belts. The direction that the flowers open is [0001] and the reverse direction is [000 1] . Figure 3.40e 80 shows the other side of the belt is smooth besides some nanoparticles grown along the axis of the belt. This phenomenon is produced by different growth rates on the (0001)-Zn and (000 1) -O terminated surfaces [7]. ZnO flowers have A two oriented relationship with the spines in type II-A, that is, the hexagonal side of the flowers is parallel or perpendicular to the axis of the spines (Figure 3.40c-d), indicating the existence of ZnO spines in two growth directions. Some developing structures are also observed. Figure 3.40f-g show that thin triangular films extend from a one-dimensional spine and integrate to form thin wings alongside the spine. It is noted that a small isolated islands appears on the axis and the flowers originate from these islands by a helical growth (see Figure 3.40g-h). The petals of the flower extend in the same direction as the wings. The TEM image of a belt-like hierarchical structure with the hexagonal side parallel to the axis of the spines is shown in Figure 3.40i and the corresponding SAED in the inset identify that the growth direction of the spine is along [1120] . An apparent contrast can be observed because the middle part is thinner than the sides of the V-shape belt. For the hierarchical structures shown in Figure 3.40c with the hexagonal side perpendicular to the axis of spines, the growth direction of the spine is along [1 100] . Figure 3.40j-l shows typical SEM images of type II-B flowers growing on a thick rectangular belt. The spines of the belt with the three main growth orientations are found as the main products and the directions that the flowers open toward here are perpendicular to the spines or their incline to them. Considering that the flat sides of the flowers are along [1120] , by comparing the angles of the directions that the flowers open toward and the growth direction of spines with Figure 3.39a -b, we can determine the growth directions of the spines as [1120] , [1 101] and [1 102] . The flowers and thin films that extend from the spines in type II have large opening angles consisting of [1 103] planes. 81 Figure 3.40 SEM images of NWs with flowers being trilled through (a) and at their tip endings (b); Belt-like hierarchical structures along [1 100] (c) and [1120] with flowers open toward [0001] (d); Some particles are observed at the ridge of the [000 1] side of the belt-like hierarchical structure(e); Developing morphologies of belt-like hierarchical structures (f)-(h); TEM image and SAED of (g) showing the belt is a single crystal with the spine along [1120] and its projection plane perpendicular to [0001]; (j)- (l) are SEM images of flowers growing on thick rectangular belt along [1120] , [1 101] and [1 102] direction. Scale bar, 10 μm for (a)-(d), (f) and (j)-(l); 5 μm for (g); 2 μm for (e) and (h); 200 nm for (i). Figure 3.37c shows typical hierarchical structure of type III, branches of a flower growing on dendrites. Most branches of flowers usually consist of a ball and a flower with multilayer petals (Figure 3.41a). A six-fold symmetry ball is always observed forming at the tip of the stem and the flowers grow on the balls (Figure 3.41b-c). The flowers here have small opening angles consisting of {1 104} planes. 82 Figure 3.41 SEM images of developing morphologies of dendrites with flowers growing at the tips (a). Balls with six symmetry form and then flowers grow on the ball (b)-(d). Scale bar, 10 μm for (a); 2 μm for (b)-(d). Figure 3.37e shows a typical hierarchical structure of type IV, branches of flowers grown on brushes, always consisting of multilayer petals. Figure 3.42a-c show brushes with three main growth orientations of [1120] , [1123] and [2243] . Figure 3.42d shows some developing branches of flower. Helical growth circling the [0001] stem is clearly observed. The flowers here have small opening angles consisting of {1 104} planes. Sometimes, balls made up of flowers can be observed in the same temperature region. Figure 3.42 (a)-(c) SEM images of flowers growing on a thick rectangular belt along [1120] , 83 [1123] and [2243] . (d) Flower growth initiates from the screw dislocations. Scale bar, 10 μm for (a)-(c) and 5 μm for (d). As has been pointed out previously, normally the formation of the hierarchical structures can be separated into two major stages. However, here there are three likely stages in the formation of these special hierarchical structures. The first stage is the fast growth of the spines of different morphologies (NWs, nanobelts and more complex structures like dendrites), which vary according to different deposition temperature regions. These different spines grew with naked Zn-terminated (0001) planes on their heads or sides, which were identified by CBED analysis. We know that the Zn-terminated surface is catalytically active and induces secondary growth, whereas the O-terminated surface is inert. Furthermore, the closer to the source, the higher the temperature and the thicker the Zn vapor and O2 vapor coming from the decomposition of ZnO at 1300 ºC. As a result, the second stage nanostructures near to the source developed into nanobrushes or thick belts through slow growth on naked (0001) surfaces, which were described in Chapter 3.3.2. The growth at the third stage is rather different to that of the second stage: flowers grew from the screw dislocations at naked (0001) sites of previously formed spines. It is worthy to note that a large amount of grey powder was found deposited at low temperature regions of about 300-400 ºC. However we found little grey powder when decreasing the reaction to one hour and only products (NWs, nanobelts and nanobrushes) at the second stage were found. The grey powder was identified as irregular thin Zn NWs by TEM with EDS. Based on this experimental analysis, we believe that the flower growth at the third stage is related to the enrichment of Zn vapor at a low temperature. With enough reaction time, more ZnO was decomposed into Zn vapor and O2. Some O2 reacted with Zn vapor to form ZnO crystals again at the deposition region and other O2 was pumped out continuously while most of the redundant Zn vapor condensed and deposited at a low temperature region. Therefore, there was a reversal in the distribution of Zn vapor concentration compared to that at the beginning of the reaction. Accompanying this change, the third grow stage carried out. Previous study [100] have shown that the growth velocity along the [0001] direction can be suppressed in the presence of high Zn vapor concentration, resulting in fast growth along < 1120 > .In our case the growth in the [0001] direction was suppressed and new fast growth directions < 2249 > and < 1126 > were found accompanied by screw dislocations, which resulted in the growth of flowers. Since Zn vapor concentration at low temperature regions is larger than that at high temperature regions, the growth speed along [0001] is far more hindered. As a consequence, flowers at low temperature regions have larger opening angles while at high 84 temperature regions small opening angles as identified in Figure 3.39h. In addition, special flower clusters were occasionally observed at low temperature regions coexisting with hierarchical structures of type I and II. We believe some Zn gas concentrate to form big liquid droplets and polycrystal coat formed outside of the droplet by oxidation and each small crystal developed into one flower resulting in flower clusters. In summary, hierarchical ZnO nanostructures with high-yield are reported. For the first time, novel ZnO nanoflowers were observed blooming at special sites of several kinds of spines, which were identified as polar (0001) planes or tips. The growth phenomena showed here demonstrated the key significance of Zn-terminated polar planes in the fabrication of hierarchical structures. The morphologies of ZnO hierarchical structures shown here are more complex than those previously reported. The spines for the as-synthesized hierarchical structures can be NWs, nanodendrites, nanobrushes, and nanofilms. The branches here are novel flowers enclosed by petals of high index planes and their growth was explained as a screw dislocation leading growth. The reversion of Zn vapor concentration distribution as confirmed by SEM observation and EDS analysis, is proposed as the cause for the growth of nanoflowers. 3.4 Summary Three different methods based on the vapor transfer process are applied to fabricate 1D ZnO nanostructures. We succeeded in realizing control over morphologies (NWs, nanocones, nanopins, nanochains, nanobelts, and nanobrushes), alignment (vertical and horizontal growth), assembly (mono & multilayered NW arrays, complexes built by tetrapods), patterning and defects (tetrapods, Y-shaped twinned nanobelts and hierarchical nanostructures decorated by flowers induced by screw dislocations) of ZnO nanostructures. PL studies show most high quality as-fabricated ZnO nanostructures exclude defective nanostructures. The spatially resolved PL study of single ZnO tetrapods reveals that defective emission originates from structural changes in the tetrapods, and is not surface state related. Vertical ZnO NW arrays fabricated directly from photoresist show a low Rw and can be used as antireflection coatings. 85 Chapter 4 Ambient Stability of ZnO NWs: Structural Degradation and Related PL 4.1 Introduction Recently, the degradation of optical properties in ZnO has gradually attracted attentions due to its importance from optical application viewpoint [9, 35]. Shan et al. studied the structural changes and PL characteristics of ZnO films exposed to air for several months [105]. They observed that the deep-level (DL) emission decreased with time which was attributed to the reduction in the number of oxygen vacancies within the films. In ZnO nanostructures, the DL emission has been frequently observed, and it has been widely acknowledged that the origin of the DL emission is mainly due to point defects, e.g., oxygen vacancies and impurity atoms formed in ZnO crystals. Efforts have been devoted to improving the optical properties by various treatments. In this chapter, we report the structural changes during the degradation of ZnO NWs in various environments and the PL study of the degraded ZnO NWs by NSOM. We continued to observe UV emission without additional DL emission generated when the structures of ZnO NWs gradually degraded. We believe that the formation of the surface defect states on ZnO NWs is not responsible to the green or DL emission. For those ZnO NWs showing DL emission, the commonly used treatment methods e.g. post-annealing and plasma surface treatments can not effectively eliminate the DL emission. 4.2 Experimental section ZnO NWs were fabricated on (100) Si substrates by a simple method of vapor-phase transport process. ZnO powder as source material was placed at the center of a tube furnace. The Si substrates coated with 5 nm Au were placed at the downstream for the nucleation and growth of ZnO NWs. The furnace was subsequently heated up to 1300 oC and held for 0.5 hour in vacuum condition (2×10-2 Torr). Finally, the ZnO NWs were found to form at the region where the substrates temperature was about 900 oC. The structure of the as-grown ZnO NWs was first characterized by XRD (PaNalytical X’Pert Pro MPD). To study the morphology and PL property of an individual ZnO NW, the as-grown NWs were scratched off the substrate and transferred onto a clean Si substrate. Afterward, the NW samples were placed into different environments such as in air, H2O, CO2, mixture of H2O and CO2, H2, and O2 respectively for 86 various periods at different temperatures. The morphology and structural changes of ZnO NWs were examined by a SEM (Philips, XL30) and a HRTEM (JEOL 2010F) equipped with EDS. The PL spectra of individual ZnO NWs were recorded by a NSOM system (Cryo2000, Nanonics Inc., Israel) at room temperature. The excitation light source was He–Cd laser (325 nm) and the excitation area on the sample was about 1 μm in diameter via a tapered optical fiber tip. 4.3 Results and discussion 4.3.1 Structure and PL studies of ZnO nanostructures in air Figure 4.1a shows a typical XRD result of the NW product formed on a (100) Si substrate. All the diffraction peaks can be well indexed as the hexagonal WZ ZnO (ICDD-PDF No. 50-0792). No other structures or crystalline impurities have been detected, indicating the good crystallinity of the product. The NW morphology is shown in Figure 4.1b in which vertically well-aligned and tapered ZnO NWs with smooth surfaces are clearly visible, and the NW diameters shrink gradually from bottom to top. High-resolution transmission electron microscopy (HRTEM) images of an individual NW illustrate the high quality and uniformity of the WZ crystal structure with no observable defects although there is always a very thin amorphous layer on the NW surface. The corresponding Fourier transform pattern shown in the inset of Figure 4.1c indicates that the NW grew along the <0001> direction. The PL spectrum taken from tapered ZnO NW ensembles exhibits only a strong UV emission peak centered at 385 nm. The DL emissions which are related to point defects including oxygen vacancy (VO), zinc vacancy (VZn), zinc interstitial (Zni), and oxygen interstitial (Oi) are not detected. 87 Figure 4.1 (a) XRD and (b) SEM image of ZnO NW product. (c) HRTEM image of a single ZnO NW. The inset is the corresponding FFT pattern. (d) Normalized PL spectrum taken from the large-area ZnO NWs. As shown in Figure 4.2a, the PL property of an individual tapered NW dispersed on the Si substrate has been investigated by collecting the PL spectra at different sites along the NWs by NOSM. The length of the NW is about 20 μm and the diameter varies from 250 nm to 80 nm. Figure 4.2b illustrates the PL spectra collected from three positions as marked by the black arrows and labeled with A and B. Obviously, only strong UV emission appears and no DL is detected. There are slight blue shifts of the UV emission peaks recorded at the positions where the diameters are relatively thin. Such size-dependent blue shifts of UV emission from ZnO NWs have been reported previously [16, 17]. This has been attributed to the reduction of the band gap caused by the nano-size confinement and domination of surface excitons. 