Spectroscopic ellipsometry and computer simulation studies of thin film morphology evolution

Since thin film morphology is an important bridge in understanding the connections between film properties and preparatory conditions, this thesis presents studies on the evolution of thin film morphology as evaluated by spectroscopic ellipsometry (SE) and computer simulation techniques.; SE is used as a major tool to characterize the amorphous germanium films in this study. The basic principles of SE are briefly introduced, some systematic errors related to the SE measurement are discussed, and problems of local minima in the modeling of SE data are discussed. A method to improve the calibration and SE measurement on dielectric surfaces is proposed.; Amorphous germanium films deposited under different conditions are systematically examined by SE. Since the adatom mobility of Ge atoms is very low, films deposited under normal conditions are usually porous; this variable porosity depends sensitively on exact deposition conditions. This is recognized as due to the shadowing effects of the depositing atoms, and is enhanced by oblique angle deposition of the films. The results obtained on these films agree well with ballistic aggregation theory. Changes in film density as a function film thickness are also studied. A prism is used to study film density on both the top and bottom surfaces. The sign of the film density gradient may change with deposition conditions, especially with changes in bombardment conditions. In order to have a better understanding of the bombardment process, ion-assisted evaporation is employed to quantitatively study the effects of the ion bombardment during deposition. Film density changes as a function of bombardment dosage are clearly seen.; A geometric film growth model is developed in the thesis to describe the films grown under low adatom mobility condition, which has been recognized as having columnar or conical structures. One important feature of this growth process is clustering, clusters competing for growth, and evolution. Films grow from the clusters and the clusters compete with each other for growth. This results in the evolution of the observed film density and surface roughness. The simulation results have been compared with experimental results, and the trends in morphology evolution can be correctly predicted in the model.