Nondestructive discrimination of biological materials by near-infrared Fourier transform Raman spectroscopy and chemometrics: Discrimination among hard and soft ivories of African elephants and mammoth tusks and prediction of specific gravity of the ivories

This paper demonstrates the usefulness of near-infrared (NIR) Fourier transform (FT) Raman spectroscopy and chemometrics in nondestructive discrimination of biological materials. The discrimination among three kinds of materials-hard ivories, soft ivories, and mammoth tusks-has been investigated as an example. NIR (1064-nm) excited FT-Raman spectra were measured in situ for these materials, and principal component analysis (PCA) of the obtained spectra was carried out over the 1800-400-cm(-1) region. The two kinds of ivories are clearly discriminated from one another on the basis of a one-factor plot. It was found that treatment of the Raman data by multiplicative scatter correction (MSC) greatly improves the ability to discriminate. Principal component weight loadings show that the discrimination relies upon the ratio of collagen and hydroxyapatite included in two kinds of ivories. The discrimination among the hard and soft ivories and mammoth tasks was made by a three-factor plot for FT-Raman spectra after the MSC treatments. Partial least-squares regression (PLSR) enabled us to make a calibration model which predicts the specific gravity of the hard and soft ivories.