Analysis of pigments and complex pigmented media using non-destructive analytical methodologies and chemometrics

The use of non-destructive analytical methodologies in conjunction with chemometrics was evaluated for complex sample compositions containing pigments. Samples were analyzed using energy dispersive X-ray fluorescence (EDXRF), attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FT-IR), and Raman microscopy.; Spectral preprocessing was used to eliminate error among the spectral data. Spectra underwent normalization, 2nd derivative calculations and mean centering. Principal component analysis (PCA) allowed for the enhancement of similarities and differences among the preprocessed data sets.; The first sample set contained inorganic pigments. Red, blue, green, brown, and orange 19th century inorganic pigments present on the American bank checks were positively identified using this method. Powdered titanium composite pigments were also identified. These pigments were characterized and classified by components present, TiO2 polymorph and extender pigment, and by manufacturing method. The earliest possible date of manufacture was determined for these powdered pigments. Unknown actual house paint samples containing titanium composite pigments were dated through characterization by methods of manufacture and components present.; The second sample set analyzed by this method was nail lacquer enamels. Nail lacquer enamels from nine common manufacturers were characterized and further interpreted for patterns among the samples by PCA. The preprocessed sample spectra were used to form a database of 2nd derivative infrared and Raman spectra. Unknown samples whose composition was present in this database were positively identified. Unknown samples with compositions not present in the database were associated with their manufacturer and brand through cluster and separation interpretation in PCA scores plots. Weight percents and ratios of nail lacquer enamel main components, dibutyl phthalate and nitrocellulose, were used to limit the size of the sample pool resulting in greater separation and clustering.; The method developed, combining non-destructive analytical techniques with spectral preprocessing, database formation, and PCA, can be applied to many sample sets. This method is beneficial for any sample set composed of extremely similar complex components and therefore spectral data, requiring separation or classification into smaller sample groups and ultimately identification.