科研之友
微信
新浪微博
Facebook
分享链接
摘要
A totally non-invasive procedure has been developed for differentiation of natural and synthetic ultramarine blue pigments on the basis of collection of UV-visible spectra in diffuse reflectance mode, followed by a chemometric treatment of the data using unsupervised pattern recognition methods. The main spectral features of natural and synthetic paint samples, i.e. reflection maxima, inflection points and reflection minima, could not be useful enough in the differentiation process; a threshold of 455 nm in the comparison of reflectance maxima has been observed, with synthetic samples peaking lower than this value and natural samples peaking higher, but it was not considered efficient in the differentiation, according to the fact that reflection maxima could be subjected to a bathochromic shift as a consequence of the addition of white pigments to blue paints. Chemometric analysis was therefore used in order to exploit information contained in the whole spectrum. To obtain an efficient classification, a proper data transformation was performed on the spectral data, using Z-score standardised variables. Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA) were the unsupervised pattern recognition methods used on the spectral data. Chemometric treatment was firstly applied to analysis of standard ultramarine blue paints and powder pigments and showed a good differentiation power, making it possible to distinguish between paints and raw lapis lazuli items and, more interestingly, between natural and synthetic ultramarine blue paints. Afterwards, PCA and HCA were applied to the analysis of blue paints on miniature painting artworks, again succeeding in the differentiation. This procedure could be used to develop a simple and totally non-invasive method for authenticating painted artworks.
- 出版日期2013