88 Figure 4.2 (a) the SEM image of a fresh tapered NW. (b) PL spectra collected at the thinner part, A, (dash line) and at the thicker part, B, (solid line) of a tapered fresh NW. (c) HRTEM of the single NW aged in air for seven weeks. Recently, Pan et al. [106] have studied the light emission in a single tapered ZnO NW by cathodoluminescence (CL). They observed that the integral intensity of UV (IUV), DL emission (IDL) and the ratio of IDL/IUV decreased with decreasing the NW diameter. They suggested that instead of the body recombination effect, the surface enhancement effect dominated the DL emission due to the larger surface-aspect ratio caused by shrinking the NW diameter. However, Chung et al [107] obtained different results that the IDL tended to decrease while IUV increased as measured from the root to the tip of a NW. They believed that the inhomogeneous distribution of point defects and surface defects from bottom to top of the NW was the main reason for the change of the emission intensity. According to these studies, it seems that different defects can induce different DL emission phenomena, and the defect properties varied in the samples prepared by different research groups. 89 4.3.2 Structure and PL studies of ZnO nanostructures in aggressive atmosphere In the present work, we choose two kinds of NWs and compare their PL properties and degradation: (1) high quality ZnO NWs without the DL emission (the density of surface states and native defects in the NWs should be very low) and (2) ZnO NWs with DL emission. Both NWs were prepared by the same method, but collected at different substrate locations. We have systematically investigated the structural degradation of single NWs by exposing them to air, carbon dioxide and water, etc. for a long period (one to seven weeks). One of the purposes of this study is to clarify whether defects can be introduced into the high-quality NWs under different environments or by treatment methods and thus PL degradation or DL emission can be induced. Figure 4.2c is a HRTEM image showing the formation of a thick amorphous layer on the NW surface after exposing the NWs in air for several weeks. This layer was identified to be amorphous carbon (a-C) by EDS. Similar carbon layers were also observed on some other semiconductor NWs such as ZnSe and ZnTe. Although the a-C layer might bring in surface states or surface defects for the ZnO NWs, the PL spectrum recorded from this NW has no change, i.e. no DL emission is generated. We then further investigated the structural degradation and related PL of individual NWs exposed to H2O, CO2 and mixed gases of CO2 and H2O, respectively. In CO2 environment, the a-C layer (about 10 nm thick) formed on the NWs was thicker than that formed in air for the same time. We believe that the NW treated in CO2 environment can absorb more carbon on its surface due to the higher CO2 concentration. Similar to the PL results obtained from the NW exposed in air, the thick carbon layer did not influence the PL property. In H2O environment, we have observed that both the crystalline structure and PL property of ZnO NW kept unchanged for seven weeks. Table 4.1 lists the details about the structure degradation and related PL properties. 90 Table 4.1 Aging conditions and results of structure and PL property Sample Aging Structure degradation PL property after aging environments and time High-quality Air (7 weeks) 4 nm carbon surface layer UV emission Intact interior structure ZnO NW High-quality H2O vapor (7 Intact surface and interior ZnO NW weeks) structure High-quality CO2 (7 weeks) 10 nm C surface layer ZnO NW UV emission UV emission Intact interior structure High-quality H2O + CO2 (3 C and poly-crystalline ZnO NW weeks) surface layer UV emission single-crystalline NW ZnO NW with post-annealing (O2, Intact surface and interior DL emission DL emission 400 ºC, 30min) structure (a little lower) ZnO NW with post-annealing (H2, Intact surface and interior DL emission DL emission 400 ºC, 30min) structure (a little stronger) ZnO NWs aged in the complex environment of CO2 and H2O, however, showed dramatic structure degradation. The SEM images in Figure 4.3a illustrate the morphology changes of a single crystalline tapered NW aged in CO2 + H2O at different periods. The NW was gradually eroded at some sites. After three weeks, the whole NW was eroded away. The structure degradation process of the NWs has been systematically characterized by HRTEM. As shown in Figure 4.3c, after aged for one week, the NW surface was changed first. A thin layer of about several nanometers was converted to polycrystalline nanoparticles wrapped in a thick amorphous layer. The core of the NW was still single crystalline, which is confirmed by the corresponding HRTEM images shown in Figure 4.3d-e, respectively. After three weeks, the eroded area on the surface of the NW was further extended (see Figure 4.3f) resulting in a thick shell (mixture of nanoparticles and carbon) wrapping the NW. The HRTEM image of area C (marked in Figure 4.3f) clearly shows the interface between the single crystalline core and the newly formed eroded shell. After six weeks, the NW was finally eroded completely. 91 Figure 4.3 (a) SEM image of a tapered NW aged in the complex environment of CO2 + H2O for different time: “1”, “2” and “3” denote sample aged for zero, one and three weeks, respectively. (b) The PL spectra of an individual NW before (“1”) and after aging in the CO2 + H2O environment for one week (“2”) and for three weeks (“3”), respectively. (c) TEM image of the NW aged for one week. (d) and (e) HRTEM image of area A and B marked in Figure 4.3c, respectively. (f) TEM image of the NW aged for three weeks. (g) HRTEM image of the area C marked in Figure 4.3f. It is interesting to note that although the NW structure gradually eroded away, the UV emission from the NW did not show obvious degradation. Moreover, no DL emission or other luminance was generated. Since the single crystalline ZnO NWs have been totally converted to polycrystalline structure wrapped by a-C (according to EDS results), and a high density of 92 defects should be introduced to the surface and interior structures of the NW. Since the PL spectra taken from the degraded NWs are very similar to that of a fresh sample, it is believed that the crystalline defects introduced extrinsically can not easily affect the PL property of the high-quality ZnO NWs. In other world, the PL property of high-quality ZnO NWs is very stable. Figure 4.4 (a) PL spectra recorded from an individual ZnO NW before and after annealing in H2 at 400ºC for 30min. (b) Another results from a ZnO NW before and after annealing in O2 at 400ºC for 30min. For those as-grown ZnO NWs already containing the UV and DL emission, we observed similar structure degradation through the treatments of different environments. Again, the UV and DL emission from these NWs showed almost no change during degradation. For ZnO thin films, it has been reported that annealing in H2 and N2 for 30min, at about 400oC resulted in the increase of the DL emission and decrease of the UV emission [108]. However, for ZnO NWs, these treatments have not caused obvious change of both UV and DL emission. The results from other normal treatment methods can be found in Table 4.1. Figure 4.4a-b show the PL spectra of the ZnO NWs annealed in H2 and N2 at 400oC for 30min. Clearly, different from ZnO thin films, the native defects generated during growth of ZnO NWs can not be easily eliminated by normal treatment methods. 4.4 Summary High-quality single crystalline ZnO NWs fabricated by thermal evaporation of ZnO 93 powder assisted by carbon reduction exhibited strong UV emission without any DL emission. The UV emission showed blue shifts as the NW diameters decreased. Under different environments of surface treatments, the crystal structure of the NWs may degrade, but the corresponding PL properties of the NWs remain unchanged. For those ZnO NWs containing DL emission, the commonly used treatment methods e.g., post-annealing and plasma surface treatments can not effectively eliminate the DL emission. The native defects generated in ZnO NWs showed distinct properties compared to that in ZnO thin films. 94 Chapter 5 Chemical Stability and Biocompatibility of 1D ZnO Nanostructures 5.1 Introduction Biosensors based on nanostructured materials such as NWs [109] and nanotubes [110] have received increasing attention in recent years. Nanostructures possess high bio-sensitivity because of the depletion or accumulation of charge carriers on their surfaces when charged biological macromolecules are bound. This surface charge affects the entire cross-sectional conduction pathway [4]. ZnO nanostructures are one type with many potential applications in optoelectronic devices, transducers, and photovoltaic devices [1, 55, and 59]. Recently, ZnO nanomaterials are being widely investigated for use in biosensors [111], bio-imaging [112], drug delivery [113] and other such biological applications. Such applications often involve direct interaction with biological systems and sometimes require a certain time to perform their functions, for example in the real time bio-imaging of protein interactions or in real time monitoring of changes in pH or glucose in vivo. This makes investigating the bioactivity, biocompatibility, chemical stability and behavior of nanomaterials in biofluids a priority. So far, little has been published on the bioactivity and biocompatibility of ZnO nanomaterials [115-118] and their chemical stability and interactions with biofluids remain poorly understood. [114,115,116,117] Wurtzite ZnO has been shown to form structures such as NWs, nanobelts [12], nanorings, nanosprings and nanohelices [4, 55]. It is believed that the polar surfaces (Zn terminated or O terminated) play an important role in the formation of these nanostructures. The differently charged surfaces show distinct reactivity and thus display different growth behaviors. To correlate the growth behaviors with the structure of nano-sized crystalline ZnO, it is important to identify the polarity of the ZnO nanocrystals experimentally. This has been the first study of the in vitro chemical stability of ZnO nanostructures with different morphologies and their etching behavior in simulated physiological solution (SPS) with ion concentrations approximately equal to those in human blood plasma. SPS has been used widely for in vitro assessment of the artificial biomaterials [118]. ZnO nanomaterials synthesized by thermal evaporation and solution methods have different “survival times” in SPS, and they show interesting anisotropic etching behaviors. The anisotropic etching behaviors can be very useful 95 for determining the polar directions of ZnO nanostructures. 5.2 Experimental section 5.2.1 Preparation and characterization of 1D ZnO nanostructures ZnO NWs and nanoribbons with different morphologies were fabricated by (1) a simple thermal evaporation method based on the self-catalyzed VS mechanism (no metal catalyst was used); and (2) a simple hydrothermal method without using any capping agents or surfactants. In brief, ZnO powder was placed at the center of an evacuated (2×10-2 Torr) tube furnace. Substrates were placed downstream in a lower temperature region (400-800 ºC) of the furnace. The furnace was heated to 1400 ºC for 2 hours. For the growth of ZnO NWs and nanobelts, carbon foil and a polycrystalline sapphire substrate were used respectively. In the solution hydrothermal method, 5 mL of 0.1 M zinc acetate ethanol solution was mixed with 35mL of 0.5 M NaOH ethanol solution to form a mixed solution which was later transferred into a Teflon-lined stainless steel autoclave (50ml) and heated to 180 °C. After 24 hours, a white precipitate had formed at the bottom of the autoclave with transparent solution above the product. The as-prepared samples were characterized by a XRD (Philips, PW1813). A drop of solution containing the nanostructures was diluted with ethanol and sonicated for 15 min. The nanostructures were then dispersed onto a holey carbon film for structural characterization using a HRTEM (JEOL, JEM-2010F) equipped with an EDS. The CBED patterns were recorded using a TEM (Philips, CM120), and the CBED simulation was performed using the JEMS software. The PL measurements were carried out using a He-Cd laser as the exciting light source. 5.2.2 In vitro experiments in SPS solution The in vitro experiments were conducted in the SPS [115] which has the same ionic composition as human plasma. One liter of SPS was prepared by dissolving 5.403g NaCl, 0.736g NaHCO3, 2.036g Na2CO3, 0.225g KCl, 0.182g K2HPO4, 0.310g MgCl2·6H2O, 11.928g HEPES (2-(4-(2-hydroxyethyl)-1-piperazinyl) ethane surfonic acid), 0.293g CaCl2, 0.072g Na2SO4 and 1.5ml 1mol·L-1 NaOH into double distilled water in sequence. The ion concentrations of this SPS are listed in Table 5.1. ZnO nanomaterials were immersed in SPS at 96 37 ºC for several hours to several weeks. After this treatment, the ZnO nanomaterial samples were washed with distilled water and dispersed on holey carbon supporting grids for structural characterization by transmission electron microscopy. Table 5.1 Ion concentrations of SPS Ion Concentrations (mmol\dm3) Na+ K+ Mg2+ Ca2+ Cl- HCO3- HPO42- SO42- 142.0 5.0 1.5 2.5 103.0 27.0 1.0 0.5 5.3 Results and discussion As shown in the XRD data in Figure 5.1a, ZnO NWs and nanobelts synthesized by our methods can be well indexed as WZ ZnO. The diffraction peaks are sharp and no other phases can be found. This indicates that these nanostructures have very good crystallinity. It has been widely acknowledged that point defects, e.g., oxygen vacancies and impurity atoms are the main defects in ZnO nanocrystals that can not be detected by electron microscopy. However, these point defects largely affect the optical properties such as PL. So far, there is no good way to determine the defect types and numbers quantitatively in nanostructured materials. However, the presence of impurities or point defects in ZnO nanostructures can be revealed by a PL spectrum [9]. Figure 5.1b shows the PL results obtained from the as-prepared ZnO nanostructures. The nanostructures from the hydrothermal method have a large green emission (due to defects, ranging from 450 nm to 650 nm) vs. the band gap emission (near 380 nm) compared to those nanostructures from the thermal evaporation method, indicating that the point defect concentration (such as ionized oxygen vacancies, antisite oxygen and zinc interstitials) [9] is high in the hydrothermal samples. In addition, according to the band gap emission intensity, the ZnO nanostructures from the thermal evaporation method have a high crystalline quality. 