Single-Crystalline ZnS Nanobelts as Ultraviolet-Light Sensors

Transcription

1 Single-Crystalline ZnS Nanobelts as Ultraviolet-Light Sensors By Xiaosheng Fang,* Yoshio Bando, Meiyong Liao, Ujjal K. Gautam, Chunyi Zhi, Benjamin Dierre, Baodan Liu, Tianyou Zhai, Takashi Sekiguchi, Yasuo Koide, and Dmitri Golberg [*] Dr. X. S. Fang, Prof. Y. Bando, Dr. U. K. Gautam, Dr. C. Y. Zhi, Dr. B. D. Liu, Dr. T. Y. Zhai, Prof. D. Golberg International Center for Young Scientists (ICYS) and World Premier International Center for Materials Nanoarchitectonics (MANA) National Institute for Materials Science (NIMS) Namiki 1-1, Tsukuba, Ibaraki (Japan) Dr. M. Y. Liao, Prof. Y. Koide Sensor Materials Center, NIMS Namiki 1-1, Tsukuba, Ibaraki (Japan) B. Dierre, Prof. T. Sekiguchi Advanced Electronic Materials Center, NIMS Namiki 1-1, Tsukuba, Ibaraki (Japan) DOI: /adma Semiconductor nanostructures, such as nanowires, nanorods, and nanobelts, are attractive building blocks for a new generation of high sensitivity and selectivity sensors primarily because of their high surface-to-volume ratios and diverse functions as both device elements and interconnects. [1 10] To date, nanostructurebased chemical, biological, gas, and UV-light sensors have been fabricated and investigated with respect to their electrical and optical properties. [11 15] UV-light sensors measure the power or intensity of incident UV radiation. This form of electromagnetic radiation has shorter wavelengths than visible light, but still longer than X-rays. According to existing classification of UV light with respect to its wavelength, the longest wavelength of the UV-A band ( nm) can reach the Earth s surface. Molecules in sunscreen lotions absorb most of the UV-B ( nm) light and protect the human skin quite efficiently, just as the molecules of the Earth s atmosphere completely absorb the UV-C ( nm) radiation, not allowing it to reach the ground. Many researches showed that UV-A light may cause a skin cancer. Therefore, the development of novel effective UV sensors that exhibit high UV-A sensitivity but are blind towards standard visible radiation is of great importance. With a wide band-gap of 3.72 and 3.77 ev for cubic zinc blend (ZB) and hexagonal wurtzite (WZ) forms, a diverse range of possible structures and morphologies, and superior chemical and thermal stabilities, ZnS provides a novel prospective alternative for UV detectors that would be particularly useful within the UV-A band. [16 18] Being compared with all existing alternative materials, such as indirect band-gap diamond (5.5 ev) and direct bandgap semiconductor ZnO (3.4 ev), ZnS has a higher potential as a UV detector in this specific wavelength regime. [19,20] Despite the abundant research on the growth and optical properties of 1D ZnS nanostructures, [21 23] a study on nanostructure-based devices (in particular, UV sensors) has not yet been initiated, and presents a challenge. Very recently, we have developed a facile and effective thermalevaporation route for the synthesis of multiangular-branched ZnS nanostructures composed of needle-shaped tips with sharp UVlight band-gap emission and broad visible-light emission. [24] Motivated by these results, we have systematically investigated the controlled growth of single-crystalline ZnS nanobelts and their cathodoluminescence (CL) properties. Herein, we designed and fabricated the first ZnS-nanobelt visible-light-blind UV-light sensors and documented their sharp UV-light emission at room temperature (RT). Enhanced stability and sensitivity of the UV sensors composed of multiple ZnS-nanobelt devices compared to the individual nanobelt-based sensors was demonstrated. Our results indeed demonstrate the uniqueness and effectiveness of ZnS nanobelts for photosensing applications. Therefore, the present single-crystalline ZnS nanobelts are envisaged to become prime candidates for future applications in nanoscale photosensors and/or photodetectors with ultrahigh sensitivity and selectivity. Figure 1 shows scanning electron microscopy (SEM) images of the as-grown ZnS nanobelts under different magnifications. A low-magnification SEM image in Figure 1a reveals that the nanobelts grow from several separated nucleation sites that seem to be connected. The inset in Figure 1a is an image of the asgrown ZnS nanobelts on an Au-coated Si substrate. These images show that the nanobelts can grow up to a length of one millimeter. Moreover, the nanobelts cover the whole substrate, and their yield varies from place to place depending on local temperature. The corresponding high-magnification SEM image in Figure 1b shows that the nanobelts have a width of up to a micrometer. The distribution of the present ZnS nanobelts is nearly similar to the aligned ultralong ZnO nanobelts, which were synthesized on an Au sheet through a one-step process using molten-salt-assisted template-free thermal evaporation. [25] A transmission electron microscopy (TEM) image reveals that the ZnS nanostructures have a belt-like geometry and are transparent (Fig. 2a). Statistics based on numerous TEM images (more than 100 nanostructures were analyzed) indicates that the typical widths of the nanobelts are in the range between 200 nm to 1 mm. An X-ray energy-dispersive spectroscopy (EDS) spectrum (Fig. 2b) acquired from an individual nanobelt and an X-Ray diffraction (XRD) pattern taken from the overall product (no shown here) confirm that the nanobelts consist of Zn and S with a stoichiometric ZnS composition. All of the diffraction peaks can be indexed to WZ-type ZnS within the experimental error. The Cu peak in Figure 2b is due to the TEM grid. The highly 2034 ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2009, 21,

2 Figure 1. a) Low-magnification SEM image. The inset in a) is an image of the as-grown ZnS nanobelts on an Au-coated Si substrate, showing that the nanobelts can grow up to 1 mm in length. b) High-magnification SEM image, showing that the nanobelts have widths of up to 1 mm. crystalline nature of the nanobelts was further verified by highresolution (HR) TEM and selected-area electron diffraction (SAED). Figure 2c presents a HRTEM image in a relatively large spot on the nanobelt, previously shown in Figure 2a. An enlarged lattice-resolved HRTEM image in Figure 2d was obtained in the marked area in Figure 2c. This implies that the nanobelts are largely defect-free. The marked interplanar d-spacings of 0.63 and 0.33 nm correspond to the (0001) and (01-10) lattice planes of a WZ ZnS. The corresponding SAED pattern is depicted in Figure 2e. The SAED patterns recorded indicated WZ structures for all numerous ZnS nanobelts observed, independent of size and shape. Moreover, most of the belts grow in the [01-10] orientation, and present ( 2110) and (0001) planes as the top/bottom and side surfaces. Figure 2f illustrates a model of the nanobelts and their enclosing facets. The corresponding atomic model is displayed in Figure 2g; the latter reveals that the side surfaces are polar, and are terminated either with Zn or S. A nanobelt, first reported by the Zhong Lin Wang s Group in 2001, is a 1D structurally controlled nanomaterial that has welldefined chemical composition, crystallographic structure, and enclosing surfaces. [3,26] Although considerable attention has been Figure 2. a) Typical TEM image of ZnS nanobelts and b) the corresponding EDS spectrum acquired from one of the individual nanobelts in a). c) HRTEM image of a relatively large area of the nanobelt, previously shown in Figure 2a. d) Enlarged lattice-resolved HRTEM image taken from the marked area in c). e) Representative SAED pattern acquired from the nanobelt in c). f) Structural model of individual nanobelts and their enclosing facets. g) Atomic model of a WZ-type ZnS nanobelt. attracted to understanding their fundamental properties and prospective applications, the unambiguous interpretation of their growth mechanism is still missing. The growth kinetics and thermodynamics involved are extremely complex. As shown in Figure 2f and g, the growth of the present ZnS nanobelts occurs with a nonpolar surface dominating the crystallization, and the polar surfaces play a relatively little role in determining the final morphology. [22] Although the literature documents that most ZnS Adv. Mater. 2009, 21, ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 2035

