Growth and structure analysis of tungsten oxide nanorods using environmental TEM
© Tokunaga et al; licensee Springer. 2012
Received: 28 November 2011
Accepted: 25 January 2012
Published: 25 January 2012
WO3 nanorods targeted for applications in electric devices were grown from a tungsten wire heated in an oxygen atmosphere inside an environmental transmission electron microscope, which allowed the growth process to be observed to reveal the growth mechanism of the WO3 nanorods. The initial growth of the nanorods did not consist of tungsten oxide but rather crystal tungsten. The formed crystal tungsten nanorods were then oxidized, resulting in the formation of the tungsten oxide nanorods. Furthermore, it is expected that the nanorods grew through cracks in the natural surface oxide layer on the tungsten wire.
Keywordstungsten oxide nanorod environmental TEM
Metal oxides such as ZnO, In2O3, and WO3 are well known as bandgap semiconductors, which led to the development of many growth methods. During the studies into these growth methods, nanoscale metal oxides were discovered. These nanoscale materials have been widely studied since the electronic characteristics of nanoscale materials are different from those of bulk-scale materials [1–5]. In particular, metal oxides with nanorod structures were studied because they have a one-dimensional structure and are thus able to be applied for electrical components such as nanoscale wires. Tungsten oxide nanorods are one of the metal oxide semiconductors that can be easily made [6–8]. Therefore, due to its semiconducting properties, it is applied in electrical devices. However, the growth mechanism of tungsten oxide nanorods has not yet been clarified, and the growth of tungsten oxide nanorods has not been successfully controlled. In this study, the tungsten oxide nanorod growth process was observed using an environmental transmission electron microscope [TEM], and the growth mechanism was examined.
The growth of tungsten oxide nanorods was conducted by heating a tungsten wire in an oxygen atmosphere inside an environmental TEM. The commercially obtained pure tungsten wire (wire diameter, 25 μm; purity, 99.99%; The Nilaco Corporation, Tokyo, Japan) was used as the primary material for the tungsten oxide nanorods, and the heater, for the wire-heated environmental TEM sample holder, which enabled the introduction of gas into the environmental TEM. The holder was equipped with electrodes and the gas-introducing nozzle; the tungsten wire was connected between the electrodes and heated by current being applied to the wire. The measurement of the temperature of the heated wire was attempted using both a thermocouple and radiation thermometer. However, due to the small size of the wire, the thermocouple could not touch the wire. Furthermore, the measurement area of a radiation thermometer is larger than the wire diameter; therefore, space and material other than the wire was included in the measurement area. As a result, the wire temperature could not be measured by either the thermocouple or the radiation thermometer. Consequently, the wire temperature was measured using the following method. First, a pure metal powder with a known melting point was set on the connected wire. Secondly, the holder was introduced into the environmental TEM and the wire was heated. Then, the current was recorded when the metal powder melted; the same procedure was repeated with other metal powders. Finally, the temperature at which each metal melted was plotted on a current-temperature graph. This graph allowed us to determine the wire temperature without a thermocouple or radiation thermometer. The sample-heating holder was inserted in the environmental TEM, and the pressure in the environmental TEM was regulated by flow-rate control of the injected oxygen gas through the nozzle. The tungsten oxide nanorods grew after the current flowed through the tungsten wire. The environmental TEM used in the present study was made by HITACHI (H-9000NAR, Tokyo, Japan) and was equipped with a Gatan imaging filter [GIF] (Tokyo, Japan), a CCD, and a camera. This machine was operated at an accelerating voltage of 300 kV. The GIF was used to determine the elemental maps and electron energy loss spectra of the samples, and the dynamic growth behavior of the samples was recorded by the camera. The growth conditions used were as follows: the wire temperature was 800°C, and the oxygen pressure in the environmental TEM was 1.0 × 10-4 Pa. These growth conditions were applied for all the samples grown. Moreover, the existence and shape of the grown material on the wire were observed by scanning electron microscopy [SEM] (HITACHI, S-4300). The wire was removed from the TEM holder for the SEM observations. In thick crystalline tungsten, it is difficult to observe the natural surface oxide layer on the wire and the behavior of the interface between the nanorods and wire due to the difficulty of the transmission of electrons for TEM analyses. In this case, a part of the tungsten wire was fabricated into a thin film, in which electrons can transmit through, by a focused ion beam [FIB] (JEM-9320FIB, JEOL Ltd., Akishima, Tokyo, Japan). The FIB was operated at an accelerating voltage of 20 kV.