97 Figure 5.1 (a) XRD patterns of ZnO NWs fabricated by the hydrothermal method (bottom) and NWs (middle) /nanobelts (top) by thermal vapor deposition. (b) PL spectra of ZnO NWs fabricated by the hydrothermal method (bottom) and NWs (middle) /nanobelts (top) by the thermal vapor deposition. Figure 5.2a-b are TEM images showing the morphologies of as-grown ZnO NWs synthesized by the hydrothermal solution method and the thermal evaporation process, respectively. The NWs from these two synthesis methods show uniform diameters of about 30 nm and smooth surfaces with a unique [0001] growth direction. According to the electron 98 diffraction patterns and high-resolution TEM (HRTEM) images taken from these NWs, the NWs contained few defects and showed high crystallinity. Figure 5.2c-d are low-magnification TEM images of NWs after they had been soaked in SPS for three days. Most ZnO NWs synthesized by the hydrothermal solution method were etched out, leaving only thin tubular shells several nanometers thick, as shown in Figure 5.2c. These shells apparently formed on the wire surface. However, the NWs fabricated by thermal evaporation were stable and still held their original shapes after soaking, with some voids several nanometers in diameter randomly distributed on their surface (see Figure 5.2d). The wires formed through the high temperature process may have been able to persist longer in the SPS solution because they had better crystalline quality than the NWs formed through the hydrothermal process as revealed by our PL studies (Figure 5.1b). ZnO NWs fabricated by thermal evaporation were observed to survive in SPS at least two weeks. Figure 5.2 (a) ZnO NWs fabricated by the hydrothermal method. (b) ZnO NWs fabricated by thermal evaporation. (c) Amorphous thin calcium phosphate shells formed on the surfaces of the ZnO NWs synthesized by the hydrothermal method. These ZnO NWs acted as the templates. 99 (d) Calcium phosphate shells coated on the ZnO NWs synthesized by thermal evaporation. (e) HRTEM image of calcium phosphate shell-ZnO NW structures and the EDS spectrum of the calcium phosphate shell. Scale bars 200 nm for (a) and (b), 100 nm for (c) and (d), 5 nm for (e). As is illustrated in the inset in Figure 5.2d, a very thin layer (several nanometers in thickness) uniformly coated each ZnO NW after immersion. Selected-area electron diffraction and HRTEM (Figure 5.2e) demonstrated that these thin shells were amorphous. The results from EDS indicated that the shells contained a high concentration of phosphorous and calcium. Other elements such as Mg, S, Cl contained in SPS can also be detected as impurities. Some very weak peaks are caused by the background noise and can not be characterized by peaks from any elements. The amorphous calcium phosphate is a precursor of hydroxyapatite and can be stabilized by being crystallized into apatite [119,120]. Biomedical researchers commonly use the formation of apatite on materials soaked in SPS to indicate the material’s bioactivity. The precipitation of amorphous calcium phosphate in these experiments shows that ZnO nanostructures behave as bioactive materials [121]. When ZnO nanostructures are soaked in SPS, ZnO(OH)42- is formed by the reaction ZnO+3H2O = ZnO(OH)42-+2H+, resulting in a negatively charged surface. Takadama et al. [120] have explained that such negative surfaces can induce the formation of amorphous calcium phosphate. Calcium ions in the SPS are attracted to the surface preferentially. This is followed by the arrival of HPO42−, resulting in a hydrated precursor cluster of calcium hydrogen phosphate. The precursor clusters grow by incorporating calcium and phosphate ions from the surrounding SPS. The calcium phosphate phase accumulates uniformly on the surface, forming a thin shell of amorphous calcium phosphate intimately interfaced with the substrate [121]. The fast reaction for the formation of ZnO(OH)42- at defect sites was probably responsible for the obvious etching which formed the voids on the surfaces of the ZnO nanostructures tested in these experiments. The etching and morphology changes were further studied using HRTEM. All the samples in this phase of the study were fabricated by the thermal evaporation method. Apart from the thin shells of amorphous calcium phosphate, numerous voids were clearly visible by TEM. As Figure 5.3a shows, the voids at the NW surface were shaped approximately as right-angled triangles. (See also the enlarged image in the inset.) Because the voids were rather small, it was hard to distinguish their profile shapes embedded in the nanostructures when the sample was thick. The voids in the nanobelts with {0001} as the dominant plane could be seen only as small “spots” with quite poor contrast at low magnification (Figure 5.3b). However, the etching 100 patterns could be observed clearly in the nanobelts grown with other dominant planes, such as (1 100) and (1120) , as shown in Figure 5.3c-e. In these cases, many voids about several nanometers in size were observed all over the nanobelts. Apart from some irregularly shaped voids, most of the voids in the nanobelts with (1 100) dominant planes were isosceles trapezoids (see the inset in Figure 5.3c and 5.3e). In nanobelts with (1120) dominant planes the voids were triangles (Figure 5.3d). It is worthy to noting that the symmetry axes of the void patterns were always along the [0001] direction. Figure 5.3 TEM images of etched ZnO nanostructures: (a) NWs viewed along [1 100] ; (b) nanobelts with a (0001) dominant plane; (c) and (e) with a (1 100) dominant plane; (d) with a (1120) dominant plane. Scale bar 50 nm for (a), (b); 200 nm for (c)-(e). The atomic structure of WZ ZnO has positively and negatively charged polar surfaces due to alternative stacking of O2- and Zn2+ ions along the c axis. Traditionally, CBED has been widely used to determine the polarity of this kind of compound. However, the sample thickness determines the contrast of CBED diffraction patterns [74]. CBED patterns taken from nanometer size (<10nm) samples show no useful contrast. To quantitatively determine the polarity of the surfaces, CBED was supplemented by theoretical simulation. Figure 5.4 shows CBED patterns taken from ZnO nanostructures with the electron beam parallel to the [1 100] direction. These patterns (shown at the top of Figure 5.4a-d) match fairly well with the simulated ones (shown at the bottom of Figure 5.4a-d) produced using JEMS simulation software. The thicknesses for the simulations of Figure 5.4 were 26 nm for 3a, 100nm for 3b, 74nm for 3c, and 28nm for 3d. These CBED results permit identifying the polarity of the surfaces in relation to the shapes and orientations of the etched voids. 101 Figure 5.3a and 5.3c-e show that the [000 1] direction always pointed to the larger flat side of each void perpendicular to the c-axis. All etched ZnO nanostructures investigated in the present work yielded the same results. This indicates that ZnO nanostructures exhibit consistent etching behavior in SPS. Figure 5.5a-b are HRTEM images of the voids viewed along the [1 100] and [1120] directions, respectively. The 3D structure of the voids can be deduced by analyzing the image contrast. Figure 5c-d illustrate the thickness changes of the void along the observing direction. Because a thin foil gives an intense image, the reversion of the profile in areas A and B reveals different concave shapes in the voids. Figure 5.6 shows the structure of the etched voids. The voids are enclosed by the {01 11} and ± (0001) planes, with the sequence of areas S(000 1 ) > S{10 11} > S(0001) . This indicates anisotropic etching at speeds along different r directions V<0001> > Vnr ⊥{10 11} > V[000 1 ] ( n stands for the normal direction of the planes). This relationship of etching speeds is consistent with ZnO’s growth speeds along different directions in neutral solution ( V<0001> > Vnr ⊥{10 1 1} > Vnr ⊥{10 11} > V[000 1 ] ) as observed by Li et al. [7] Because the etching behavior of ZnO crystals in SPS is due to the anisotropy of ZnO crystals and independent of the morphology of the ZnO nanostructure, the observed behavior can be used to determine the polar directions, especially the [0001] and [000 1] directions. This is important for understanding the growth of nano-sized crystals dominated by polar surfaces. By immersing a sample in SPS and examining it with TEM, the polar directions can be identified immediately from the shapes of the etched voids. The immersion time can be shortened to several hours at an elevated etching temperature. Figure 5.4 CBED experimental patterns (left) and simulated patterns (right) for the samples shown in (a) Figure 5.3a, (b) Figure 5.3c, (c) Figure 5.3d, (d) Figure 5.3e. Currently, three TEM techniques are widely applied for determining the polar surfaces of 102 ZnO nanostructures: 1) CBED [74]; 2) HRTEM [122] and 3) the electron energy loss spectrum (EELS) technique [100]. Table 5.2 compares these three methods. The key requirement for forming a high quality CBED pattern is a sample thicker than one extinction distance (about 100 nm at 120 kV) to give a strong dynamic effect. Samples in nanostructure studies are always thinner than 100 nm. In this situation, low accelerating voltages for the electron beam and a cooling holder are used to improve contrast and increase the number of the Kossel-Moellenstedt fringes (or decrease the extinction distance in the material). The identification process is time-consumed and is not useful for the samples less than 10 nm thick. In addition, this method is very sensitive to the defects in the sample. Identifying the polar direction using HRTEM largely depends on the quality of the HRTEM images and the image simulation, as has been discussed by Ding et al [121]. Moreover, the orientation of the nanocrystals severely limits the applicability of this technique. Compared to CBED, the advantage of HRTEM is that it can be used to study samples less than 10 nm thick. The EELS method determinates the polar directions by comparing the element peak intensities in the EELS spectra collected at two beam conditions with proper collection angles. The experimental procedure is complicated, requiring a dedicated and skilled operator, and the results are very sensitive to factors such as contamination, the thickness of the sample and collection angles [74]. Figure 5.5 HRTEM images showing the typical morphologies of the voids viewed along (a) [1 100] and (b) [1120] directions. Insets are corresponding Fourier transform patterns. (c) and (d) are the profile graphs of the outlined areas A and B shown in (a) and (b). Scale bar 5nm for a) and b). 103 Figure 5.6 (a) Schematic model of an etched void. (b) and (c) 3D projections of the voids viewed along the [1 100] and [1120] directions. d) and e) are profile images of cross-sections of (b) and (c) cut perpendicular to the [0001] direction. The chemical etching (CE) method proposed in the present work is, by contrast, relatively simple, reliable, timesaving and easy to manipulate, with less constraints on sample geometry and preparation, equipment, and technique. It can be used to investigate many samples quickly for any sample thicknesses and growth direction. In addition, this method might be applied to other polar crystals if a proper etching solution is chosen. The chemical stability of ZnO nanostructures in SPS has important implications for medical applications. ZnO can potentially be used in biosensors, where a reasonable exposure time is often required for sensing in biological systems. Indeed, working times from several hours to several days or even several weeks are often needed. In addition, the behavior of ZnO nanostructures in vivo is often similar to that of many bioactive materials, which could potentially allow its use in many applications [122]. 