3 nanobelts are grown along the [0001] orientation and have the (2-1-10) and (01-10) planes as enclosing facets, [17,21,22,27 29] the [01-10] axis is also an alternative growth direction when the synthesis conditions are changed or a vapor liquid solid (VLS) mechanism becomes effective. [30,31] The growth direction is sometimes dependent on belt width or morphology. For example, Ding et al. [32] have found that the ZnS nanobelts growing along the [01-10] directions normally have larger widths. In the present experiments, the Au film deposited on the substrate must play the key role for the growth along the [01-10] direction. As shown in the Supporting Information (Fig. S1) and Figure 1a, the ZnS nanobelts grew from some seeds on the substrate. These seeds offer nucleation centers for the nanobelt growth as well as restrict their further growth along a given growth direction, such as [01-10]. Cathodoluminescence (CL) is the emission of light as a result of electron bombardment. As compared to other techniques, such as photoluminescence (PL) spectroscopy, CL has the advantage of higher spatial resolution, up to few nanometers. [33] Using the same characterization means described in our previous work, [24] the optical properties of individual ZnS nanobelts dispersed on standard C-coated TEM copper grids were studied using highspatial-resolution SEM. Figure 3c displays a typical CL spectrum acquired from a pure ZnS nanobelt (seen in the SEM image in Fig. 3a). The spectrum consists of a narrow and strong UV peak centered on 337 nm and broad, low-intensity luminescence in the visible region (550 nm). Similar CL spectra with hardly noticeable intensity and peak-position differences were collected from many other individual ZnS nanobelts. Figure 3b is the corresponding CL image of the nanobelt showing fairly uniform luminescence that arises from the nanobelt alone. In order to study spatial variations of the optical properties, the CL spectra were collected in different spots along the individual ZnS nanobelts. Figure 3d plots a number of CL spectra acquired at various spots along the line shown in Figure 3a. The shapes of CL spectra are fairly uniform, in which all peaks are located at the same positions, while the intensities of the UV peaks are slightly decreased. Although much research has been devoted to the optical properties, such as CL and PL, of 1D ZnS nanostructures, very few studies have been reported on their possible UV band-gap emission at RT. This is mainly due to the high sensitivity of the 1D ZnS nanostructure optical properties to the synthetic conditions, crystal size and shape, and intrinsic defects, such as vacancies, interstitials, and others. As the literature documents, for a large number of 1D ZnS nanostructures of various shapes, there are usually two characteristic optical peaks visible at RT. [34 36] Thus, one can obtain the UV band-gap emission at RTonly if one learns the origins of visible-light peaks and then suppress them on purpose. Through the analysis of numerous experiments, we showed that the first visible-light peak is related to elemental sulfur species on the ZnS nanostructure surfaces. This was substantiated using structural analysis by HRTEM, X-ray photoelectron spectroscopy (XPS), PL and excitation techniques, etc. [37] The microstructure and crystallography within a given nanostructure could vary except in the case of a perfect singlecrystalline structure. Currently, we have been investigating the origin of the CL peaks and their relation to the microstructure within the same single nanostructure using high-resolution CL measurements and HRTEM. The details will be presented elsewhere. Briefly, this indicates that the second visible-light band emission originated mainly from the defects or twin planes. This inspired us to carry out a synthesis where no external sulfur source was used, and a majority of defect-free single-crystalline ZnS nanostructures were obtained by decreasing evaporation rates. Figure 4a shows an optical microscopy image of ZnS nanobeltbased UV sensors. Schematics of an individual ZnS nanobeltbased sensor and a representative SEM image of a singlecrystalline ZnS nanobelt device are shown in Figure 4b and c (obtained from the marked area in Fig. 4a), respectively. The Cr/Au parallel electrodes (4 mm apart) are visible in Figure. 