H/Ge/Si(100)表面上的熱反應

Abstract

摘要 本文研究氫吸附在Ge/Si(100)表面的熱反應, 矽(100)樣品上鍺覆蓋層的厚度分別是0, 0.4, 0.8, 1.2單分子層。本實驗是利用由核心層光電子激發術進行觀測最上層的原子。由鍺3d核心電子層能譜的變化可知, 不論鍺的厚度是0或1.2個單分子層, 最上層氫部份是從H-Ge-Ge-H的單氫鍵結雙子上脫附, 另一部份是從H-Ge-Si-H複合的單氫鍵結雙子上脫附。因為Ge-H鍵的鍵結能比Si-H鍵要弱, 所以氫在較低溫時便從Ge-H鍵上脫附, 到較高溫時才從Si-H鍵上脫附。如上述, 從覆蓋了1.2個單分子層的表面上得到的鍺3d能譜亦可觀察到類似的脫附過程, 即氫分別從Ge-H及Si-H上脫附。類似的脫附過程顯示了Si-H 鍵結的出現, 但是樣品H/Ge/Si(100)未加熱前的表面已完全被鍺原子覆蓋了, 因此我們可以確定在氫脫附過程中表層附近的矽, 鍺原子之間發生了交換作用。因為交換作用都伴隨著氫的脫附, 所以在第一次氫脫附後露出的Ge-Ge雙子可能是造成交換作用的關鍵。在氫曝氣之後的矽2p能譜中, 有一個與表面狀態有關的分量出現。即使表面上覆蓋了1.2單分子層的鍺, 這個分量仍然存在, 這顯示此分量應該不是由表面下的矽原子所造成。從內體激發的光電子在穿過表面時可能會消耗部份能量以激發Ge-H鍵和Si-H鍵的震動。此現象和電子能量損失能譜 (electronic energy loss spectrum) 是同一原理, 因此我們認為這個分量可能是由於從內體激發的光電子消耗能量後所造成的。我們在本文中也會討論可行的實驗以驗證此想法。

ABSTRACT This thesis investigates the thermal reactions on the H/Ge/Si(100) surfaces by high-resolution core-level photoemission. The thickness of the Ge overlayers on the Si(100) sample are 0, 0.4, 0.8, and 1.2 monolayer (ML). The evolution of the Ge 3d spectra indicates that hydrogen atoms desorb partly from the H-Ge-Ge-H monohydride and partly from the H-Ge-Si-H mixed monohydride for all Ge coverages. Because the activation energy of the Ge-H dangling bond is weaker than that of the Si-H, hydrogen atoms desorb from Ge-H bonds at a temperature lower than Si-H. As indicated above, the similar desorption process of hydrogen, which indicates the existence of the Si-H bonds, is observed from the Ge 3d spectra of 1.2-ML Ge/Si surface. However, the corresponding Si 2p spectra show that the 1.2-ML Ge/Si surface is completely covered by Ge. Thus one can assure that the exchange reaction does occur between the Si and the Ge atoms near the surface layer accompanied with hydrogen desorption. Because the exchange reaction always occurs accompanied with hydrogen desorption, the Ge-Ge dimers after initial hydrogen desorption might be the precursor of the exchange reaction. A surface-related component appears in the Si 2p spectra after exposure of hydrogen on the sample. This component persists even for H-terminated1.2-ML Ge/Si(100) surface. Therefore one can rule out that the new component is originated from the subsurface. Instead of the origin from the subsurface, we suggest the shift of the new component should be due to an energy loss of the bulk component to the excitation of the Si-H and the Ge-H stretching mode. Possible experiments to confirm this suggestion are discussed.