5.4 Summary The in vitro chemical stability and etching of ZnO nanostructures in simulated physiological solution (SPS) were studied using electron microscopy. Calcium hydrogen phosphate thin layers were observed to be uniformly deposited on the surfaces of ZnO nanomaterials in SPS. Electron diffraction and high-resolution transmission electron microscopy revealed that the calcium hydrogen phosphate layers were amorphous and had excellent interfacial contact with the nanocrystals. ZnO nanostructures fabricated by thermal 104 evaporation were found to survive much longer in SPS than those fabricated using a hydrothermal solution method. The shapes of the voids formed in the ZnO nanostructures by the etching in SPS can be used to deduce the polar directions of ZnO nanostructures. Table 5.2 A comparison of methods for identifying the polarity of ZnO nano crystals Methods CBED [74] HRTEM [122] EELS [100] Process Record CBED Record HRTEM Collect EELS patterns; images; spectra; Compare with Compare with Compare the simulated atomic model. element peak CE Directly observe by TEM intensities. patterns. Real-time No No No Yes Sample Thickness > 10 Thickness < 5 Proper No limit. geometry nm. Flat and free nm. Flat and thickness. of defects. free of defects. Equipments Normal TEM HRTEM Special EELS Normal TEM Time Yes Yes Yes No Reliability High Low High High Feasibility for No No No Yes consuming bulk quantity of samples 105 Chapter 6 1D TiO2-ZnO Nanohybrids 6.1 Introduction Semiconductor nanostructured composites are of interest in many technological applications, such as biolabels, electroluminescent displays, photochemical solar cells, photo catalysis, and sensors. In recent years, considerable effort has been devoted to combining semiconductor nanoparticles with suitable materials to synergize the properties of both components, which has led to many promising applications such as the enhancement of photocatalytic performance through the deposition of metal or metal islands on TiO2 and ZnO nanoparticles [ 123 , 124 ]. In general, photocatalytic efficiency is limited by the fast recombination of photogenerated charge carriers. In semiconductor/metal nanocomposites, the photo-induced charge carriers are trapped by the metal component, which promotes interfacial charge-transfer processes [125]. Compared to pure TiO2 and ZnO, coupled TiO2/ZnO(TZO) polycrystals display a greater photocatalytic activity, such as in the degradation of phenol, 2-chlorophenol, and pentachlorophenol [126] and the decomposition of salicylic acid [127]. The enhanced photocatalytic activity of coupled TZO can be interpreted based on the spatially separation of electrons and holes [128]. However, most previous work on the combination of TZO are simply core-shell (surface coating) or sandwich (particles randomly adhere to each other) geometries leaving bad interface structures, which limit the interfacial charge-transfer processes of the photogenerated charge carriers and thus the photocatalytic activity of the nanocomposite material [125, 126]. It remains a big challenge to fabricate this kind of nanohybrids with good interface structures. Here, we report a simple method for the preparation of TZO nanohybrid structures by the site-specific deposition of TiO2 on ZnO NRs, tetrapods, and NW arrays. TEM studies revealed each ZnO NR to be assembled with one TiO2 cap at the Zn terminated (0001) surface. The polarity of the ZnO (0001) surface plays an important role in the formation of TZO nanohybrid structures. The TZO nanohybrids show uniform and atomically flat interfaces between ZnO and TiO2 with tunable crystal phases, which can be amorphous, anatase and rutile. These TZO nanohybrid structures are expected to demonstrate an enhanced photocatalytic activity due to the improved structures for a better interfacial charge-transfer/spatially separation process of photogenerated charge carriers. Therefore, we investigate the photocatalytic activities of TZO products and their annealing products with different TiO2 phase structures for the understanding of the relationship between their structures and photocatalytic property. We also find that our methods can be used to assemble 106 TiO2 on the 0001 surface of other ZnO nanostructures such as tetrapods, nanofilms, nanoflowers, and NW arrays whatever methods they are fabricated by. By further annealing at higher temperature we convert TZO NW arrays to Zn2TiO4/ZnO nanostructures. 6.2 Synthesis and characterization of TZO nanohybrids 6.2.1 Experimental section The method for the fabrication of TZO nanohybrid structures has three steps: 1) preparation of ZnO NRs; 5 mL of 0.1 M zinc acetate ethanol solution was mixed with 35mL of 0.5 M NaOH ethanol solution to form a suspension solution. The suspension solution was later transferred to a Teflon-lined stainless steel autoclave (50 ml) and heated at 180 °C for 24 hours. 2) preparation of TiO2 nanotubes; 1 g of anatase TiO2 nanoparticles are treated with a NaOH (10 M) aqueous solution in a Teflon vessel at 150 °C for 12 h. 3) 200 mg TiO2 nanotubes (about 9 nm in diameter and hundreds of nanometers in length; amorphous TiO2 powder can also work but the result is not good as TiO2 nanotubes) are mixed with the final reaction solution in step 1 and heated at 180 °C for another 24 hours. Off-white precipitation products that were obtained at the bottom of the autoclave were sonicated for 30 min and then allowed to stand for 30 minutes to separate into two layers. The upper white TZO and the lower off-white superfluous amorphous TiO2 nanoparticles (NPs) are collected separately and washed with ethanol and DI water several times and then centrifuged and dried at 80 °C for further study. In addition, for comparison, the reactions in the absence of the TiO2 nanotubes (or zinc acetate) were also carried out with other conditions unchanged. 35mL of 0.5 M NaOH ethanol solution was mixed with 5 mL of 0.1 M zinc acetate ethanol solution (or 200 mg of the TiO2 nanotubes) to form a suspension solution. The suspension solution was later transferred into a Teflon-lined stainless steel autoclave (50 ml) and heated at 180 °C for 24 hours. Without the TiO2 nanotubes, only ZnO NWs (with diameters ranging from several nanometers to tens of nanometers and lengths of several micrometers) were obtained. Without zinc acetate, the product consists of amorphous TiO2 nanoparticles only (tens of nanometers in size). The as-prepared samples were characterized by XRD (Philips, PW1813). Morphology and structure characterization were carried out by a SEM (Philips, XL-30) and a HRTEM (JEOL, 2010F) equipped with an EDX. The CBED patterns were recorded by using a TEM (Philips, CM120), and the CBED simulation was performed by using the JEMS simulation software. The differential scanning calorimetry (DSC) and thermogravimetry analysis (TGA) 107 (NETZSCH STA 449C) measurements were carried out in a flowing Ar atmosphere at a constant heating rate of 10 ºC/min. The PL measurements were carried out using a 325 nm He-Cd laser as the exciting light source. The UV-Vis diffuse-reflectance spectrum (DRS) measurement was conducted with a UV-Vis spectrophotometer (Lambda 20) and an integrating sphere (Labsphere) with a sampling spot of 10 mm×10 mm at normal incidence. The surface area was performed by a Brunauer-Emmett-Teller (BET) surface area and pore size analyzer (Beckman Coulter, SA 3100). Chemical bonding information of the samples was studied by a Fourier transform infrared spectrometer (FTIR; Bio-Rad, FTS 600). 6.2.2 Morphology and structure characterization Figure 6.1 are SEM images showing the as-grown ZnO NRs and their hybrid structures (capped by TiO2 nanoparticles). The as-grown NRs are tens of nanometers in diameter and several micrometers in length (Figure 6.1a). After reacting with the TiO2 nanotubes, the ZnO NRs were capped with nanoparticles with the NRs’ diameter and length unchanged (Figure 6.1b). Each NR has one particle assembled at one end only, and no TiO2 nanoparticles formed on the side surfaces of the NRs. The diameter of TiO2 particle is slightly larger than that of the NR attached. Figure 6.1 SEM images of (a) ZnO NRs, (b) ZnO NRs capped with TiO2 particles. Figure 6.2a shows a representative XRD pattern of as-prepared TiO2-ZnO NR hybrid product. The diffraction peaks can be well indexed as a WZ-type hexagonal-phase ZnO with the lattice parameters a=3.252Å and c=5.208 Å, which is in a good agreement with the calculated values a=3.253 and c=5.209 (JCPDS 80-0075). The typical morphology of the 108 fabricated TZO nanohybrids is illustrated in Figure 6.2b. Each NR has one TiO2 particle assembled at one end only (Figure 6.2c). Some TiO2 particles were observed with the nanohybrids in the reaction solution because of the excessive TiO2 source that was used. The selected-area electron diffraction (SAED) pattern (Figure 6.2d) of these nanohybrids shows that the ZnO NRs to be composed of WZ-type hexagonal-phase (JCPDS 80-0075) crystals. Figure 6.2 (a) XRD pattern of the as-prepared nanohybrid product. The reflections of ZnO crystal are marked by the indices. The Al peaks come from the aluminum holder in XRD measurement. (b) TEM image showing the morphology of the fabricated ZnO/TiO2 nanohybrid product. (c) TiO2 nanoparticles assembled at one end of the ZnO NR only. (d) SAED pattern of the nanohybrid product. A high-resolution TEM (HRTEM) study showed the ZnO NRs to be single crystalline, and [0001] to be the preferential growth direction. This is illustrated in the HRTEM image in Figure 6.3a, which clearly shows the lattice spacing of 0.52nm that corresponds to the inter-plane spacing of the (0001) planes of the WZ-type hexagonal ZnO crystals. The ZnO NRs were well crystallized with no impurities being detected within the limit of EDX. The TiO2 particles that assembled to the NRs were identified to be amorphous, and the stoichiometric ratio of Ti to O as measured by EDX was about 1:2 (see Figure 6.3b). The interfaces between the ZnO NRs and 109 TiO2 were flat and clearly visible when the electron beam was perpendicular to the NR axes. We believe that the TiO2 nanoparticles might nucleate and grow on the end surfaces of the ZnO NRs. Another possibility, however, is that TiO2 particles in the solution might fuse preferentially with the NRs tops and attached on the tips. Some cap-like TiO2 particles can be seen in Figure 6.3c. Similar nanoparticles were observed in the experiment without adding zinc acetate. Wurtzite-type ZnO has a polar crystalline structure that consists of Zn-terminated (0001) and O-terminated ( 000 1 ) surfaces. The CBED technique is conventionally used in TEM to determine the polarity of semiconductor compounds [74, 98]. The CBED patterns are formed with a converged electron probe focusing at the sample area in the nanometer range. Figure 6.3d illustrates the CBED pattern along the [1 100] zone axis. Because the diameters of most of the TZO NRs were small, a thick TZO NR of about 80nm in diameter was used with a low accelerating voltage (80 kV) of the electron beam for the CBED study to achieve a better contrast and to allow the determination of the number of Kossel-Moellenstedt fringes (or decrease in the extinction distance of the material) in the CBED pattern. In the CBED study, the thickness of the NR was estimated to be of the same order as the diameter. The fringes and intensities that are shown in diffraction disks (0002) and ( 0002) in Figure 6.3d varied with the sample thickness. It can clearly be seen that the two disks are asymmetrical: the central diffraction fringe in the (0002) disk is a bright single line, whereas the central fringe in the ( 0002) disk consists of bright double lines. This experimental CBED pattern matches fairly well with the simulated pattern (produced using JEMS simulation software) that is shown in Figure 6.3e. The best match was found for the sample thickness near 65 nm (with electron beam energy of 80 kV). We investigated a number of TZO NRs by using CBED and obtained the same result, namely, that the polar surfaces of the NRs that were covered with TiO2 caps had Zn-terminated (0001) surfaces. 110 Figure 6.3 (a) HRTEM image of an individual ZnO/TiO2 structure. (b) EDX spectra recorded by focusing the electron beam on the NR and the TiO2 cap, respectively. The C and Cu signals came from a carbon-supporting film that was prepared on a copper grid. (c) The amorphous caps (the hole of each of which is indicated by the arrows) that may have become detached from the ZnO/TiO2 nanohybrids during TEM sample preparation. (d) The CBED pattern taken along the [1120] direction. (e) The corresponding simulated CBED pattern. 6.3 Growth processes and mechanism of TZO nanohybrids To understand the mechanism of the formation of TZO nanohybrids, we investigated the initial growth process of ZnO NRs and TiO2 nanoparticles. ZnO NRs of lengths of 100-200 nm were formed without any TiO2 deposition at the tips or on the side surfaces within the first hour of reaction (see Figure 6.4a). During the second hour of reaction, a thin layer of amorphous TiO2 gradually appeared on the (0001) planes and at the tips of the ZnO NRs. After three hours, the TiO2 layers became obvious, as marked by the arrows in Figure 6.4b. As the reaction time increased, these thin layers of TiO2 grew to cover the entire tips of the ZnO NRs, developing a cap-like morphology (see Figure 6.2b). It was noted that the length of the NRs increased very 111 slowly after the formation of the amorphous TiO2 on the NR tips. These experimental results confirm that the fastest NR growth occurs along the [0001] direction at a much faster rate than that along the [ 000 1 ] direction [7]. Due to the presence of the TiO2, the growth of the ZnO NRs along the [0001] direction was retarded by the “caps.” Therefore, for the same reaction condition, the ZnO NRs that grew in the presence of TiO2 were much shorter than the ZnO NRs that grew without TiO2. Based on the observed growth condition and structure, we believe the formation of the TZO nanohybrids to be due to the site-specific deposition of TiO2 on the polar planes and tips of the ZnO NRs. It is suggested that for the ZnO with a WZ structure the Zn-terminated polar surface is chemically active in the growth of nanostructures [4]. The inherent asymmetry and anisotropy result in the preferential growth of the crystal along the c-axis, which is terminated by Zn ions [7]. The net dipole moment diverges as the length of the NR increases, and the electrostatic potential increases monotonically, which cannot be compensated by surface reconstruction because ZnO ± (0001) is quite stable [8]. To cancel out the polarity, a rearrangement of the charges on the outermost layers is proposed through the following three principal mechanisms. (1) The creation of surface states and the transfer of negative charges from the O-face to the Zn-face, (2) the removal of surface atoms, and (3) the deposition of positively (negatively) charged impurity atoms on the O (or Zn) surface [ 129]. In this experiment, the formation of the nanohybrids is believed to have been caused by the third compensation mechanism, with the outermost Zn-terminated (0001) faces of the NRs adsorbing the anions and the outermost O-terminated ( 000 1 ) faces adsorbing the cations. Figure 6.5 is the schematic illustration showing the growth process of TZO nanohybrids. Amorphous TiO2 that was dissolved in alkali ethanol at a high temperature and under a high pressure generated a certain concentration of titanium hydrate colloids with negative charges [130]. These negative colloids were captured by the positive Zn-terminated (0001) faces of the NRs, which resulted in the deposition of amorphous TiO2 at the tips and the canceling out of the polarity of the NRs. For the O-terminated ( 000 1 ) tips, however, we found that after the formation of the TiO2 caps at the (0001) tips, the (000 1 ) tips became tapered or rounded. These tapered or rounded growth morphologies is caused by a fast growth along Zn-terminated polar directions (10 1 1) and (10 11) and a relatively slow growth along O-terminated (000 1 ) direction. In addition, we observed that the TiO2 nanocaps could be assembled on ZnO NRs of different diameters. TZO nanohybrids can also be achieved by using pre-synthesized ZnO NRs as the starting materials and reacting them with TiO2 alkali ethanol solution. By controlling the aspect ratios and 112 diameters of the ZnO NR source material, it would be easy to fine-tune the aspect ratios and sizes of the TiO2 caps on the nanohybrids. Figure 6.4 Typical TEM images of the products after (a) 1 hour and (b) 3 hours of reaction. The inset in (b) is an enlarged TEM image to show the big difference of the contrasts for TiO2 particle and ZnO NR. Amorphous particles are marked with arrows. (c) DSC and TGA results for the nanohybrid product. The inset in (c) is the enlarged DSC curve for the temperature range of 250 to 500oC. The scale bar is 100 nm. Figure 6.5 Schematic illustration showing the growth mechanism of TZO nanohybrids. 