Results and discussion
The growth mechanism has often been discussed in other papers written about the growth of nanorod structures; the vapor-liquid-solid [VLS] and vapor-solid [VS] growth mechanisms are well known . The VLS growth mechanism is the method in which the vapor is melted in a catalyst and then segregated. The VS mechanism is the method in which the original sources are dissolved in vapor and then crystals formed on the substrate. Catalysts are needed for VLS growth, and there were no catalysts on the top of the nanorod in Figure 5a. Therefore, the growth mechanism of the WO3 nanorod was not VLS. Moreover, origin gases are needed in the case of the VS mechanism. In this study, the only origin gas was oxygen, and tungsten gas was not introduced in the environmental TEM. The possibility of the evaporation of tungsten oxide, which existed originally or formed by heating in oxygen on the wire, was imagined. However, the heating temperature was 800°C, which is lower than the required 1,400°C for the evaporation of tungsten oxide . As a result, the VS growth mechanism was not reasonable for the nanorod growth mechanism. In this study, the oxygen and tungsten originated from vapor and tungsten wire, respectively, so it is presumed that the tungsten, which was supplied from the tungsten wire, was oxidized by oxygen in vapor, and then WO3 nanorods were grown on the wire. These deliberations and results show that WO3 nanorods are grown from the tungsten prominence seen in Figure 4b by lateral growth.
Next, the growth of WO3 nanorods from the tungsten prominence is discussed. Engel et al. investigated the tungsten face that most easily absorbed oxygen and determined that the (110) face of tungsten absorbed the most oxygen . Figure 4 suggests that the side edge of the prominence was (110) of tungsten and oxygen absorbed preferentially on (110). It was also inferred that WO3 formed preferentially on (110) of the edge of the tungsten prominence, and then oxygen absorbed on the (010) face of the WO3 formed on the tungsten prominence. After that, the (010) and (001) faces of WO3, which absorbed oxygen easily and are the closest and close-packed planes , grew. The origin of the tungsten is the bottom of the tungsten wire, so this acts as the tungsten supply for the nanorods. Therefore, the growth of WO3 on the edge of the nanorod starts from the bottom to the top of the nanorod. The reason that WO3 nanorod growth disappears at the area covered by the natural oxide layer when the tungsten wire was heated is that the tungsten prominence, which has the planes that easily absorb oxygen, do not grow.
In summary, the mechanism of the WO3 nanorod growth was determined to be as follows: cracks occurred in the surface of the natural tungsten oxide layer when the tungsten was heated, after which tungsten diffused through the cracks of natural tungsten oxide layer from the tungsten wire to form a highly crystalline prominence. The (110) plane of the tungsten prominence was preferentially oxidized to form WO3. Tungsten and oxygen are supplied to the WO3 surface from the bottom tungsten wire and atmosphere, respectively, resulting in continual growth of the WO3 nanorods. To obtain further evidence for the proposed growth mechanism, a part of the oxide layer on the tungsten substrate needs to be fine-fabricated by FIB, electron beam lithography, etc., and then heated in an oxygen atmosphere, and the appearance of WO3 nanorod growth will have to be confirmed.
WO3 nanorods were grown by heating a tungsten wire in an oxygen atmosphere, and the growth of WO3 was observed by environmental TEM and HRTEM. In particular, the initial and the middle growth were observed. The growth mechanism involving the initial formation of cracks in the surface natural oxide layer on the tungsten wire followed by the formation of a tungsten prominence that was subsequently oxidized to form the WO3 nanorods was proposed. The tungsten and oxygen were supplied from the tungsten wire and the oxygen atmosphere, respectively. WO3 nanorod growth was suggested by TEM observation.
