シリカガラスの光学的性質に影響を与える構造欠陥

Abstract

Amorphous SiO2 such as bulk silica glasses and thin films have been one of key materials in modern optoelectronic industries. They are currently used in communication technologies as an optical fiber, fiber grating and fiber amplifier, thin films for electrical insulation in the dynamic random access memory (DRAM), and optical lenses and photo-masks in the excimer laser lithography. The property essential for the applications is its wide band gap amounting ∼9eV. However, bulk silica glasses commercially available and silica thin films do show photo-responses to sub-band gap lights like KrF emission and unexpected trapping of charges, and the behavior has strong dependency on the preparation histoty. A great number of studies were carried out to clarify the relation between the properties and structural imperfections in the materials and formation mechanisms of the defects. There are two categories of the imperfections: one is dopant- or impurity-related imperfections and another is non-stoichiometry related defects. These defects constitute gap states in a-SiO2. The structural identification was usually done by absorption- and emission-spectroscopy in visible-ultraviolet region and electron spin resonance (ESR). The experimentally proposed models were compared with the predictions by theoretical calculation of energy levels. Recent development of excimer laser lithography technique led us to recognition that a latent member, which has been unnoticed because of no response to the optical absorption or emission in the visible-uv range and ESR absorption, exists in the family of active centers in a-SiO2, that is a strained Si-O-Si bond originated from the planar three membered ring. In contrast, the puckered four membered ring is unstrained. Although it has been pointed out that there was a wide distribution in Si-O-Si bond angle from 90° to 180° by X-ray analysis or 29Si solid state NMR, the physical and chemical responses of the Si-O-Si bonds with a particular bond angle could not be differrentiated. Very recently it was clarified that a strained Si-O-Si bonds, in other words chemically excited bonds, has an optical absorption locating on the band edge. The excited bond can be scavenged by fluorine doping, because it is chemically reactive. In the present review we hope to show the unresolved optical and electric responses of silica glasses can be comprehensively understood by taking the presence of the strained bonds into consideration.