113 6.4 Phase transition of TiO2 in TZO nanohybrids The annealing of the nanohybrids at different temperatures revealed the occurrence of a phase transition in the TiO2 caps. Figure 6.4c shows the differential scanning calorimetry (DSC) and thermogravimetry analysis (TGA) results for the nanohybrid product. The exothermic peaks with weight loss observed at ~91°C and ~242°C correspond to the removal of physically absorbed water and organic components. The small exothermic peaks that were observed at ~330 °C and ~430 °C without weight loss are associated with the phase transition of TiO2 from an amorphous state to a crystalline state, such as an anatase or rutile structure. As observed in the low magnification TEM images of the nanohybrids (see the insets in Figure 6.7a-b), there is no change on the morphologies of the ZnO NRs and the TiO2 caps after annealing the samples at 300 ºC and 600 ºC. Note that there are no apparent peaks for anatase and rutile TiO2 in the XRD patterns of annealed products (Figure 6.6) since the amount of TiO2 coupled to ZnO NRs are quite small, estimated less than 5 % by weight so that they can not be detected by XRD. Figure 6.6 XRD patterns of TZO, TZO300 and TZO600 products. The apparent sharp peaks are all attributed to WZ ZnO. 114 Figure 6.7 HRTEM images showing the phase transformation by annealing at (a) 300°C and (b) 600°C for 2 hours. The insets show the low magnification TEM images and the FFT patterns of the areas that are marked by the dashed lines in the main picture. The images were taken from the zone axes of [1120] for ZnO, [201] for anatase TiO2 (Figure 6.7a), and [100] for rutile TiO2 (Figure 6.7b). c) Anatase crystallites formed at the outermost shell of the cap after illumination with a convergent electron beam. The insets show images of the TiO2 caps before and after electron beam illumination. The scale bar is 5 nm. Figure 6.7a-b illustrate the HRTEM images of TiO2 caps that were annealed at 300 °C and 600 °C in air, respectively. Generally, the annealed TiO2 caps consisted of several crystallites with almost the same orientation. In studying the HRTEM images and the corresponding FFT, we identified that the amorphous TiO2 was converted to the anatase phase at 300 °C and then to the rutile phase at 600°C (see the insets in Figure 6.7a-b). The HRTEM image in Figure 6.7 was 115 taken with the electron beam parallel to the zone axes of [1120] for ZnO, [201] for the anatase TiO2, and [100] for the rutile TiO2. We observed that the temperature that was needed for the crystallization of the amorphous TiO2 to occur was quite low. This may be attributable to the size effect of the nanoparticles (a low melting temperature compared to the bulk material) or to organic impurities that remained on the surface of the TiO2 caps [131]. When the amorphous TiO2 cap was illuminated by a strong convergent electron beam, the apparent crystallization of the amorphous TiO2 to the anatase structure occurred at the outermost shell of the cap. This is shown in Figure 6.7c, in which the area of crystallization is shaded dark to contrast it with the inner part. The FFT analysis revealed the TiO2 crystallites of the anatase and rutile phases to have the following orientation relationships with the ZnO NRs. 2 anatase || (0001) ZnO, 2 anatase || (112) TiO [201]TiO 2 rutile || (0001) ZnO, and 2 rutile || (011) TiO [100]TiO [1120] ZnO, [1120] ZnO . The inter-plane spacing d (112)/TiO 2 anatase (2.33Å) was close to d (011)/TiO 2 rutile (2.46Å), and both planes were parallel to the (0001) ZnO after the phase transition. We found the crystallites that were formed by annealing or electron beam irradiation to have a very similar orientation relationship with the ZnO NRs. The interface structures of almost all of the nanohybrids were similar and atomically flat. It is expected that such uniform interfaces will generate good electrical contact between the ZnO and TiO2 nanocrystals, which may enhance the interfacial charge-transfer processes of the photogenerated charge carriers and thus the photocatalytic activity of the nanocomposite material. Because different crystal phases of TiO2 show distinct chemical properties, the present synthesis method also provides an effective way to fabricate and control TZO nanocomposite structures for use in a host of technological applications. 116 6.5 Enhanced photocatalytic performance of TZO nanohybrids The UV-Vis absorption spectra of amorphous TiO2, ZnO NRs, TZO nanohybrids and the annealing products TZO300 and TZO600 are compared in Figure 6.8. ZnO NRs and all the TZO products have strong absorption in the UV region with the absorption edge ca. 380 nm, corresponding to the ZnO bandgap 3.26 eV. The amorphous TiO2 NPs have a strong adsorption in the deep-UV region with the absorption edge ca. 313 nm with the corresponding bandgap 3.96 eV [132,133]. The greater absorbance below 320 nm of TZO products ensures that the TiO2 nanoparticles attached to ZnO NRs dominate the deep UV absorption and hence may enhance the use of UV light compared with pure ZnO NRs. Figure 6.8 UV-Vis absorption spectra of amorphous TiO2 NPs, ZnO NRs, TZO nanohybrids and the annealing products TZO300 and TZO600. Previous studies involving semiconductor-semiconductor or semiconductor-metal composites [ 134 ] such as CdS-ZnO, CdS-SWCNT, ZnO-SWCNT, CdS-AgI, CdS-TiO2, TiO2-Au, and ZnO-Au have shown charge equilibration when subjected to bandgap excitation. Photoinduced electron transfer between composites can be established by the emission quenching as shown in these systems. ZnO NRs with a bandgap of 3.26 eV has undergone charge separation under UV-excitation (λ< 380 nm). These charge carriers (eCB and hVB) either directly recombined and/or got trapped at the vacancies (et and ht) [135]. Earlier studies have established that the green emission from ZnO colloids arise from oxygen vacancies [9]. A green emission of at most around 525 nm is a useful probe to monitor 117 the charge transfer processes at the ZnO surface. In the present investigation we monitored the green emission of ZnO to investigate the charge transfer interaction with TiO2. Figure 6.9 shows the PL spectra of ZnO NRs and TZO products. With the combination of TiO2 nanoparticles and fine interfaces, ZnO NRs are expected to intensely interact with the deposited TiO2. As a result we expect a decrease in the emission yield of ZnO. More than ca. 50 % emission quenching of ZnO can be achieved by the assembly of TiO2 nanoparticles on ZnO NRs. It is noted that the emission quenching of TZO300 and TZO600 products cannot simply be attributed to the charger transfer interaction because the PL spectra is largely affected by the annealing treatment; For our case, the green emission intensity decreases with the increase of the annealing temperature for pure ZnO nanostructures in air. Figure 6.9 PL spectra of ZnO NRs, TZO nanohybrids, TZO300 and TZO600 The photocatalytic experiments have been carried out in our home-made photocatalytic reactor system as shown in Figure 6.10. Eighty milligrams of each catalyst was suspended in 200 mL of a methyl blue (MB) aqueous solution (20 ppm), and then the mixture was put into a quartz reaction tank and kept in the dark for 30 min to obtain the equilibrium adsorption state. The concentration of the MB solution slightly decreased while it was kept in the dark, so that the C0 value was slightly smaller than 20 ppm at t=0. UV irradiation was carried out using a 500W high-pressure mercury lamp (the strongest emission being 368 nm) cooled by circulating water. After a given irradiation time, about 3 mL of the mixture was extracted, and the catalysts 118 were separated from the suspensions by centrifuge. The degradation process was monitored by a UV-Vis spectrophotometer (Lambda 20) (measuring the absorption of MB at 664 nm). Figure 6.10 The home-made photocatalytic reactor system To evaluate the photooxidation capability of TZO nanohybrids, we examined the decomposition of the methylic blue (MB) dye in a solution through the samples of TZO, TZO300, and TZO600 under UV light irradiation as a function of time (Figure 6.11). For comparison, we also carried out decomposition of the MB dye in solution over the amorphous TiO2 NPs and ZnO NRs reference photocatalyst through UV light irradiation (Figure 6.11a). To show that the decomposition of MB dye over TZO is neither caused by catalysis nor photolysis, we carried out the decomposition experiment in the dark with TZO (catalysis) and under full arc light irradiation without catalysts (Blank experiment in Figure 6.11a). In these experiments, the MB concentration remained unchanged as a function of time, that is, TZO is actively photocatalytic under UV light. From the almost exponential decay at the initial stage (0~15 mins), it is revealed that the decomposition kinetics essentially follow first order kinetics, with classical equation ln(C/C0) =-k•τ, where k is so-called pseudo-first rate kinetic constant and τ is time. By fitting the curves in the initial 0~15 mins, the k values are obtained and provided in Figure 6.11, which represent a good measure of the overall photo-degradation rate of all investigated structures. It is evident that the TZO nanohybrid structures without post treatment show the fastest decomposition rate of MB with k=0.138 min-1, about five times that of the 119 component ZnO NRs. The photocatalytic activities of annealing products decrease when the annealing temperature increases in the TiO2 phase on the TZO nanohybrid structures which are transformed from amorphous to anatase and rutile. Figure 6.11 (a) Photodegradation of MB by TZO products amorphous, ZnO NRs, amorphous TiO2 NPs and the blank experiment. (b) A comparison of photocatalytic activity of ZnO NRs, TZO, TZO300 and TZO600. 120 Figure 6.12 Illustration of photoinduced charge transfer and separation in the interface of TZO heterostructures. Figure 6.13 FTIR spectra of the as-synthesized TZO, TZO300 and TZO600 samples in the wavenumber ranges of 4000-400 cm-1. The broad absorptions at about 3352 and 1639 cm-1 are assigned to the hydroxyl groups of chemisorbed and/or physisorbed H2O molecules on the samples. A strong absorption band near 540 cm-1 reveals the vibration properties of ZnO NRs. Other unsigned peaks are attributed to remnant organic species in the samples. The enhanced photocatalytic activities of TZO products can be understood as follows. The TZO nanohybrids formed type II semiconductor heterostructures [136] as shown in Figure 6.12 [129]. In this case, the lowest energy states for electrons and holes are found in different 121 semiconductors; therefore, the energy gradient existing at the interfaces tends to spatially separate electrons and holes that are excited by UV light on different sides of the heterostructures. That is, the electrons transfer from the conduction band (CB) of ZnO to the CB of TiO2 under illumination, and the holes transfer from the valence band (VB) of TiO2 to the VB of ZnO. This process isolated active electrons and holes, and hence gives rise to a decrease in the electron-hole pair recombination rate, and an increase in lifespan, which directly results in an intense emission quenching as revealed by our PL results. It also increases the availability of the pairs (electron and hole) on the surface of the photocatalysts and thus enhances the redox process. The fine atomic flat interfaces of our TZO heterostructures also further favor the charge transfer/isolation process. It is noteworthy that the annealing treatment depresses the photocatalytic activities of TZO products. This may be a result of the following reasons. First, as revealed by Figure 6.12, apparently a larger energy gradient/difference between VBZnO and VBTiO2 better enables the active holes to transfer from TiO2 to ZnO. This difference decreases as the bandgap energy (Eg) of TiO2 structures reduce after annealing treatments (Eg/rutile (3.0 eV) < Eg/anatase(3.2 eV) <Eg/amorphous(3.96 eV)), and hence the ability of the charge transfer/separation is also weakened. Second, specific surface area and surface hydroxyl groups can strongly affect photocatalytic activities of nanomaterials and sometimes they are the dominating factor compared to the crystal structure [137]. In our case, the increase in the annealing temperature results in a decrease in the specific surface area of TZO products (TZO 14.643 m2/g; TZO300 13.763 m2/g; TZO600 8.418 m2/g) and the quantity of surface hydroxyl groups (see caption of Figure 6.13), lower the photocatalytic activity. Through this, the reaction rates of annealed TZO products are still larger than or close to that of ZnO NRs. In summary, we have demonstrated a simple method of synthesizing ZnO/TiO2 nanohybrid structures by the site-specific deposition of TiO2 on ZnO NRs. The polarity of the ZnO (0001) surface plays an important role in the formation of ZnO/TiO2 nanohybrid structures. Annealing at different temperatures gives rise to the phase transformation of amorphous TiO2 to anatase phase and rutile phase nanocrystals with good interface structures. Compared with the components of ZnO NRs and amorphous TiO2 nanoparticles, the combined TZO products demonstrate higher catalytic activity. This was explained by an enhanced charge transfer/separation process resulting from the novel type II heterostructures with fine interfaces, which was supported by the emission quenching in the PL studies. The catalytic performance of annealing products varies with the annealing temperatures and results from the combined action of the changes in the Eg of TiO2 phase structures, the specific surface area and the quantity of surface hydroxyl groups. 