4b. Figure 3. a) SEM image of a ZnS nanobelt and b) its corresponding CL image. c) CL spectrum recorded from the ZnS nanobelt seen in the SEM image in a). d) CL spectra acquired at various spots on the nanobelt along the line marked in a). Figure 4. a) Optical microscopy image of an individual ZnS-nanobeltbased UV-light sensor array composed of individual structures. b) Schematic diagram of the individual ZnS-nanobelt-based UV-light sensor configuration. c) SEM image of a single-crystalline ZnS-nanobelt device taken from the marked area in a). d) I V characteristics of an individual ZnS-nanobelt-based UV-light sensor illuminated with light of different wavelengths ranging from 320 to 600 nm ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2009, 21,

4 The nanobelt in Figure 4c has a typical width and thickness of 400 and 20 nm, respectively, with its uncovered part exposed to the incident light. The dark current in the device was below the detection limit (10 14 A) of the current meter. Figure 4d shows the typical current voltage (I V) characteristics of an individual ZnS nanobelt-based UV sensor illuminated with light of different wavelengths from 320 to 600 nm. The I V curve shows a nonlinear behavior that may be due to blocking at the metal/ semiconductor contacts. The I V curves acquired upon illumination with sub-band-gap light with a wavelength of 600 nm were highly resistant. [6] In contrast, upon the illumination with a photon energy above E g (3.7 ev, 335 nm), such as 320 nm UV light, the resistance decreased by several orders of magnitude, as depicted in Figure 4d. The UV-light sensitivity decreased markedly under vacuum conditions, indicating that surface trapping played an important role in the observed UV response. Electron hole pairs were generated under illumination with photon energy above E g, and one type of carrier was trapped at the surface, enhancing the sensitivity due to increasing density of the other carrier type. The spectral response at 320 nm is 0.12 AW 1 at a 5 V bias. This corresponds to an external quantum efficiency of around 50%, if the photoconductive gain is assumed to be 1. The efficiency can be much higher at higher electric fields due to carrier injection. The present performance is comparable to a near-uv-light photodetector based on hybrid polymer/zinc oxide nanorods prepared by low-temperature solution processes. The latter device exhibited a response of 0.18 AW 1 at 300 nm by applying a bias of 2V. [38] Figure 5a displays the spectral photoresponse of an individual ZnS nanobelt-based UV sensor for a bias of 10 V at different wavelengths from 300 to 630 nm. The highest sensitivity was observed at the wavelength of 300 nm. The photoresponse of an individual ZnS-nanobelt-based sensor increases by over three orders of magnitude upon illumination with a 320 nm light compared to its response to visible light. High spectral selectivity combined with high photosensitivity indicates bright prospects for use of the present ZnS nanobelts as visible-light-blind UVlight photodetectors in many areas. In particular, since ZnS materials are operative at much higher temperatures than conventional Si semiconductors, the present detectors can be effectively used during combustion monitoring of gases, during which the UV-light emission is a characteristic process. [1,39] Figure 5b shows the time response of an individual ZnSnanobelt-based UV sensor to a pulsed incident