Abstract
A self-bracketing search technique employing classical dispersion analysis is used to calculate dispersion parameter values by minimizing the deviation between experimental and calculated reflectance or transmission data. An evaluation of the method on multiband spectra is accomplished by using synthetic spectra that contain both overlapping and nonoverlapping band systems. The experimental spectra are obtained by measuring the reflectance at normal incidence from a large single crystal or by measuring the transmittance of a thin film deposited on a transparent substrate. Upon application of the technique to real spectra, the deviation between the calculated and real spectra produces a constant value rather than approaching zero, as is the case with synthetic spectra. Reasons for this phenomenon are discussed.
Get full access to this article
View all access options for this article.