This work is supported by a research grant from the Murata Science Foundation and a Grant-in-Aid for Young Scientists (B program, no. 22760537) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.
- Ocana M, Morales MP, Serna C: The growth of α-Fe2O3ellipsoidal particles in solution. J Coll and Inter Sci 1995, 171: 85–91. 10.1006/jcis.1995.1153View ArticleGoogle Scholar
- Dai Y, Zhang Y, Li QK, Nan CW: Synthesis and optical properties of tetrapod-like zinc oxide nanorods. Chem Phys Lett 2002, 358: 83–86. 10.1016/S0009-2614(02)00582-1View ArticleGoogle Scholar
- Berger O, Fischer WJ, Melev V: Tungsten-oxide thin films as novel materials with high sensitivity and selectively to NO2, O3and H2S. Part I: preparation and microstructural characterization of the tungsten-oxide thin films. J Mater Sci: Mater in Electro 2004, 15: 463–482.Google Scholar
- Hiralal P, Unalan HE, Wijayantha KGU, Kursumovic A, Jefferson D, MacManus-Driscoll JL, Amaratunga GAJ: Growth and process conditions of aligned and patternable films of iron(III) oxide nanowires by thermal oxidation of iron. Nanotechnology 2008, 19: 455608–455614. 10.1088/0957-4484/19/45/455608View ArticleGoogle Scholar
- Li ZL, Liu F, Xu NS, Chen J, Deng SZ: A study of control growth of three-dimensional nanowire networks of tungsten oxides: from aligned nanowires through hybrid nanostructures to 3D networks. J Cry Growth 2010, 312: 520–526. 10.1016/j.jcrysgro.2009.11.036View ArticleGoogle Scholar
- Gu G, Zheng B, Han WQ, Roth S, Liu J: Tungsten oxide nanowires on tungsten substrates. Nano Lett 2002, 2: 849–851. 10.1021/nl025618gView ArticleGoogle Scholar
- Vaddiraju S, Chandrasekaran H, Sunkara MK: Vapor phase synthesis of tungsten nanowires. J Am Chem Soc 2003, 125: 10792–10793. 10.1021/ja035868eView ArticleGoogle Scholar
- Liu Z, Bando Y, Tang C: Synthesis of tungsten oxide nanowires. Chem Phys Lett 2003, 372: 179–182. 10.1016/S0009-2614(03)00397-XView ArticleGoogle Scholar
- Weast RC: CRC Handbook of Chemistry and Physics. Boca Raton: CRC Press; 1988.Google Scholar
- Lee M, Flom DG: Hardness of polycrystalline tungsten and molybdenum oxides at elevated temperatures. J Am Ceram Soc 1990, 7: 2117–2118.View ArticleGoogle Scholar
- Lee YH, Choi CH, Jang YT, Kim EK, Ju BK: Tungsten nanowires and their field electron emission properties. App Phys Lett 2002, 81: 745–747. 10.1063/1.1490625View ArticleGoogle Scholar
- Massalski TB, Murray JL, Bennett LH, Baker H: Binary Alloy Phase Diagram. Materials Park: American Society for Metals; 1986.Google Scholar
- Wagner RS, Ellis WC: Vapor-liquid-solid mechanism of single crystal growth. App Phys Lett 1964, 4: 89–90. 10.1063/1.1753975View ArticleGoogle Scholar
- Samsonov GV: The Oxide Handbook. New York: IFI/Plenum Data Corporation; 1973.View ArticleGoogle Scholar
- Engel T, Hagen TVD, Bauer E: Adsorption and desorption of oxygen on stepped tungsten surfaces. Surf Sci 1977, 62: 361–378. 10.1016/0039-6028(77)90088-7View ArticleGoogle Scholar
- Li YB, Bando Y, Golberg D, Kurashima K: WO3nanorods/nanobelts synthesized via physical vapor deposition process. Chem Phys Lett 2003, 367: 214–218. 10.1016/S0009-2614(02)01702-5View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.