Studies in atomic-scale control of functional titanium nitride thin films by reflection high energy electron diffraction

氮化钛，TiN，具有共价、金属和离子性质，这使得它在微电路中的扩散障碍等技术应用和基础研究中都具有吸引力。为了研究TiN薄膜有趣的物理性质，原位观察TiN薄膜的生长是更可行的。尽管最近有报道称，氮离子注入到室温下在氯化钠衬底上蒸发的钛薄膜中可以外延生长TiN薄膜，但对氮化过程中TiN薄膜的晶体结构和电子结构的变化的了解还不够全面。因此，在400千伏分析和高分辨电子显微镜结合离子加速器中，将62 keV的氮离子(N_2~+)注入到蒸发钛薄膜中。采用电子能量损失谱和反射高能电子衍射的原位电子显微镜观察，结合离散变分X-α分子轨道计算，揭示了N注入引起的蒸发钛薄膜的晶体结构和电子结构的变化。A(001)取向的NaCl型TiN_y是由(03 5)取向的hcp-Ti转变为(001)取向的fcc-Ti和N在八面体(O-)位上占据而外延形成的，而(110)取向的TiN_y是通过氮化(110)取向的TiH_x形成的，H从TiH_x中优先释放，而不是N在fcc-Ti亚晶格的O位上占据。在氮离子注入初期，TiHx生长区域的体等离激元损耗峰向低损耗能量方向移动，而HCP-Ti生长区域的体积等离激元损耗峰随N剂量的增加而逐渐向高损耗能量方向移动。对Mulliken键重叠布居的分析表明，N在O位的占据导致了钛-钛键的减弱和钛-氮共价键的形成。

Titanium nitride, TiN, has covalent as well as metallic and ionic properties, which make it fascinating for both technological applications such as diffusion barriers in microcircuits, and fundamental research. In order to investigate the interesting physical properties of TiN films, in-situ observations of the growth of TiN films are more feasible. Although it has been recently reported in light of ex-situ experiments that TiN films could be epitaxially grown by N-implantation into Ti films evaporated on NaCl substrates at room temperature, knowledge of changes in the crystallographic and electronic structures during nitriding Ti films, is incomplete. Thus, nitrogen ions (N_2^+) with 62 keV were implanted into evaporated-Ti films in the 400 kV analytical and high resolution transmission electron microscopy (TEM) combined with ion accelerators. Observations by in-situ TEM equipped with electron energy loss spectroscopy and reflection high energy electron diffraction, along with the discrete variational Xα molecular orbital calculations, revealed changes in the crystallographic and electronic structures of evaporated-Ti films due to N-implantation. A (001)-oriented NaCl-type TiN_y is epitaxially formed by the transformation of (03 5)-oriented hcp-Ti to (001)-oriented fcc-Ti and by the occupation of N in the octahedral (O-) sites, whereas a (110)-oriented TiN_y is formed by nitriding a (110)-oriented TiH_x. The release of H from the TiH_x occurs preferentially rather than the occupation of N in the O-sites of fcc-Ti sublattice. The loss peak due to volume plasmon of areas where TiH_x has grown in the as-grown Ti film shifts to lower loss energy in the early N-implanting stage, while that of areas, where hcp-Ti has grown, gradually shifts to higher loss energies with increasing N dose. Analysis of Mulliken bond overlap populations determines that occupation of N in the O-sites gives rise to weakening Ti-Ti bonds and formation of Ti-N covalent bonds.