122 6.6 Assemble TiO2 nanoparticles on other ZnO nanostructures TiO2 nanoparticles can be assembled on the (0001) tip of ZnO nanostructures whatever the fabrication methods and their morphologies are according to our proposed site-specific deposition mechanism. We find there is no difference for the fabrication of TZO nanohybrids using ZnO NWs by hydrothermal methods or by CVD methods. We successfully assemble TiO2 nanoparticles on the tips of pre-prepared ZnO tetrapods (Figure 6.14a), nanowalls (Figure 6.14b), nanoflowers (Figure 6.14c), and NW arrays (Figure 6.14d). Figure 6.14 The SEM images of ZnO nanostructures with TiO2 nanoparticles assembled on their 0001 ends. 123 6.7 Zn2TiO4/ZnO NW heterostructures Zinc titanate (Zn2TiO4) is an inverse spinel, which has been used as a catalyst and pigment in industry. It is one of the leading regenerable catalysts and has been demonstrated to be a good sorbent for removing sulfur-related compounds at high temperatures. As a dielectric material, its physical, electrical, and optical properties have been studied for various applications. Only recently, syntheses of single-crystal Zn2TiO4 NWs have been achieved by using ZnO NWs as a template [138]. The fabrication process is coating ZnO NWs with a layer of Ti/TiO2 by magnet sputter and then annealing them at high temperature to form Zn2TiO4 structures by the so called “Kirkendall effect”. With a similar approach, other ZnO based ternary compound nanotubes and NWs such as ZnFe2O4, ZnAl2O4, ZnGa2O4, Zn2SnO4, and Zn2SiO4 and the like, are fabricated [139]. Figure 6.15 The SEM images of (a) ZnO NW arrays, (b) TZO NW arrays, (c) Zn2TiO4/ZnO NW arrays, and (d) Schematic illustration of the formation process of Zn2TiO4/ZnO heterostructures. Here, we synthesize a new type of Zn2TiO4/ZnO NW heterostructure based on the intended assembly of TiO2 nanoparticles on ZnO NWs. The fabrication process contains three steps: 1) Preparation of ZnO NW arrays on Si substrates as described in Chapter 3. 2) Preparation of TiO 2 124 nanotubes; 1 g of anatase TiO2 nanoparticles are treated with a NaOH (10 M) aqueous solution in a Teflon vessel at 150 °C for 12 h. 3) Preparation of TZO NW arrays; 200 mg TiO2 nanotubes are mixed with the reaction solution for the synthesis of ZnO NRs in Chapter 6.2.1. Then the ZnO NW arrays are placed into the mixed solution and heated at 180 °C for 24 hours. 3) Annealing at 800 ºC in the air for two hours. The typical morphologies of products in each step are shown in Figure 6.15. The tips of the final products seem to be melted and show irregular shapes. Figure 6.16 (a) The low magnification TEM image of a single hydrogen TiO2 hydrate/ZnO NW heterostructure. (b) SAED pattern recorded at the nanoparticle. (c) The HRTEM image of nanoparticle illuminated by electron beam and its FFT. Figure 6.16a shows the typical morphology of a single TiO2/ZnO NW structure. The nanoparticle capped on the ends of ZnO NW shows a regular shape. EDS shows there are only titanium and oxide in the nanoparticles and the SAED pattern (Figure 6.16b) recorded on the 125 nanoparticle shows regular diffraction points, revealing a crystalline structure. We suggest it a hydrogen TiO2 hydrate while we need further studies to confirm its structure. This hydrate is not stable under electron beam illumination and can be converted into a cubic TiO phase (Hongquiite, JCPDS 86-2352) as shown in Figure 6.16c-d. Further TEM studies in Figure 6.17 show that the TiO2 nanoparticles located at the tips of ZnO NWs after the annealing treatment have been converted into a single crystalline Zn2TiO4 with cubic spinel structure. HRTEM recorded at the interface shows that there are two series of oriented relationships between the Zn2TiO4nanoparticles and ZnO NWs: (11 1) Zn (11 1) Zn 2 TiO 4 2 TiO 4 || (0001) ZnO, [220]Zn 2TiO 4 || (0001) ZnO, [220]Zn 2TiO 4 || [1120] ZnO ; || [1 100] ZnO. The above two series of lattice match configurations are illustrated in Figure 6.18. In most cases, hexagonal structured substrates with [0001] orientation are used for heteroepitaxial growth of cubic structured crystals along the [111] direction because of the same triangle lattices and an easy lattice match. The epitaxial relationship is <110> || <11-20> when the lattice mismatch is less than 25%. Otherwise the epitaxial lattice will rotate 30º, that is <110> || <1-100>, to minimize the lattice mismatch lowering the interface energy. In our case, the lattice mismatches in Figure 6.18a-b are -8.75% and 5.82% respectively. The small differences in lattice mismatch make these two epitaxial configurations (Figure 6.18) favorable and are frequently found in our TEM observations. 126 Figure 6.17 (a) and (e) the low magnification TEM images of two single Zn2TiO4/ZnO NW heterostructures with different oriented relationships. (c) and (g) are the HRTEM images near the interface of the heterostructures. (b) and (d) are FFT patterns corresponding to area A and B respectively; (f) and (h) are FFT patterns corresponding to area C and D respectively. 127 Figure 6.18 Schematic illustration of two series of oriented relationships between the spinel Zn2TiO4 and WZ ZnO. Solid-state reactions of type AO + B2O3 AB2O4 are a common method for the fabrication of spinel oxides. Traditional studies on spinel-formation reactions are usually conducted at planar interfaces or in the form of powder mixtures by bringing two solid binary oxides, or a solid oxide and a vapor or liquid phase, into contact at high temperatures (>1000 ºC). The growth process of classical spinel oxides involves the so-called Wagner’s cation counter diffusion mechanism [140], viz. cations migrating through the reaction interface in opposite directions and the oxygen sublattice remaining essentially fixed. This mechanism applies to many types of spinels, for example, ZnFe2O4, MgAl2O4, and Mg2TiO4. Another growth mechanism involves the diffusion of both cations and oxygen into a counterpart in an effective unilateral transfer into the spinel [141]. This means that an inert marker plane placed at the initial interface as found at the ZnO/spinel interface for the ZnO–Al2O3 reaction, whereas in the case of the MgO–Al2O3 reaction, the marker plane is within the spinel layer. However, observation of the marker plane is rarely reported, especially in nanomaterials. Our TiO2/ZnO NW heterostructures with special configuration provide a good sample for the study of spinel growth mechanisms by the observation of the marker plane position. Our HRTEM results show that no Zn2TiO4 structures were found in the wire part revealing that ZnO unilaterally diffuse into spinel Zn2TiO4 with the marker plane at the initial interface of the TiO2/ZnO NW. This is contrary to previously reported results that TiO2 incorporated in the ZnO lattice as a solid solution because of the faster diffusion rate of Ti4+ than that of Zn2+. EDS elemental mapping 128 results of oxygen, titanium and zinc of single TZO nanohybrid and Zn2TiO4/ZnO heterostructures further support our claim (Figure 6.19). Apparently, zinc diffuses from the wire into the TiO2 nanoparticle while the titanium remains at the nanoparticle. We notice that the part of the NW connecting the Zn2TiO4 nanoparticle quickly tapers off (Figure 6.19b). Therefore the formation of the spinel Zn2TiO4 obtains ZnO from the surface diffusion at the tip of ZnO NW. Figure 6.19 TEM of one single TZO and corresponding EDS elemental mapping of oxygen, titanium and zinc of (a) single TZO nanohybrid and (b) Zn2TiO4/ZnO heterostructures In summary, a novel Zn2TiO4/ZnO NW heterostructure is fabricated. ZnO unilaterally transfers into TiO2 by surface diffusion to form the spinel Zn2TiO4 structure. Two series of orientation relationships are observed. 129 Chapter 7 Conclusions & Future work l Conclusions 1D ZnO nanostructures with controlled morphologies, defects and alignment have been fabricated by three simple thermal vapor transfer methods. The crystal structures, interfaces, growth mechanisms and optical properties of ZnO nanostructures have been investigated by SEM, TEM, and PL spectroscopy. The present work is also concerned with the patterned growth and assembly of ZnO nanostructures as well as the stability of ZnO NWs. Firstly, fabrication and patterning of high-quality ZnO NW arrays were realized by high temperature pyrolysis of ZnO under vacuum conditions using a layer of PR coated on a substrate that is stable at a temperature higher than 700ºC. As-fabricated ZnO NWs show excellent alignment, crystal quality, and optical properties that are independent of the substrates used. The Raman study reveals that the PRs are carbonized. HRTEM investigations show that the carbonized PRs are amorphous structure and ZnO is found nucleating with c-axis parallel to the surface of amorphous PR. The carbonized PRs provide perfect nucleation sites for the growth of aligned ZnO NWs and also perfectly connect to the NWs to form ideal electrodes. The c-axis alignment of ZnO NWs is attributed to the texturing of initial ZnO nucleuses at relatively high deposition temperatures. This approach is further extended to realize large area growth of different forms of ZnO NW arrays (e.g., the horizontal growth and multilayered ZnO NW arrays) on other kinds of carbon-based materials. In addition, the as-synthesized vertically aligned ZnO NW arrays show a low weighted reflectance (Rw) and can be used as antireflection coatings. Moreover, non c-axis growth of 1D ZnO nanostructures (e.g., nanochains, nanobrushes and nanobelts) and defect related 1D ZnO nanostructures (e.g., Y-shaped twinned nanobelts and hierarchical nanostructures decorated by flowers induced by screw dislocations) is also present. Secondly, uniform ZnO NWs and tetrapods have been fabricated with a high yield using direct oxidization of pure Zn at high temperatures in air. Spatially-resolved PL measurements were taken on single NWs and ZnO tetrapods fabricated by NSOM. Multiple peaks of defect emissions occurred at the core of the tetrapods and the intensity ratio GL/BE decreases as it gets closer to the leg end. In contrast, no GL is detectable for the NWs, showing that under our zinc-rich growth condition, defect-free NWs could be obtained. The very different PL observed in these two structures strongly suggests that the defects leading to GL originates from structural changes in the tetrapods, and not surface related. On the other hand, those ZnO 130 tapered structures fabricated by a modified carbon thermal method with the assistance of Au catalysts display strong UV emission, indicating a good crystallization quality. Thirdly, two basic issues that will be encountered with when using ZnO nanostructures as nanodevices are studied: 1) the stability, structural degradation and related PL property of ZnO NWs under different environments of surface treatments. For high-quality ZnO NWs, the UV emission shows no change and no DL emission was generated during the structural degradation. For those ZnO NWs showing GL emission, the commonly used treatment methods e.g., post-annealing can not effectively eliminate the GL emission. 2) The chemical stability and biocompatibility of ZnO nanostructures in SPS. ZnO nanostructures fabricated by the thermal evaporation method were found to survive much longer in SPS than those fabricated using a hydrothermal solution method. Calcium hydrogen phosphate amorphous layers structures have been observed to have excellent interfacial contacts with ZnO NWs. The shapes of the voids formed in the ZnO NWs are due to the interesting anisotropic etching behaviors in SPS which can be used to identify the polar directions of ZnO nanocrystals. Finally, TZO nanohybrid structures have been found to form through the site-specific deposition of TiO2 on ZnO NRs using hydrothermal reaction. TEM studies have revealed each ZnO NR to be assembled with one TiO2 cap at the Zn terminated (0001) surface. The polarity of the ZnO (0001) surface plays an important role in the formation of the TZO nanohybrid structures. The TZO nanohybrids contain uniform and atomically flat interfaces between ZnO and TiO2 with tunable crystal phases, which can be amorphous, anatase and rutile through annealing treatments. These nanohybrid structures demonstrate an enhanced photocatalytic activity due to the improved interface structures for a better interfacial charge-transfer/spatially separation process of photogenerated charge carriers. The site-specific deposition method has also been applied to assemble TiO2 on the (0001) surfaces of other ZnO nanostructures such as tetrapods, nanoribbons, nanoflowers and NW arrays produced by different synthesis techniques. Through high temperature annealing, the TZO nanohybrid structures can be further converted into Zn2TiO4/ZnO nanostructures and show certain orientation relationships. HRTEM observations and EDS mapping results reveal that ZnO unilateral transferred into TiO2 by surface diffusion to form the spinel Zn2TiO4 structure. These nanohybrid structures may synergize the properties of both components and lead to many promising applications. 131 l Future work 1. Controlled growth and assembly of ZnO nanostructures by vapor transport process (a) Diameter & density-controlled growth: ultrathin and high aspect ratio. (b) Shape-controlled growth: nanocones and nanopins. (c) Assembly of ZnO NWs: horizontal growth, multilayered growth, tree-like hierarchical growth. (d) Doped ZnO NWs (e) Site-specific patterned growth of single & multi- NW arrays: large area. (f) Growth mechanism: VS & VLS or others (g) NW array growth on transparent glass substrates (h) Controlled growth of non c-axis ZnO nanomaterials 2. Controlled growth and assembly of ZnO NWs by solution methods (a) NW array growth on transparent glass substrates and flexible plastic substrates (b) Site-specific patterned growth of single & multi- NW arrays: large area. 3. ZnO NWs as templates for other nanomaterials and heterostructures Carbon fibers, ternary ZnO based nanostructures, axial heterostructures and radial core-shell heterostructures. 