Evaluation of the effect of different paint cross section preparation methods on the performances of Fourier transformed infrared microscopy in total reflection mode

The stratigraphical characterization of polychrome samples may be considerably limited by sample preparation procedure both in terms of embedding material and surface polishing. Indeed, the use of synthetic embedding media may contaminate the sample, while the surface morphology obtained after polishing may affect the performances of different analytical techniques such as FTIR microscopy both in attenuated total reflection (ATR) and in total reflection mode. In particular, in ATR an optimal contact between the crystal and the sample is necessary, while in total reflection sample roughness affects the shape and the intensities of the spectra. Recently, different sample preparations were studied and compared taking into account the performance of FTIR microscopy in ATR mode but, as this approach presents some disadvantages, such as the risk of damaging the sample due to the pressure of the crystal, the present research work has been focused on total reflection, which works in non-contact mode. Two sample preparation methods based on the use of cyclododecane and KBr, recently proposed for their capability to reduce contamination effects, were selected to evaluate the performances of FTIR microscopy in total reflection mode on samples characterized by a different surface morphology. In particular the sample prepared with epoxy resin and CDD has been subjected to a simple cut, producing an inhomogeneous surface, with a high roughness, while the sample embedded with KBr has been prepared with a fine polishing procedure, obtaining an optically flat surface. Both of the preparation procedures were applied to obtain two paint cross-sections from fragments collected in the same area from a mural temple (15th century) located in Thubeken Lhakhang, Nepal. The cross sections were analysed with FTIR microscopy in total reflection mode, in order to compare the effect of the preparation methods on the spectral response. Line-scan imaging measurements were carried out on both of the samples and the resulting chemical images were re-constructed by a chemometric approach based on principal component analysis.

The stratigraphical characterization of polychrome samples may be considerably limited by sample preparation procedure both in terms of embedding material and surface polishing. Indeed, the use of synthetic embedding media may contaminate the sample, while the surface morphology obtained after polishing may affect the performances of different analytical techniques such as FTIR microscopy both in attenuated total reflection (ATR) and in total reflection mode.In particular, in ATR an optimal contact between the crystal and the sample is necessary, while in total reflection sample roughness affects the shape and the intensities of the spectra.Recently, different sample preparations were studied and compared taking into account the performance of FTIR microscopy in ATR mode but, as this approach presents some disadvantages, such as the risk of damaging the sample due to the pressure of the crystal, the present research work has been focused on total reflection, which works in non-contact mode. Two sample preparation methods based on the use of cyclododecane and KBr, recently proposed for their capability to reduce contamination effects, were selected to evaluate the performances of FTIR microscopy in total reflection mode on samples characterized by a different surface morphology. In particular the sample prepared with epoxy resin and CDD has been subjected to a simple cut, producing an inhomogeneous surface, with a high roughness, while the sample embedded with KBr has been prepared with a fine polishing procedure, obtaining an optically flat surface.Both of the preparation procedures were applied to obtain two paint cross-sections from fragments collected in the same area from a mural temple (15th century) located in Thubeken Lhakhang, Nepal. The cross sections were analysed with FTIR microscopy in total reflection mode, in order to compare the effect of the preparation methods on the spectral response. Line-scan imaging measurements were carried out on both of the samples and the resulting chemical images were re-constructed by a chemometric approach based on principal component analysis. (C) 2013 Elsevier B.V. All rights reserved.