Study Of Data Processing For Single-wave Null Ellipsometry

Ellipsometry is an optical analytical technique used to determine the optical properties and morphology of a surface from measurements of the change in polarization state of reflecting light. The method is widely used in many fields such as Physics, chemistry, materials and photographic science, biology, as well as biomedical engineering because of its essentially nondestructive character and remarkable sensitivity to minute interfacial effects. But the ellipsometer equation is an transcendental equation and there is no direct analytical solution for unknowns from the measured angles ?m and ?m. Then to find an algorithm to inverse the data ?m, and ?m and to determine the three unknowns is the primary problem. This dissertation presents a new method, Simulated-Annealing algorithm, to solve the problem, and gets good results after some tests.The main contents of the dissertation are: analyzing the complexity of the data processing for ellipsometry by simulation, discussing the influence of measurement error and searching precision on the result-hunting; discussing the process that how to eliminate the azimuth-angle errors resulting from polarizer, compensator, and analyzer by Four-zone averaging; introducing the Simulated-Annealing algorithm to the data processing , designing and writing a set of programs for data processing; optimizing the cool parameters in Simulated-Annealing algorithm by simulation calculation; processing the measured data for standard thin film which is provided by the manufacturer and "SiNx/glass" layers which are made by ourselves, and get the reliable results.