shows the XRD results. For the sample without In inclusion, it has two additional peaks not associated with ZnO, but with ZnO2 or Zn. When the percentage of In incorporated in the mixed powders increases, the intensity of the peak of ZnO (002) is enhanced, while the intensity of ZnO (101) peak is diminished. Moreover, other peaks which do not belong to ZnO disappear. This result clearly shows that the crystalline purity of zinc oxide can be improved with the inclusion of In. From those reported previously [12], we now attempt to understand the above result. When the velocity of crystallization of ZnO is too fast, the morphology of the crystal will be a hexagonal awl. However, when the velocity of crystallization is steady, the crystal of ZnO will be hexagonal rods or wires. Under this circumstance, (101) faces would not appear and most crystal faces are (002) or (100). Our SEM results shown above are indeed in good agreement with this description. Because there is no any additional peaks appearing in the XRD data, the fraction of the inclusion of In in the growth process is therefore negligibly small, and its role is mainly to serve as a surfactant located on the newly grown surface, which will decrease the growth rate of zinc oxide. Besides, it can prevent the incorporation of impurities into the grown crystal and improve the crystalline purity. To further confirm our conjecture here, Raman scattering experiments were performed as shown in Fig. 2(b). We can clearly see that the half-width of the E2 mode of ZnO decreases with increasing percentage of In inclusion and the peak intensity also increases. The result illustrates that the crystalline quality can be greatly improved with the inclusion of In in the growth process.