UV light (320 nm) measured by a current meter under and without UV light. This experiment highlights that the time response is faster than the limit of our measurement set-up (0.3 s). The photoconductivity response of an individual ZnS nanobelt-based UV sensor is fully reversible and periodic. The possible origins of the observed current fluctuations are surface species absorption/ desorption or appearance of defects. Figure 5. a) Spectral photoresponse measured from an individual ZnSnanobelt-based UV-light sensor for a bias of 10 V. b) Time response of the photoconductor upon 320 nm light illumination measured by a current meter under and without UV light. Figure 6. A multiple ZnS-nanobelts-based sensor and its performance. a,b) Optical images of the device at different magnifications. The nanobelts were grown on a prepatterned substrate produced using optical lithography and a predesigned mask (the process was similar to that used for the fabrication of the individual nanobelt-based UV sensors). c) Spectral photoresponse measured at a bias of 20 V within nm wavelengths. d) Conductance response of the device under 320 nm UV-light illumination for the light-on and light-off states. Adv. Mater. 2009, 21, ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 2037

5 A multiple-zns-nanobelt-based sensor and its performance are shown in Figure 6. Figure 6a and b depicts optical images of the devices at different magnifications. A Cr layer was used to protect the SiO 2 layer on the Si substrate. The nanobelts grew selectively on the top and side surfaces of the electrodes. The interspaces between electrodes were densely filled with wellaligned nanostructures. The nanobelt lengths were enough to join each two neighboring electrodes in one circuit. The nanobelt physical contact to electrodes was significantly improved with the addition of post-synthesis Cr/Au (10 nm/100 nm) layers. The present sensor photoresponse at a bias of 20 V and wavelengths is shown in Figure 6c. The photoresponse shows an increase above the threshold excitation energy, ~335 nm (~3.7 ev), which is close to the bandgap energy of ZnS. However, some of the sensor photoresponses have a turning point between ~240 nm and ~310 nm so that the photoresponse displays an increase between the threshold excitation energy and the turning point, and then a small decrease after the turning point. The origins of these minor changes seem to be interesting for further investigations. Figure 6d illustrates a conductance response of the device upon 320 nm light illumination measured for the light-on and light-off states. The stability and photocurrent intensity are significantly enhanced compared with an individual ZnSnanobelt-based UV-light sensor performance. In summary, we have demonstrated an effective approach for the synthesis of single-crystalline ZnS nanobelts possessing sharp UV-light emission at RT via right selection of source materials and by controlling their evaporation and agglomeration rates. Individual-ZnS- nanobelt-based UV-light and multiple- ZnS-nanobelts-based sensors were successfully fabricated, and showed high potential as visible-light-blind UV-light photodetectors and ultrafast optoelectronic switches. The sensor characteristics, including a spectral response, I V curves under various light illuminations, and time response were studied. The photoresponsivity of ZnS-nanobelt-based UV-light sensors exhibited over three orders of magnitude gain under UV-light illumination as compared to visible light. The high spectral selectivity combined with high photosensitivity and fast response time (<0.3 s) render the present single-crystalline ZnS nanobelts particularly valuable for new visible-light-blind UV-light photodetectors, especially in the UV-A region. Experimental The ZnS nanobelts were synthesized by chemical vapor deposition (CVD) using a simple conventional horizontal tube furnace with a 36 mm inner-diameter quartz tube at C. High-purity commercial ZnS powder (10 mm, 99.99%) was used as a precursor, and was put