FIM Observation of Dynamic Processes of Thin Film Growth by Vapor deposition.

利用以原子分辨率著称的场离子显微镜（FIM）观察气相沉积金属薄膜的精细结构，研究了薄膜的生长。除了使用场蒸发观察内部结构的传统方法（方法1）外，我们还尝试通过一种新方法（方法2）直接观察生长表面，该方法是在将金属气相沉积到尖端时进行FIM观察，交替打开和关闭成像电压。方法1对钨上铜沉积的研究，与常识相反，表明即使尖端温度很低（55K），铜原子也能够从衬底表面的第一个触点移动约7-9nm。此外，在高温尖端（200-400K）沉积铜的情况下，发现撞击原子沿撞击原子的动量方向在衬底表面上运动。在覆铜针尖的热处理下，再结晶的成核也发生在铜镀层和钨基体的界面上。结果发现，铜与铁和铬的结果存在一些差异。在铜的情况下，没有得到单晶像，但在钨的低折射率面上观察到几个同心圆阶环。在铁和铬的情况下，观察到相反的单晶图像。沉积薄膜面积的扩展也略有不同。新方法2有几个基本问题，其中最重要的是成像电压对生长本身的影响。尽管有这些麻烦，一些有趣的结果已经开始出现，例如，原子在低折射率平面上的固定稍微困难，但在钨的阶环上的固定很容易，等等。更详细的结果即将公布。

Thin film growth has been studied by observing fine structures of vapor-deposited metal thin films with a field ion microscope (FIM) well-known with its atmic resolution. In addition to the traditional method (Method 1) to observe the inside structure using field evaporation, we have attempted to directly observe the growing surface by a new method (Method 2) in which the FIM observation is performed during vapor-depositing a metal onto a tip with alternating the imaging voltage on and off. The study with Method 1 on copper deposition on tungsten, contrary to common sense, showed that copper atoms were able to move approximately 7-9nm on the substrate surface from the first hit point even though the tip temperature was very low at 55K.In addition, in the case of deposition of copper onto a high temperature tip (200-400K), the impinged atoms are found to move on the substrate surface in the direction of the momentum of the impinging atoms. The nucleation for the recrystallization under the heat treatment of a copper-covered tip was also revealed to take place at the interface of the copper deposit and the tungsten substrate. Some differences were found between the result with copper and those with iron and chromium. In the case of copper, a single crystal image was not obtained but several concentric step rings were observed on the low index planes of tungsten. On the contrarya single crystal image was observed in the case of iron and chromium. The spreading of the deposited thin film area was also slightly different to each other. With new Method 2, several fundamental problems have come to loght, most important one being the influence of the imaging voltage on the growth itself. Despite of these troubles, some interesting results have begun to come out, e.g., it is slightly hard for atoms to fix on the low index planes but easy to fix at the step rings of tungsten, and so on. More detailed results are being expected now.