Fe-ZnO is a candidate material in the field of room temperature magnetic semiconductors and dilute magnetic semiconductors(DMS). However, to accomplish DMS, the localized magnetic phase has to be formed inside the semiconductor. Many studies have been reported over the past 15 years; however, it was difficult to produce but also to analyse even if it is done[1]. In this study we focus on these dual difficulties, but specifically on the analytical difficulty.
Fe-ZnO nanorods have been fabricated using colloidal synthesis by a wet chemistry method[2] which can control the length and diameter of the nanorod product. This method has been successfully applied to synthesis of Co-ZnO, but not yet in the case of Fe-ZnO. We first synthesized ZnO:Fe by boiling Fe-spearate, Zn-spearate and Na- oleate in a solution of 1-octadecene. After mixing Fe-spearate and Zn-spearate at the specified
ratio, various morphologies of nanorods were formed along the doping concentration during the synthesis reaction(Figure 2). In order to detect the Fe inside of ZnO, we utilized an aberration-corrected scanning transmission electron microscope (Jeol ARM 200CFG) combined with electron energy loss spectroscopy(Gatan inc). Magnetic characteristics of Fe-ZnO nanorods were evaluated with the SQUID technique.
Figures 2 and 3 present TEM micrographs, magnetic characteristics and STEM-EELS sepctra. Various aspect ratios of nanorods can be found through the TEM images. The length and diameter range of our nanorods are 50 ~ 100 nm and 10 ~ 20 nm, respectively. Magnetic characteristics of the 100-nm nanorod were measured and evidenced ferromagnetism. There is considerable Z-contrast changing along the long axis of the nanorod, which suggests rod composition variation, By Z-contrast mechanism there is more Zn atom in the brighter region and smaller in dark region. The intensity ratio of Fe L2,3 is a useful tool for evaluating chemical state. In the dark region, the increasing Fe L2 partial ratio from Fe L3 suggests that Fe oxidation state is also increasing[3]. On the other hand, Zn spectrum becomes shapeless in the dark region, probably due to the lattice distorsion. In order to estimate the exact chemical state of Fe, we applied "Hartree-Slater and modified double-step hydrogenic continuum models" to remove backgound from the Fe L23 edge spectrum. Furthermore, if Fe is substituted for Zn in the nanorod to form the magnetic phase, a lot of Zn vacancies can be formed, and HRSTEM reveals strong strain in the dark contrast region. It is necessary to calculate the defect formation energy and electronic structure variation in details.
[1] T. Dietl et al. Science 2000, 287, 1019
[2] Yang Y et al. Am Chem Soc. 2010, 132(38), 13882-94
[3] Schmid, H.K. et al. W. Micron 2006, 37, 426?432
This work was supported from KBSI project T34520.