into an alumina boat that was placed in the center of the tube furnace. Silicon substrates coated with a Au layer 3 nm thick were prepared using an electric gun deposition system (ULVAC UEP C) and were placed at the downstream position of the source material. The tube furnace was purged with high-purity argon (Ar) at a constant flow rate of 500 sccm (standard cubic centimeter per minute) for 3 h prior to heating in order to remove traces of oxygen from the furnace. Then, the furnace was heated to C in 1 h and kept at this temperature for 30 min with high-purity Ar simultaneously serving as a protecting medium and carrying gas. The Ar flow rate was kept at 100 sccm during the heating and cooling processes. After the system was cooled down to RT, white-colored wool-like products were deposited onto a silicon substrate. The as-synthesized products were analyzed and characterized using powder X-ray diffraction (XRD, RINT 2200HF), field-emission SEM (JSM-6700F), and TEM (JEM-3000F) coupled with X-ray energy dispersive spectrometerry (EDS). Spatially resolved CL spectra from individual structures were collected with a high-resolution CL system at an accelerating voltage of 5 kv and a current of 1000 pa at RT using an ultrahigh-vacuum scanning electron microscope (UHV-SEM) with a Gemini electron gun (Omicron, Germany) equipped with a CL system [40]. For the fabrication of an individual nanobelt-based UV sensor, ZnS nanobelts were dispersed on a thermally oxidized Si substrate covered with a 300 nm SiO 2 layer. The Cr/Au (10 nm/100 nm) interdigitated electrodes were patterned on top of the nanobelts using optical lithography with the assistance of a predesigned mask and electron beam deposition. A predesigned patterned substrate produced using the same fabrication process and the analogous electrodes was also used to fabricate a sensor made of multiple ZnS nanobelts. In the latter case, the nanobelts were not predispersed on the substrate. Different-magnification optical images of the patterned substrate are shown in Figure S2 in the Supporting Information. The gap between electrodes could be tuned using different masks. During the guided ZnS growth on the prepatterned substrates, the synthetic conditions within a CVD system were the same as those during the initial synthesis of ZnS nanobelts (Fig. 1). After the CVD system was cooled down to RT, white-colored wool-like products were found deposited onto the substrate. Due to the partial damage of the patterns and usage of a Au catalytic layer during the growth, the Cr/Au (10 nm/100 nm) interdigitated electrodes were repeatedly electron-beam deposited using 50 and 100 mm Au wires as a mask to cover the aligned nanobelts. A small quantity of nanobelts was damaged during the second deposition of interdigitated electrodes. Fabrication of the perfectly aligned singlecrystalline 1D ZnS nanostructures and producing devices using them and ZnS surface functionalization seem to be one of the alternative promising routes. The I V characteristics of an individual ZnS nanobelt-based sensor and of a multiple ZnS-nanobelt-based sensor were measured in air using an Advantest picoammeter R8340A and a dc voltage source R6144. The spectral responses were recorded by measuring a dc current at different wavelengths from 230 to 630 nm using a xenon lamp (500 W). The time responses of sensors to light irradiation were measured by a current meter after shutting the UV light. The incident-light power was calibrated by a UVenhanced Si photodiode [19]. Acknowledgements This work was in part supported by the World Premier International Center for Materials Nanoarchitectonics (MANA) of the National Institute for Materials Science (NIMS), Tsukuba, Japan, and the Nanotechnology Network Project of the Ministry of Education, Culture, Sports, Science and Technology (HEXT), Japan. The authors are indebted to Dr. J. Yan, Dr. P. M. F. J. Costa, A. Futaesaku, Y. Uemura, Dr. S. G. Ri, Dr. S. Hara, H. Sugaya, Y. Misawa, N. Akihiko, and K. Tamura for their technical assistances and kind help. Supporting Information is available online from Wiley InterScience or from the author. Received: August 21, 2008 Revised: January 9, 2009 Published online: March 26, 2009 [1] a) Y. Huang, X. F. Duan, C. M. Lieber, Small 2005, 1, 142. b) C. Yang, C. J. Barrelet, F. Capasso, C. M. Lieber, Nano. Lett. 2006, 6, c) W. Lu, C. M. Lieber, Nat. Mater. 