4. Home-made combined SEM/TEM-CL system for the study of optical properties of ZnO nanostructures Realize the growth, reaction, characterization, manipulation, and measurement of nanomaterials etc. all in an electron microscopy. 5. FED, FET, DSSC and heating boiling application based on nanostructures 6. Biosafety, biocompatibility and bio-application of nanostructures 132 Publication Accepted and submitted paper [1] C. Cheng, K. K. Fung, and N. Wang, “Enhanced Photocatalytic Performance of TiO2/ZnO Hetero Nanostructures” submitted to J. Phys. Chem. C [2] L Feng, C.Cheng, B. D. Yao, N. Wang and M. M. T. Loy “Size dependence of excitonic luminescence at low-temperature of single ZnO nanostructures by photoluminescence spectroscopy” submitted to App. Phys. Lett. [3] M. Lei, L. Feng, C. Cheng, M.M.T. Loy and N. Wang “Structural Degradation and Related Photoluminescence Property of ZnO Nanowires” submitted to Solid State Communication. [4] C. D. Gu,, C. Cheng, H. Huang, T. L. Wang, N. Wang, and T. Y. Zhang, Growth and Photocatalytic Activity of Dendrite-like ZnO@Ag Heterostructure Nanocrystals” Crystal Growth & Design (accepted, 2009) [5] C. Cheng, M. Lei, L. Feng, T. L. Wong, K. M. Ho, K. K. Fung, M. M. T. Loy, D. Yu, and N. Wang, High-Quality ZnO Nanowire Arrays Directly Fabricated from Photoresists, ACS Nano 3 (2009) 53-58 [6] C. Cheng, and N. Wang “Synthesis, Characterization and Growth Mechanism of ZnO/TiO2 Nanohybrid Arrays” Mater. Res. Soc. Symp. Proc. 1035 (2008) 1035-L02-11 [7] C. Cheng, R. Xin, Y. Leng, D. Yu, and N. Wang, “ Chemical Stability of ZnO Nanostructures in Simulated Physiological Environments and Its Application in Determining Polar Directions” Inorg. Chem. 47 (2008) 7868-7873 [8] L Feng, C. Cheng, M. Lei, N. Wang, and M. M. T. Loy “Spatially Resolved Photoluminescence Study of Single ZnO Tetrapods” Nanotechnology 19 (2008) 405702 133 [9] C. Cheng, K.F. Yu, Y. Cai, K.K. Fung, and N. Wang, “Site-Specific Deposition of Titanium Oxide on Zinc Oxide Nanorods”, J. Phys. Chem. C 111 (2007)16712-16716 [10] J. Lin, C. Cheng, J. Zhang, Y. Huang, F. J. Shi, X. X. Ding, C. Tang, and S. R. Qi, “ Controllable Growth of Zinc Oxide Micro- and Nanocrystals by Oxidization of Zn-Cu Alloy”, J. Solid State Chem. 178 (2005) 819-824 [11] J. Lin, Y. Huang, X.X. Ding, C. Cheng, C. Tang, and S. R. Qi, “Metal Oxide Coating on Carbon Nanotubes by a Methanol-Thermal Method” J. Nano Sci. Nanotech.5 (2005) 932-936 [12] Z. W. Gan, X. X. Ding, Z. X. Huang, X. T. Huang, C. Cheng, C. Tang, and S. R. Qi “ Growth of Boron Nitride Nanotube Film in Situ” Appl. Phys. A 81 (2004) 527-529 Conference papers [1] “Growth and annealing of ZnO/TiO2 nanohybrids by hydrothermal method” C. Cheng, K.F. Yu, Y. Cai, K.K. Fung, and N. Wang HK IAS - US ICMR Workshop on Advanced Materials, HKUST, Hong Kong. 12 -15 September, 2007 (Poster presentation) [2] “ZnO/TiO2 nanohybrid structures synthesized by site-specific deposition” C. Cheng, and N. Wang 2007 MRS Fall Meeting, Boston, MA, USA 26 - 30 November, 2007 (Oral presentation and conference paper) [3] “Study of the Bioactivity and Stability of ZnO Nanostructures in Biofluids” C. Cheng, R. Xin, Y. Leng, D. Yu, K. K. Fung, and N. Wang, 2008 MRS International Materials Research Conference, Chongqing, China 9 - 12 June, 2008 (Poster presentation) 134 [4] “Study of the Bioactivity and Stability of ZnO Nanostructures in Biofluids” C. Cheng, R. Xin, Y. Leng, D. Yu, K. K. Fung, and N. Wang, Croucher Advanced Study Institute, “Advanced Microscopies: Opportunities and Challenges in Nanomaterial and Surface Research” HKU, Hong Kong 8 -13 December, 2008 (Poster presentation) [5] “Optical and Structural Studies of ZnO Nanostructures” M.M.T. Loy, L. Feng, C. Cheng, M. Lei and N. Wang 2008 International Conference on Nanoscience + Technology (ICN+T), Keystone, CO, USA 21-25 July, 2008 (Presentation) 135 Reference 1. Ozgur, U; Alivov, YI; Liu, C; et al. A comprehensive review of ZnO materials and devices JOURNAL OF APPLIED PHYSICS, 98 (4): Art. No. 041301 AUG 15 2005 2. Minne, SC; Manalis, SR; Quate, CF Parallel atomic force microscopy using cantilevers with integrated piezoresistive sensors and integrated piezoelectric actuators APPLIED PHYSICS LETTERS, 67 (26): 3918-3920 DEC 25 1995 3. SBERVEGLIERI, G RECENT DEVELOPMENTS IN SEMICONDUCTING THIN-FILM GAS SENSORS SENSORS AND ACTUATORS B-CHEMICAL, 23 (2-3): 103-109 FEB 1995 4. Wang, ZL Splendid One-Dimensional Nanostructures of Zinc Oxide: A New Nanomaterial Family for Nanotechnology ACS NANO, 2 (10): 1987-1992 OCT 2008 5. Pan, H; Feng, YP Semiconductor Nanowires and Nanotubes: Effects of Size and Surface-to-Volume Ratio ACS NANO, 2 (11): 2410-2414 NOV 2008 6. Wander, A; Schedin, F; Steadman, P; et al. Stability of polar oxide surfaces PHYSICAL REVIEW LETTERS, 86 (17): 3811-3814 APR 23 2001 7. Li, WJ; Shi, EW; Zhong, WZ; et al. Growth mechanism and growth habit of oxide crystals JOURNAL OF CRYSTAL GROWTH, 203 (1-2): 186-196 MAY 1999 8. Diebold, U; Koplitz, LV; Dulub, O Atomic-scale properties of low-index ZnO surfaces APPLIED SURFACE SCIENCE, 237 (1-4): 336-342 OCT 15 2004 9. Djurisic, AB; Leung, YH Optical properties of ZnO nanostructures SMALL, 2 (8-9): 944-961 AUG 2006 10. Wang, XD; Ding, Y; Summers, CJ; et al. Large-scale synthesis of six-nanometer-wide ZnO nanobelts JOURNAL OF PHYSICAL CHEMISTRY B, 108 (26): 8773-8777 JUL 1 2004 136 11. Hur, TB; Hwang, YH; Kim, HK Quantum confinement in Volmer-Weber-type self-assembled ZnO nanocrystals APPLIED PHYSICS LETTERS, 86 (19): Art. No. 193113 MAY 9 2005 12. Gu, Y; Kuskovsky, IL; Yin, M; et al. Quantum confinement in ZnO nanorods APPLIED PHYSICS LETTERS, 85 (17): 3833-3835 OCT 25 2004 13. Gopel, W; Lampe, U Influence of Defects on the Electronic-Structure of Zinc-Oxide Surfaces PHYSICAL REVIEW B, 22 (12): 6447-6462 1980 14. Hong, SS; Joo, T; Park, WI; et al. Time-resolved photoluminescence of the size-controlled ZnO nanorods APPLIED PHYSICS LETTERS, 83 (20): 4157-4159 NOV 17 2003 15. Wang, RC; Liu, CP; Huang, JL; et al. ZnO symmetric nanosheets integrated with nanowalls APPLIED PHYSICS LETTERS, 87 (5): Art. No. 053103 AUG 1 2005 16. Chen, YQ; Jiang, J; He, ZY; et al. Growth mechanism and characterization of ZnO microbelts and self-assembled MATERIALS LETTERS, 59 (26): 3280-3283 NOV 2005 17. Greene, LE; Law, M; Goldberger, J; et al. Low-temperature wafer-scale production of ZnO nanowire arrays ANGEWANDTE CHEMIE-INTERNATIONAL EDITION, 42 (26): 3031-3034 2003 18. Djurisic, AB; Choy, WCH; Roy, VAL; et al. Photoluminescence and electron paramagnetic resonance of ZnO tetrapod structure ADVANCED FUNCTIONAL MATERIALS, 14 (9): 856-864 SEP 2004 19. Shi, WS; Cheng, B; Zhang, L; et al. Influence of excitation density on photoluminescence of zinc oxide with different morphologies and dimensions JOURNAL OF APPLIED PHYSICS, 98 (8): Art. No. 083502 OCT 15 2005 20. Huang, MH; Wu, YY; Feick, H; et al. Catalytic growth of zinc oxide nanowires by vapor transport ADVANCED MATERIALS, 13 (2): 113-116 JAN 16 2001 137 21. Ng, HT; Chen, B; Li, J; et al. Optical properties of single-crystalline ZnO nanowires on m-sapphire APPLIED PHYSICS LETTERS, 82 (13): 2023-2025 MAR 31 2003 22. Yang, Q; Tang, K; Zuo, J; et al. Synthesis and luminescent property of single-crystal ZnO nanobelts by a simple low temperature evaporation route APPLIED PHYSICS A-MATERIALS SCIENCE & PROCESSING, 79 (8): 1847-1851 DEC 2004 23. Lin, BX; Fu, ZX; Jia, YB Green luminescent center in undoped zinc oxide films deposited on silicon substrates APPLIED PHYSICS LETTERS, 79 (7): 943-945 AUG 13 2001 24. Liu, X; Wu, XH; Cao, H; et al. Growth mechanism and properties of ZnO nanorods synthesized by plasma-enhanced chemical vapor deposition JOURNAL OF APPLIED PHYSICS, 95 (6): 3141-3147 MAR 15 2004 25. Garces, NY; Wang, L; Bai, L; et al. Role of copper in the green luminescence from ZnO crystals APPLIED PHYSICS LETTERS, 81 (4): 622-624 JUL 22 2002 26. Xu, CX; Sun, XW; Zhang, XH; et al. Photoluminescent properties of copper-doped zinc oxide nanowires NANOTECHNOLOGY, 15 (7): 856-861 JUL 2004 27. Reynolds, DC; Look, DC; Jogai, B Fine structure on the green band in ZnO JOURNAL OF APPLIED PHYSICS, 89 (11): 6189-6191 Part 1 JUN 1 2001 28. Van Dijken, A; Meulenkamp, EA; Vanmaekelbergh, D; et al. The kinetics of the radiative and nonradiative processes in nanocrystalline ZnO particles upon photoexcitation JOURNAL OF PHYSICAL CHEMISTRY B, 104 (8): 1715-1723 MAR 2 2000 29. Pearton, SJ; Norton, DP; Heo, YW; et al. ZnO spintronics and nanowire devices JOURNAL OF ELECTRONIC MATERIALS, 35 (5): 862-868 MAY 2006 30. Vanheusden, K; Seager, CH; Warren, WL; et al. Correlation between photoluminescence and oxygen vacancies in ZnO phosphors APPLIED PHYSICS LETTERS, 68 (3): 403-405 JAN 15 1996 138 31. Meng, XQ; Shen, DZ; Zhang, JY; et al. The structural and optical properties of ZnO nanorod arrays SOLID STATE COMMUNICATIONS, 135 (3): 179-182 JUL 2005 32. Wang, YG; Lau, SP; Zhang, XH; et al. Evolution of visible luminescence in ZnO by thermal oxidation of zinc films CHEMICAL PHYSICS LETTERS, 375 (1-2): 113-118 JUN 25 2003 33. Zhao, QX; Klason, P; Willander, M; et al. Deep-level emissions influenced by O and Zn implantations in ZnO APPLIED PHYSICS LETTERS, 87 (21): Art. No. 211912 NOV 21 2005 34. Xu, PS; Sun, YM; Shi, CS; et al. The electronic structure and spectral properties of ZnO and its defects NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESEARCH SECTION B-BEAM INTERACTIONS WITH MATERIALS AND ATOMS, 199: 286-290 JAN 2003 35. Huang, MH; Mao, S; Feick, H; et al. Room-temperature ultraviolet nanowire nanolasers SCIENCE, 292 (5523): 1897-1899 JUN 8 2001 36. Law, M; Sirbuly, DJ; Johnson, JC; et al. Nanoribbon waveguides for subwavelength photonics integration SCIENCE, 305 (5688): 1269-1273 AUG 27 2004 37. Kind, H; Yan, HQ; Messer, B; et al. Nanowire ultraviolet photodetectors and optical switches ADVANCED MATERIALS, 14 (2): 158-+ JAN 16 2002 38. Fan, ZY; Chang, PC; Lu, JG; et al. Photoluminescence and polarized photodetection of single ZnO nanowires APPLIED PHYSICS LETTERS, 85 (25): 6128-6130 DEC 20 2004 39. Soci, C; Zhang, A; Xiang, B; et al. ZnO nanowire UV photodetectors with high internal gain NANO LETTERS, 7 (4): 1003-1009 APR 2007 40. Xiang, B; Wang, PW; Zhang, XZ; et al. Rational synthesis of p-type zinc oxide nanowire arrays using simple chemical vapor deposition NANO LETTERS, 7 (2): 323-328 FEB 2007 139 41. Yuan, GD; Zhang, WJ; Jie, JS; et al. P-type ZnO nanowire arrays NANO LETTERS, 8 (8): 2591-2597 AUG 2008 42. Yang, Y; Sun, XW; Tay, BK; et al. A p-n homojunction ZnO nanorod light-emitting diode formed by As ion implantation APPLIED PHYSICS LETTERS, 93 (25): Art. No. 253107 DEC 22 2008 43. Wadeasa, A; Beegum, SL; Raja, S; et al. The demonstration of hybrid n-ZnO nanorod/p-polymer heterojunction light emitting diodes on glass substrates APPLIED PHYSICS A-MATERIALS SCIENCE & PROCESSING, 95 (3): 807-812 JUN 2009 44. Jeong, MC; Oh, BY; Ham, MH; et al. Electroluminescence from ZnO nanowires in n-ZnO film/ZnO nanowire array/p-GaN film heterojunction light-emitting diodes APPLIED PHYSICS LETTERS, 88 (20): Art. No. 202105 MAY 15 2006 45. Park, WI; Yi, GC Electroluminescence in n-ZnO nanorod arrays vertically grown on p-GaN ADVANCED MATERIALS, 16 (1): 87-+ JAN 5 2004 46. Gonzalez-Valls, I; Lira-Cantu, M Vertically-aligned nanostructures of ZnO for excitonic solar cells: a review ENERGY & ENVIRONMENTAL SCIENCE, 2 (1): 19-34 2009 47. Keis, K; Magnusson, E; Lindstrom, H; et al. A 5% efficient photo electrochemical solar cell based on nanostructured ZnO electrodes SOLAR ENERGY MATERIALS AND SOLAR CELLS, 73 (1): 51-58 MAY 2002 48. Chiba, Y; Islam, A; Watanabe, Y; et al. Dye-sensitized solar cells with conversion efficiency of 11.1% JAPANESE JOURNAL OF APPLIED PHYSICS PART 2-LETTERS & EXPRESS LETTERS, 45 (24-28): L638-L640 JUL 2006 49. Zhao, MH; Wang, ZL; Mao, SX Piezoelectric characterization of individual zinc oxide nanobelt probed by piezoresponse force microscope NANO LETTERS, 4 (4): 587-590 APR 2004 140 50. Wang, XD; Song, JH; Liu, J; et al. Direct-current nanogenerator driven by ultrasonic waves SCIENCE, 316 (5821): 102-105 APR 6 2007 51. Lee, YJ; Ruby, DS; Peters, DW; et al. ZnO nanostructures as efficient antireflection layers in solar cells NANO LETTERS, 8 (5): 1501-1505 MAY 2008 52. Wan, Q; Yu, K; Wang, TH; et al. Low-field electron emission from tetrapod-like ZnO nanostructures synthesized by rapid evaporation APPLIED PHYSICS LETTERS, 83 (11): 2253-2255 SEP 15 2003 53. Fan, ZY; Wang, DW; Chang, PC; et al. ZnO nanowire field-effect transistor and oxygen sensing property APPLIED PHYSICS LETTERS, 85 (24): 5923-5925 DEC 13 2004 54. Lupan, O; Chai, G; Chow, L Fabrication of ZnO nanorod-based hydrogen gas nanosensor MICROELECTRONICS JOURNAL, 38 (12): 1211-1216 DEC 2007 55. Wang, ZL ZnO nanowire and nanobelt platform for nanotechnology MATERIALS SCIENCE & ENGINEERING R-REPORTS, 64 (3-4): 33-71 APR 3 2009 56. Wang, ZL; Kong, XY; Zuo, JM Induced growth of asymmetric nanocantilever arrays on polar surfaces PHYSICAL REVIEW LETTERS, 91 (18): Art. No. 185502 OCT 31 2003 57. BURTON, WK; CABRERA, N; FRANK, FC ROLE OF DISLOCATIONS IN CRYSTAL GROWTH NATURE, 163 (4141): 398-399 1949 58. A.H. Cottrell, Dislocations and Plastic Flow in Crystals, Oxford University Press, London, 1963. 