|Budget Amount *help
¥6,000,000 (Direct Cost: ¥6,000,000)
Fiscal Year 1988: ¥2,000,000 (Direct Cost: ¥2,000,000)
Fiscal Year 1987: ¥2,000,000 (Direct Cost: ¥2,000,000)
Fiscal Year 1986: ¥2,000,000 (Direct Cost: ¥2,000,000)
As a means of evaluating semiconductor/insulator interfaces, crystal modeling method is proposed. this crystal model consists of plastic balls and sticks representing atoms and bonds, respectively. inputting the coordinates of atoms based on the crystal model into computer, energy relaxation is carried out calculating keating energy. In order to evaluate an interface, the interface model is first constructed, then the keating of the interface-forming atoms is calculated which in turn gives informations about stability of the interface.
As the result of the evaluation of low-index interfaces such as (100),(110) or (111), increasing energy was obtained in the order:(111), (110), (100). The stability of the interfaces decreased in the same order. Particularly, insulator/Si(100) substrate interface was evaluated in several ways, varying the shape of the interface. as the result, a pyramidal-shape interface surrounded by four (111) facets was found to be the most stable interface, which also
agrees with the result of XPS measurement. Therefore, actual insulator/semiconductor interface is considered to form this type of shape.
The interface was also evaluated by DV-Xa method, the energy of each orbital of interfaceforming atoms one of the molecular-orbital methods. Using this method, can be calculated and oxidation reaction at intefaces can be analyzed from electronic structure point of view. as the result, oxidation reaction was found to take place in the inner part of crystal instead of the interface itself.
We have also examined the SOI structure formed by L-SPE technique, in which semiconductor crystal is formed on top an insulator layer, giving totally different characteristics to the interface from those of the oxidation interface. Tt the growth front, three-phase boundary region, consisting of c-Si, a-Si and a-SiO_2, is formed. The formation of microtwins is explained in terms of the distortion energy of atoms at the growth front. Also the formation of microtwins at c-Si/SiO_2 interface is explained by the stability of (111) interface. Furthermore, delayed crystal growth in the three-phase region, consisting of c-Si, a-Si and a-SiN, can be attributed to comparatively large lattice mismatch between c-Si and a-SiN. This result is a good agreement with modeling evaluation.
As a conclusion, analysis based on modeling method is considered to be valid as far as its agreement with the experimental results is concerned. Less