J. Ling et al. /Journal of Solid State Chemistry 178(2005)819-824 Xp nt of Cu/Zn= 6/4 and the experiment of brass 41.45 wt% of Zn)(as shown in Fig. 2e and Fig. 3b) though these two kinds of reactants have approximately the same proportion of Zn. We believe the difference of morphology of products is caused by the difference of alloy melting degree. Brass has higher alloy melting degree than the mixed powders of Cu/Zn and this leads to a difference of control in the releasing of metal gas and then the difference in the morphology of final products. Further detailed investigation is needed In addition, for the cases of Zn mixed with a large of diagram it is much difficult to form a liquid Cu-Zn alloy at this experimental temperature. We only collected 2m33 some radiation-flower-like microwires with a large diameter (Fig. 2f). This morphology often exists in the thesized in the 1 catalyst via a CVD method, which may be attributed to a catalyst-assisted VLS growth [101 We also collected some bubble-like compact crusts at the alumina boat after the reaction when Cu-Zn alloy was used as reactant. The compositional and morpho logical studies for the byproducts can help us to further understand the proposed CGVA method. XRD analysis indicates that the byproducts mainly consist of ZnO and little zn. Moreover. SEM examination shows a three- tiered structure of the compact oxide crust(Fig. 4a). The exterior of the compact crust was covered by a large amount of white Zno micropods with a uniform umm diameter about lum(Fig. 4b), possibly caused by the oxidation of little Zn vapor at the surface of brass. Fig 4c shows the morphology appearing in the middle layer f the compact crust, which consists of large Zno microcrystal particles with a regular shape. We believe that the growth of Zno nuclei from the oxidation of a liquid Zn layer at high temperature may result in these Zno microcrystals. Much Zn gas aggregates inside the 1500 compact oxide crust and forms the fungus-like mor- phology(Fig. 4d). Based on above experimental results it is reasonable to suppose an enrichment of Zn vapor yss % ahead So 5 o 6s the brass as reactan, we need to break the bubble-like content at the surface of brass and a slow releasing of Zn crust, which obstructs the supply of oxygen, appearing Fig3.(a)SEM. (b)TEM micrograph and(c)typical XRD pattern for at the surface at initial reaction stage to meet with the he as-synthesized product of Zno nanowires(A: peaks of purity of latter growth of ZnO. However, in the latter reaction, zinc, B: peaks of purity of zinc nitride). the presence of Cu seems helpful to cease the formation of the crust According to the TEM observation(Fig. 5), we find ome un-developed ZnO multiply twinned structure [ll] realizes the control of Zn vapor pressure and then which explicitly shows that the following growth strongly confines the growth of the finally synthesized process: each grain within the Zno multiply twin Zno with various sizes and shapes by appropriately particles develops to a thick branch(Fig 5a)at the adjusting vaporization ratio during the reactions. step and then forms three thin nanowires(Fig. 5b) by Furthermore, experiments results show that there is a the incorporation of atoms at a whisker side surface and ig difference of morphology of products of the the diffusion of atoms along the lateral surface confinedvapor generation. However, here the presence of Cu realizes the control of Zn vapor pressure and then strongly confines the growth of the finally synthesized ZnO with various sizes and shapes by appropriately adjusting vaporization ratio during the reactions. Furthermore, experiments results showthat there is a big difference of morphology of products of the experiment of Cu/Zn ¼ 6/4 and the experiment of brass (41.45 wt% of Zn) (as shown in Fig. 2e and Fig. 3b), though these two kinds of reactants have approximately the same proportion of Zn. We believe the difference of morphology of products is caused by the difference of alloy melting degree. Brass has higher alloy melting degree than the mixed powders of Cu/Zn and this leads to a difference of control in the releasing of metal gas and then the difference in the morphology of final products. Further detailed investigation is needed. In addition, for the cases of Zn mixed with a large proportion of Cu, according to Cu–Zn binary phase diagram it is much difficult to form a liquid Cu-Zn alloy at this experimental temperature. We only collected some radiation-flower-like microwires with a large diameter (Fig. 2f). This morphology often exists in the microwires or nanowires synthesized in the presence of catalyst via a CVD method, which may be attributed to a catalyst-assisted VLS growth [10]. We also collected some bubble-like compact crusts at the alumina boat after the reaction when Cu–Zn alloy was used as reactant. The compositional and morpho￾logical studies for the byproducts can help us to further understand the proposed CGVA method. XRD analysis indicates that the byproducts mainly consist of ZnO and little Zn. Moreover, SEM examination shows a three￾tiered structure of the compact oxide crust (Fig. 4a). The exterior of the compact crust was covered by a large amount of white ZnO micropods with a uniform diameter about 1um (Fig. 4b), possibly caused by the oxidation of little Zn vapor at the surface of brass. Fig. 4c shows the morphology appearing in the middle layer of the compact crust, which consists of large ZnO microcrystal particles with a regular shape. We believe that the growth of ZnO nuclei from the oxidation of a liquid Zn layer at high temperature may result in these ZnO microcrystals. Much Zn gas aggregates inside the compact oxide crust and forms the fungus-like mor￾phology (Fig. 4d). Based on above experimental results, it is reasonable to suppose an enrichment of Zn vapor content at the surface of brass and a slowreleasing of Zn vapor at the reaction temperature. Actually, when using the brass as reactant, we need to break the bubble-like crust, which obstructs the supply of oxygen, appearing at the surface at initial reaction stage to meet with the latter growth of ZnO. However, in the latter reaction, the presence of Cu seems helpful to cease the formation of the crust. According to the TEM observation (Fig. 5), we find some un-developed ZnO multiply twinned structure [11], which explicitly shows that the following growth process: each grain within the ZnO multiply twinned particles develops to a thick branch (Fig. 5a) at the first step and then forms three thin nanowires (Fig. 5b) by the incorporation of atoms at a whisker side surface and the diffusion of atoms along the lateral surface confined ARTICLE IN PRESS Fig. 3. (a) SEM, (b) TEM micrograph and (c) typical XRD pattern for the as-synthesized product of ZnO nanowires (A: peaks of purity of zinc, B: peaks of purity of zinc nitride). 822 J. Ling et al. / Journal of Solid State Chemistry 178 (2005) 819–824