2007, 6, 841. [2] Y. G. Sun, H. H. Wang, Adv. Mater. 2008, 19, [3] Z. W. Pan, Z. R. Dai, Z. L. Wang, Science 2001, 291, [4] Y. N. Xia, P. D. Yang, Y. G. Sun, Y. Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim, H. Yan, Adv. Mater. 2003, 15, ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2009, 21,

6 [5] a) L. Li, X. S. Fang, G. H. Li, J. Mater. Sci. Technol. 2007, 23, 166. b) L. Li, N. Koshizaki, G. H. Li, J. Mater. Sci. Technol. 2008, 24, 550. [6] K. H. Liu, P. Gao, Z. Xu, X. D. Bai, E. G. Wang, Appl. Phys. Lett. 2008, 92, [7] Q. G. Li, R. M. Penner, Nano. Lett. 2005, 5, [8] J. S. Jie, W. J. Zhang, Y. Jiang, X. M. Meng, Y. Q. Li, S. T. Lee, Nano. Lett. 2006, 6, [9] C. Soci, A. Zhang, B. Xiang, S. A. Dayeh, D. P. R. Aplin, J. Park, X. Y. Bao, Y. H. Lo, D. Wang, Nano. Lett. 2007, 7, [10] a) S. L. Ji, C. H. Ye, J. Mater. Sci. Technol. 2008, 24, 457. b) X. S. Fang, Y. Bando, U. K. Gautam, C. H. Ye, D. Golberg, J. Mater. Chem. 2008, 18, 509. [11] a) Z. L. Wang, Adv. Mater. 2003, 15, 432. b) C. M. Lieber, Z. L. Wang, MRS Bull. 2007, 32, 99. [12] A. Kolmakov, Y. X. Zhang, G. S. Cheng, M. Moskovits, Adv. Mater. 2003, 15, 997. [13] C. N. LaFratta, D. R. Walt, Chem. Rev. 2008, 108, 614. [14] C. S. Lao, M. C. Park, Q. Kuang, Y. L. Deng, A. K. Sood, D. L. Polla, Z. L. Wang, J. Am. Chem. Soc. 2007, 129, [15] J. H. He, Y. Y. Zhang, J. Liu, D. Moore, G. Bao, Z. L. Wang, J. Phys. Chem. C. 2007, 111, [16] X. S. Fang, C. H. Ye, L. D. Zhang, Y. H. Wang, Y. C. Wu, Adv. Funct. Mater. 2005, 15, 63. [17] X. S. Fang, Y. Bando, G. Z. Shen, C. H. Ye, U. K. Gautam, P. M. F. J. Costa, C. Y. Zhi, C. C. Tang, D. Golberg, Adv. Mater. 2007, 19, [18] X. S. Fang, Y. Bando, D. Golberg, J. Mater. Sci. Technol. 2008, 24, 512. [19] M. Y. Liao, Y. Koide, J. Alvarez, Appl. Phys. Lett. 2005, 87, [20] Y. Y. Lin, C. W. Chen, W. C. Yen, W. F. Su, C. H. Ku, J. J. Wu, Appl. Phys. Lett. 2008, 92, [21] a) X. S. Fang, L. D. Zhang, J. Mater. Sci. Technol. 2006, 22, 721. b) X. S. Fang, U. K. Gautam, Y. Bando, D. Golberg, J. Mater. Sci. Technol. 2008, 24, 520. [22] D. Moore, Z. L. Wang, J. Mater. Chem. 2006, 16, [23] U. K. Gautam, X. S. Fang, Y. Bando, J. H. Zhan, D. Golberg, ACS Nano 2008, 2, [24] X. S. Fang, U. K. Gautam, Y. Bando, B. Dierre, T. Sekiguchi, D. Golberg, J. Phys. Chem. C 2008, 112, [25] W. Z. Wang, B. Q. Zeng, J. Yang, B. Poudel, J. Y. Huang, M. J. Naughton, Z. F. Ren, Adv. Mater. 2006, 18, [26] Z. L. Wang, Annu. Rev. Phys. Chem. 2004, 55, 159. [27] D. Moore, Y. Ding, Z. L. Wang, J. Am. Chem. Soc. 2004, 126, [28] X. S. Fang, Y. Bando, C. H. Ye, G. Z. Shen, D. Golberg, J. Phys. Chem. C. 2007, 111, [29] X. S. Fang, Y. Bando, C. H. Ye, D. Golberg, Chem. Commun. 2007, [30] D. Morre, C. Ronning, C. Ma, Z. L. Wang, Chem. Phys. Lett. 2004, 385, 8. [31] Q. Li, C. Wang, Appl. Phys. Lett. 2003, 83, 359. [32] Y. Ding, X. D. Wang, Z. L. Wang, Chem. Phys. Lett. 2004, 398, 32. [33] B. G. Yacobi, D. B. Holt, Cathodoluminescence Microscopy of Inorganic Solids, Plenum Press, New York [34] Y. W. Wang, L. D. Zhang, C. H. Liang, G. Z. Wang, X. S. Peng, Chem. Phys. Lett. 2002, 357, 314. [35] Y. C. Zhu, Y. Bando, D. F. Xue, Appl. Phys. Lett. 2003, 82, [36] L. Chai, J. Du, S. Xiong, H. Li, Y. Zhu, Y. T. Qian, J. Phys. Chem. C. 2007, 111, [37] a) C. H. Ye, X. S. Fang, G. H. Li, L. D. Zhang, Appl. Phys. Lett. 2004, 85, b) C. H. Ye, X. S. Fang, M. Wang, L. D. Zhang, J. Appl. Phys. 2006, 99, [38] Y. Y. Lin, C. W. Chen, W. C. Yen, W. F. Su, C. H. Ku, J. J. Wu, Appl. Phys. Lett. 2008, 92, [39] M. Razeghi, A. Rogalski, J. Appl. Phys. 1996, 79, [40] T. Sekiguchi, Mat. Res. Soc. Symp. Proc. 2000, 588, 75. Adv. Mater. 2009, 21, ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 2039