well as pretreated reference samples. In the first step, the provenance of the reference samples based on the spectra from previously developed Raman database
was established. Then, the reference samples were subjected to the same conservation procedure that was used for the archeological beads and the newly applied
superficial layer was matured in the museum environment for 1 year. The analysis
of treated and untreated reference samples has shown that the applied lacquer does
not interfere with the Raman spectral bands used for amber identification. It was
then possible to determine the provenance of both sets of studied objects, beads
which underwent conservation treatment and reference samples. The recorded
Raman spectra suggested the use of succinite (Baltic amber) as a raw material in the
manufacturing process of the archeological beads.
Analyses of larger data sets collected with the Raman spectroscopy require new
approaches which can go beyond simple comparisons of the spectra. Successful
application of the chemometric methods in investigations of cultural heritage
objects has been documented in several recent papers [137–139]. In particular, the
Raman spectroscopy coupled with partial least squares (PLS) method was used for
the classification of amber samples according to their provenance and geological
age [135]. The Raman spectra of samples from Czech Republic, Baltic region, and
various objects dating back to the Upper Cretaceous and Cenozoic ages were
analyzed. In this work, an alternative chemometric approach was proposed to select
additional intervals on the spectra (Table 16.1). Several metrics (number of misclassifications and permutation test) which evaluated the capability of the alternative PLS models were used to predict future measurements. The achieved results
indicated the possibility to assess the geological age and provenance of amber
considering an interval of wave numbers instead of some single wave numbers.
Furthermore, differences in the Raman spectra were highlighted using the Variable
Importance Projection methodology over the PLS model. The applied model clearly
distinguished amber from Baltic, the Czech Republic and resins from Cenozoic and
Upper Cretaceous periods and supported classical results interpretations.
16.3 Vis
Color is one of the most important aspects of an artwork. Human eyes allow
resolving between approximately two million shades of colors even though our
vision capabilities are based only on three types of receptors sensitive to red, green,
and blue light. Analysis of the Vis spectra of an artifact, recorded with a spectral
resolution of many hundreds of points, could provide information on the chemical
nature of colorant used by an artist, their alteration caused by natural degradation
and/or human interventions. The analysis is usually done by comparing measured
spectra to a pigment/lake/dye database containing reflectance spectra collected for
different pure colorants dispersed in various binders (different oils, egg tempera,
varnishes). The spectra recorded in the reflectance mode contain also the scattered
light, where intensity and spectral characteristics depend on many factors (type of
500
T. Łojewski and B. Łydżba-Kopczyńska
was established. Then, the reference samples were subjected to the same conservation procedure that was used for the archeological beads and the newly applied
superficial layer was matured in the museum environment for 1 year. The analysis
of treated and untreated reference samples has shown that the applied lacquer does
not interfere with the Raman spectral bands used for amber identification. It was
then possible to determine the provenance of both sets of studied objects, beads
which underwent conservation treatment and reference samples. The recorded
Raman spectra suggested the use of succinite (Baltic amber) as a raw material in the
manufacturing process of the archeological beads.
Analyses of larger data sets collected with the Raman spectroscopy require new
approaches which can go beyond simple comparisons of the spectra. Successful
application of the chemometric methods in investigations of cultural heritage
objects has been documented in several recent papers [137–139]. In particular, the
Raman spectroscopy coupled with partial least squares (PLS) method was used for
the classification of amber samples according to their provenance and geological
age [135]. The Raman spectra of samples from Czech Republic, Baltic region, and
various objects dating back to the Upper Cretaceous and Cenozoic ages were
analyzed. In this work, an alternative chemometric approach was proposed to select
additional intervals on the spectra (Table 16.1). Several metrics (number of misclassifications and permutation test) which evaluated the capability of the alternative PLS models were used to predict future measurements. The achieved results
indicated the possibility to assess the geological age and provenance of amber
considering an interval of wave numbers instead of some single wave numbers.
Furthermore, differences in the Raman spectra were highlighted using the Variable
Importance Projection methodology over the PLS model. The applied model clearly
distinguished amber from Baltic, the Czech Republic and resins from Cenozoic and
Upper Cretaceous periods and supported classical results interpretations.
16.3 Vis
Color is one of the most important aspects of an artwork. Human eyes allow
resolving between approximately two million shades of colors even though our
vision capabilities are based only on three types of receptors sensitive to red, green,
and blue light. Analysis of the Vis spectra of an artifact, recorded with a spectral
resolution of many hundreds of points, could provide information on the chemical
nature of colorant used by an artist, their alteration caused by natural degradation
and/or human interventions. The analysis is usually done by comparing measured
spectra to a pigment/lake/dye database containing reflectance spectra collected for
different pure colorants dispersed in various binders (different oils, egg tempera,
varnishes). The spectra recorded in the reflectance mode contain also the scattered
light, where intensity and spectral characteristics depend on many factors (type of
500
T. Łojewski and B. Łydżba-Kopczyńska