Effective Sample Size in Diffuse Reflectance Near-IR Spectrometry.

Two independent methods for determination of the effectively sampled mass per unit area are presented and compared. The first method combines directional-hemispherical transmittance and reflectance measurements. A three-flux approximation of the equation of radiative transfer is used, to separately determine the specific absorption and scattering coefficients of the powder material, which subsequently are used to determine the effective sample size. The second method uses a number of diffuse reflectance measurements on layers of controlled powder thickness in an empirical approach. The two methods are shown to agree well and thus confirm each other. From the determination of the effective sample size at each measured wavelength in the visible-NIR region for two different model powder materials, large differences was found, both between the two analyzed powders and between different wavelengths. As an example, the effective sample size ranges between 15 and 70 mg/cm(2) for microcrystalline cellulose and between 70 and 300 mg/cm(2) for film-coated pellets. However, the contribution to the spectral information obtained from a certain layer decreases rapidly with increasing distance from the powder surface. With both methods, the extent of contribution from various depths of a powder sample to the visible-NIR diffuse reflection signal is characterized. This information is valuable for validation of analytical applications of diffuse reflectance visible-NIR spectrometry.