59. Fan, ZY; Lu, JG Zinc oxide nanostructures: Synthesis and properties JOURNAL OF NANOSCIENCE AND NANOTECHNOLOGY, 5 (10): 1561-1573 OCT 2005 60. Jun, YW; Choi, JS; Cheon, J Shape control of semiconductor and metal oxide nanocrystals through nonhydrolytic colloidal routes ANGEWANDTE CHEMIE-INTERNATIONAL EDITION, 45 (21): 3414-3439 2006 141 61. Pacholski, C; Kornowski, A; Weller, H Self-assembly of ZnO: From nanodots, to nanorods ANGEWANDTE CHEMIE-INTERNATIONAL EDITION, 41 (7): 1188-+ 2002 62. Gao, PX; Wang, ZL Substrate atomic-termination-induced anisotropic growth of ZnO nanowires/nanorods by the VLS process JOURNAL OF PHYSICAL CHEMISTRY B, 108 (23): 7534-7537 JUN 10 2004 63. Ramirez, D; Pauporte, T; Gomez, H; et al. Electrochemical growth of ZnO nanowires inside nanoporous alumina templates. A comparison with metallic Zn nanowires growth PHYSICA STATUS SOLIDI A-APPLICATIONS AND MATERIALS SCIENCE, 205 (10): 2371-2375 OCT 2008 64. Qiu, YJ; Yu, J; Zhou, XS; et al. Synthesis of Porous NiO and ZnO Submicro- and Nanofibers from Electrospun Polymer Fiber Templates NANOSCALE RESEARCH LETTERS, 4 (2): 173-177 FEB 2009 65. Fan, HJ; Werner, P; Zacharias, M Semiconductor nanowires: From self-organization to patterned growth SMALL, 2 (6): 700-717 JUN 2006 66. Xu, S; Wei, Y; Kirkham, M; et al. Patterned Growth of Vertically Aligned ZnO Nanowire Arrays on Inorganic Substrates at Low Temperature without Catalyst JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 130 (45): 14958-+ NOV 12 2008 67. Daisuke Shindo & Tetsuo Oikawa Analytical electron microscopy for materials science Originally published in Japanese by Kyoritsu Shuppan Co., Ltd., Tokyo, JAPAN, 1999 68. Spence, J. C. H.; Zuo, J. M. Electron Microdiffraction Plenum Press: New York 1992 69. Buxton, Bf; Eades, Ja; Steeds, Jw; Et al. Symmetry of Electron-Diffraction Zone Axis Patterns PHILOSOPHICAL TRANSACTIONS OF THE ROYAL SOCIETY OF LONDON SERIES A-MATHEMATICAL SCIENCES, 281 (1301): 171-& 1976 142 PHYSICAL AND ENGINEERING 70. Tanaka, M.; Terauchi, M.; Kaneyama, T. Convergent beam Electron Diffraction I-III, JEOL: Tokyo, 1988. 71. Durig, U; Pohl, Dw; Rohner, F Near-Field Optical-Scanning Microscopy JOURNAL OF APPLIED PHYSICS, 59 (10): 3318-3327 MAY 15 1986 72. Feng, L; Cheng, C; Lei, M; et al. Spatially resolved photoluminescence study of single ZnO tetrapods NANOTECHNOLOGY, 19 (40): Art. No. 405702 OCT 8 2008 73. Ding, Y; Wang, ZL; Sun, TJ; et al. Zinc-blende ZnO and its role in nucleating wurtzite tetrapods and twinned nanowires APPLIED PHYSICS LETTERS, 90 (15): Art. No. 153510 APR 9 2007 74. Cheng, C; Yu, KF; Cai, Y; et al. Site-specific deposition of titanium oxide on zinc oxide nanorods JOURNAL OF PHYSICAL CHEMISTRY C, 111 (45): 16712-16716 NOV 15 2007 75. Liu, YZ; Yuan, HT; Zeng, ZQ; et al. Inversion domain boundary in a ZnO film PHILOSOPHICAL MAGAZINE LETTERS, 87 (9): 687-693 2007 76. Chang, PC; Chien, CJ; Stichtenoth, D; et al. Finite size effect in ZnO nanowires APPLIED PHYSICS LETTERS, 90 (11): Art. No. 113101 MAR 12 2007 77. Djurisic, AB; Choy, WCH; Roy, VAL; et al. Photoluminescence and electron paramagnetic resonance of ZnO tetrapod structure ADVANCED FUNCTIONAL MATERIALS, 14 (9): 856-864 SEP 2004 78. Wang, ZG; Zu, XT; Zhu, S; et al. Green luminescence originates from surface defects in ZnO nanoparticles PHYSICA E-LOW-DIMENSIONAL SYSTEMS & NANOSTRUCTURES, 35 (1): 199-202 AUG 2006 79. Yu, WD; Li, XM; Gao, XD Homogeneous-catalytic synthesis of tetrapodlike ZnO nanocrystals and their photoluminescence properties CHEMICAL PHYSICS LETTERS, 390 (1-3): 296-300 MAY 21 2004 143 80. Liu, M; Kitai, Ah; Mascher, P Point-Defects and Luminescence-Centers in Zinc-Oxide and Zinc-Oxide Doped With Manganese JOURNAL OF LUMINESCENCE, 54 (1): 35-42 AUG 1992 81. Zollfrank, C; Rambo, CR; Batentschuk, M; et al. Spatially resolved luminescence properties of ZnO tetrapods JOURNAL OF MATERIALS SCIENCE, 42 (15): 6325-6330 AUG 2007 82. Stichtenoth, D; Ronning, C; Niermann, T; et al. Optical size effects in ultrathin ZnO nanowires NANOTECHNOLOGY, 18 (43): Art. No. 435701 OCT 31 2007 83. Yao, BD; Chan, YF; Wang, N Formation of ZnO nanostructures by a simple way of thermal evaporation APPLIED PHYSICS LETTERS, 81 (4): 757-759 JUL 22 2002 84. Makarov, AV; Zbezhneva, SG; Kovalenko, VV; et al. Mass spectrometric study of nanocrystalline ZnO vaporization INORGANIC MATERIALS, 39 (6): 594-598 JUN 2003 85. Lv, YY; Wang, WZ; Liu, SY; Ji ZP Inorganic Chemistry Series, Vol. VI, Zinc race Science Press, China, 1995 86. Yan, YH; Chan-Park, MB; Zhang, Q Advances in carbon-nanotube assembly SMALL, 3 (1): 24-42 JAN 2007 87. Lee, C; Bae, SY; Mobasser, S; et al. A novel silicon nanotips antireflection surface for the micro sun sensor NANO LETTERS, 5 (12): 2438-2442 DEC 2005 88. Li, Y; Lee, EJ; Cai, W; et al. Unconventional method for morphology-controlled carbonaceous nanoarrays based on electron irradiation of a polystyrene colloidal monolayer ACS NANO, 2 (6): 1108-1112 JUN 2008 89. Sun, Y; Fuge, GM; Ashfold, MNR Growth of aligned ZnO nanorod arrays by catalyst-free pulsed laser deposition methods CHEMICAL PHYSICS LETTERS, 396 (1-3): 21-26 SEP 21 2004 144 90. Wei, YG; Ding, Y; Li, C; et al. Growth of Vertically Aligned ZnO Nanobelt Arrays on GaN Substrate JOURNAL OF PHYSICAL CHEMISTRY C, 112 (48): 18935-18937 DEC 4 2008 91. Greene, LE; Law, M; Tan, DH; et al. General route to vertical ZnO nanowire arrays using textured ZnO seeds NANO LETTERS, 5 (7): 1231-1236 JUL 2005 92. Claeyssens, F; Freeman, CL; Allan, NL; et al. Growth of ZnO thin films - experiment and theory JOURNAL OF MATERIALS CHEMISTRY, 15 (1): 139-148 2005 93. Ng, HT; Li, J; Smith, MK; et al. Growth of epitaxial nanowires at the junctions of nanowalls SCIENCE, 300 (5623): 1249-1249 MAY 23 2003 94. Park, BY; Taherabadi, L; Wang, CL; et al. Electrical properties and shrinkage of carbonized photoresist films and the implications for carbon microelectromechanical systems devices in conductive media JOURNAL OF THE ELECTROCHEMICAL SOCIETY, 152 (12): J136-J143 2005 95. Lao, JY; Huang, JY; Wang, DZ; et al. Hierarchical oxide nanostructures JOURNAL OF MATERIALS CHEMISTRY, 14 (4): 770-773 FEB 21 2004 96. Zhang, P; Qi, H; Zhang, X; et al. Field emission from lotiform-like ZnO nanostructures synthesized via thermal evaporation method APPLIED PHYSICS A-MATERIALS SCIENCE & PROCESSING, 84 (1-2): 165-169 JUL 2006 97. Lao, JY; Huang, JY; Wang, DZ; et al. ZnO nanobridges and nanonails NANO LETTERS, 3 (2): 235-238 FEB 2003 98. Zou, K; Qi, XY; Duan, XF; et al. Sb-induced bicrystal ZnO nanobelts APPLIED PHYSICS LETTERS, 86 (1): Art. No. 013103 JAN 3 2005 99. Dai, Y; Zhang, Y; Bai, YQ; et al. Bicrystalline zinc oxide nanowires CHEMICAL PHYSICS LETTERS, 375 (1-2): 96-101 JUN 25 2003 145 100.Zhang, ZH; Liu, HH; Jian, JK; et al. Transmission electron microscopy investigation of self-assembly ZnO twinning nanostructures APPLIED PHYSICS LETTERS, 88 (19): Art. No. 193101 MAY 8 2006 101.Xu, C; Chun, J; Rho, K; et al. Ferromagnetic ZnO bicrystal nanobelts fabricated in low temperature APPLIED PHYSICS LETTERS, 89 (9): Art. No. 093117 AUG 28 2006 102.Xu, C; Chun, J; Kim, D; et al. Structural characterization and low temperature growth of ferromagnetic Bi-Cu codoped ZnO bicrystal nanowires APPLIED PHYSICS LETTERS, 91 (15): Art. No. 153104 OCT 8 2007 103.Oba, F; Sato, Y; Yamamoto, T; et al. Current-voltage characteristics of cobalt-doped inversion boundaries in zinc oxide bicrystals JOURNAL OF THE AMERICAN CERAMIC SOCIETY, 86 (9): 1616-1618 SEP 2003 104.Huang, BH; Chen, SY; Shen, PY {10 11} and {1121} -specific growth and twinning of ZnO whiskers JOURNAL OF PHYSICAL CHEMISTRY C, 112 (4): 1064-1071 JAN 31 2008 105.Wang, N; Cai, Y; Zhang, RQ Growth of nanowires MATERIALS SCIENCE & ENGINEERING R-REPORTS, 60 (1-6): 1-51 MAR 31 2008 106.Shan, FK; Liu, GX; Lee, WJ; et al. Aging effect and origin of deep-level emission in ZnO thin film deposited by pulsed laser deposition APPLIED PHYSICS LETTERS, 86 (22): Art. No. 221910 MAY 30 2005 107.Pan, N; Wang, XP; Li, M; et al. Strong surface effect on cathodoluminescence of an individual tapered ZnO nanorod JOURNAL OF PHYSICAL CHEMISTRY C, 111 (46): 17265-17267 NOV 22 2007 108.Chung, TF; Zapien, JA; Lee, ST Luminescent properties of ZnO nanorod Arrays grown on Al : ZnO buffer layer JOURNAL OF PHYSICAL CHEMISTRY C, 112 (3): 820-824 JAN 24 2008 146 109.Kang, JS; Kang, HS; Pang, SS; et al. Investigation on the origin of green luminescence from laser-ablated ZnO thin film THIN SOLID FILMS, 443 (1-2): 5-8 OCT 22 2003 110.Hahm, J; Lieber, CM Direct ultrasensitive electrical detection of DNA and DNA sequence variations using nanowire nanosensors NANO LETTERS, 4 (1): 51-54 JAN 2004 111.Wang, J Carbon-nanotube based electrochemical biosensors: A review ELECTROANALYSIS, 17 (1): 7-14 JAN 2005 112.Sun, XW; Wang, JX; Wei, A Zinc oxide nanostructured biosensor for glucose detection JOURNAL OF MATERIALS SCIENCE & TECHNOLOGY, 24 (4): 649-656 JUL 2008 113.Dorfman, A; Kumar, N; Hahm, J Nanoscale ZnO-enhanced fluorescence detection of protein interactions ADVANCED MATERIALS, 18 (20): 2685-+ OCT 17 2006 114.Khopade, AJ; Arulsudar, N; Khopade, SA; et al. Ultrathin antibiotic walled microcapsules BIOMACROMOLECULES, 6 (1): 229-234 JAN-FEB 2005 115.Wang, B; Feng, WY; Wang, M; et al. Acute toxicological impact of nano- and submicro-scaled zinc oxide powder on healthy adult mice JOURNAL OF NANOPARTICLE RESEARCH, 10 (2): 263-276 FEB 2008 116.Sayes, CM; Reed, KL; Warheit, DB Assessing toxicity of fine and nanoparticles: Comparing in vitro measurements to in vivo pulmonary toxicity profiles TOXICOLOGICAL SCIENCES, 97 (1): 163-180 MAY 2007 117.Zhou, J; Xu, NS; Wang, ZL Dissolving behavior and stability of ZnO wires iin biofluids: A study on biodegradability and biocompatibility of ZnO nanostructures ADVANCED MATERIALS, 18 (18): 2432-+ SEP 18 2006 118.Wesselkamper, SC; Chen, LC; Gordon, T Quantitative trait analysis of the development of pulmonary tolerance to inhaled zinc oxide in mice RESPIRATORY RESEARCH, 6: Art. No. 73 JUL 18 2005 147 119.Kim, HM; Himeno, T; Kokubo, T; et al. Process and kinetics of bonelike apatite formation on sintered hydroxyapatite in a simulated body fluid BIOMATERIALS, 26 (21): 4366-4373 JUL 2005 120.Takadama, H; Kim, HM; Kokubo, T; et al. Mechanism of biomineralization of apatite on a sodium silicate glass: TEM-EDX study in vitro CHEMISTRY OF MATERIALS, 13 (3): 1108-1113 MAR 2001 121.Liu, XY; Fu, RKY; Chu, PK; et al. Mechanism of apatite formation on hydrogen plasma-implanted single-crystal silicon APPLIED PHYSICS LETTERS, 85 (16): 3623-3625 OCT 18 2004 122.Ding, Y; Wang, ZL Profile imaging of reconstructed polar and non-polar surfaces of ZnO SURFACE SCIENCE, 601 (2): 425-433 JAN 15 2007 123.Kenny, S; Buggy, M; Hill, RG The influence of hydroxyapatite: Zinc oxide ratio on the setting behavior and mechanical properties of polyalkenoate cements JOURNAL OF MATERIALS SCIENCE-MATERIALS IN MEDICINE, 12 (10-12): 901-904 2001 124.Pacholski, C; Kornowski, A; Weller, H Site-specific photodeposition of silver on ZnO nanorods ANGEWANDTE CHEMIE-INTERNATIONAL EDITION, 43 (36): 4774-4777 2004 125.Cozzoli, PD; Curri, ML; Giannini, C; et al. Synthesis of TiO2-Au composites by titania-nanorod-assisted generation of gold nanopartictes at aqueous/nonpolar interfaces SMALL, 2 (3): 413-421 MAR 2006 126.Jakob, M; Levanon, H; Kamat, PV Charge distribution between UV-irradiated TiO2 and gold nanoparticles: Determination of shift in the Fermi level NANO LETTERS, 3 (3): 353-358 MAR 2003 127.Serpone, N; Maruthamuthu, P; Pichat, P; et al. Exploiting The Interparticle Electron-Transfer Process in The Photocatalyzed Oxidation of Phenol, 2-Chlorophenol and Pentachlorophenol - Chemical Evidence For Electron and Hole Transfer between Coupled 148 Semiconductors JOURNAL OF PHOTOCHEMISTRY AND PHOTOBIOLOGY A-CHEMISTRY, 85 (3): 247-255 JAN 25 1995 128.Sukharev, V; Kershaw, R Concerning the role of oxygen in photocatalytic decomposition of salicylic acid in water JOURNAL OF PHOTOCHEMISTRY AND PHOTOBIOLOGY A-CHEMISTRY, 98 (3): 165-169 AUG 23 1996 129.Marci, G; Augugliaro, V; Lopez-Munoz, MJ; et al. Preparation characterization and photocatalytic activity of polycrystalline ZnO/TiO2 systems. 1. Surface and bulk characterization JOURNAL OF PHYSICAL CHEMISTRY B, 105 (5): 1026-1032 FEB 8 2001 130.Dulub, O; Diebold, U; Kresse, G Novel stabilization mechanism on polar surfaces: ZnO(0001)-Zn PHYSICAL REVIEW LETTERS, 90 (1): Art. No. 016102 JAN 10 2003 131.Yang, HG; Li, CZ; Gu, HC; et al. Rheological behavior of titanium dioxide suspensions JOURNAL OF COLLOID AND INTERFACE SCIENCE, 236 (1): 96-103 APR 1 2001 132.John, AK; Surender, GD Low temperature process for the synthesis of rutile phase titania through vapor phase hydrolysis JOURNAL OF MATERIALS SCIENCE, 40 (11): 2999-3001 JUN 2005 133.Naik, V. M.; Haddad, D.; Naik, R.; et al Optical Properties of Anatase, Rutile and Amorphous Phases of TiO2 Thin Films Grown at Room Temperature by RF Magnetron Sputtering Mat. Res. Soc. Symp. Proc. 755, DD11.12.1, 2003 134.King, DM; Du, XH; Cavanagh, AS; et al. Quantum confinement in amorphous TiO2 films studied via atomic layer deposition NANOTECHNOLOGY, 19 (44): Art. No. 445401 NOV 5 2008 135.Casavola, M; Buonsanti, R; Caputo, G; et al. Colloidal strategies for preparing oxide-based hybrid nanocrystals EUROPEAN JOURNAL OF INORGANIC CHEMISTRY, (6): 837-854 FEB 2008 149 136.Vietmeyer, F; Seger, B; Kamat, PV Anchoring ZnO particles on functionalized single wall carbon nanotubes. Excited state interactions and charge collection ADVANCED MATERIALS, 19 (19): 2935-+ OCT 5 2007 137.Ivanov, SA; Piryatinski, A; Nanda, J; et al. Type-II core/shell CdS/ZnSe nanocrystals: Synthesis, electronic structures, and spectroscopic properties JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 129 (38): 11708-11719 SEP 26 2007 138.Li, Y; Sasaki, T; Shimizu, Y; et al. Hexagonal-Close-Packed, Hierarchical Amorphous TiO2 Nanocolumn Arrays: Transferability, Enhanced Photocatalytic Activity, and Superamphiphilicity without UV Irradiation JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 130 (44): 14755-14762 NOV 5 2008 139.Yang, Y; Sun, XW; Tay, BK; et al. Twinned Zn2TiO4 spinel nanowires using ZnO nanowires as a template ADVANCED MATERIALS, 19 (14): 1839-+ JUL 16 2007 140.Fan, HJ; Yang, Y; Zacharias, M ZnO-based ternary compound nanotubes and nanowires JOURNAL OF MATERIALS CHEMISTRY, 19 (7): 885-900 2009 141.Carter Re Mechanism of Solid-State Reaction between Magnesium Oxide and Aluminum Oxide and Between Magnesium Oxide and Ferric Oxide JOURNAL OF THE AMERICAN CERAMIC SOCIETY 44 : 116 1961 142.Branson Dl Kinetics and Mechanism of Reaction Between Zinc Oxide and Aluminum Oxide JOURNAL OF THE AMERICAN CERAMIC SOCIETY 48 : 591 1965 150