As an example for an iron gall ink manuscript, a handwritten memo on a legal land description from the year 1769 was analyzed at ANKA in collaboration with the Institute for Instrumental Analysis of the Forschungszentrum Karlsruhe. Contactfree and non-destructive methods of investigation are obviously essential when handling such irreplaceable materials. Micro focused synchrotron X-ray fluorescence (µSXRF) measurements at ANKA fulfill these requirements, and at the same time allow detailed investigation of microscopic areas of the sample.
The µSXRF data show that a different ink was used in lines 10-11 of the manuscript than in lines 4-5, as indicated by the different zinc to iron and copper to iron ratios. This correlates well with infrared spectroscopy of the same manuscript, which also revealed local differences in the state of paper degradation. The results reveal the role of ink composition in accelerating the deterioration of historic manuscript paper, due to catalysis of cellulose oxidation by the zinc and copper contaminants in the ink. Such insights into the chemistry of the ageing process will provide a basis for developing effective conservation procedures.
||Figure 1: Original iron gall ink manuscript on legal land description of the year 1769; left: photo of the manuscript; right top: representative µSXRF spectra (21.5 keV excitation energy, 20 µm capillary, measuring time: 500 sec) of the iron gall ink and paper; right bottom: zinc to copper ratio of iron gall ink in different lines of the manuscript
The usefulness of FT-IR technique with synchrotron radiation in the characterization of materials in ancient paintings has already been demonstrated in a number of studies. SR allows focusing on areas of only a few microns, as a consequence of the high brightness and high collimation of the beam resulting in spectra with very good signal-noise ratio.
SR-FT-IR microspectroscopy has allowed us to isolate and identify the different compounds from very small samples. The painting layers are formed by a complex and heterogeneous mixture of substances (inorganic and organic pigments and binders, reaction compounds, alteration compounds) and were applied forming submillmetric layers (typical painting layers thickness between 20 to 100 μm).
Figure 2. Example of one of the spectra obtained with far IR microspectroscopy, spot size 50x50 μm2, 32 scans. The measurement corresponds to the minium Pb3O4 identified in a sample of wall painting from the 15th century from the Unha church (Val d’Aran).
Following the initial proposal, the samples analysed were paintings from the xv century.
Three types of materials were analysed:
1) Altarpieces belonging to the gothic period collection at “Museu Nacional d’Art de Catalunya” MNAC (24 samples).
2) Wall paintings from the Unha church in Val d’Aran (10 samples).
3) Reference materials (26 samples).
Small particles (few microns) extracted from the samples (<1mm2 size) are first placed and pressed in the diamond cell only one of the windows of the cell (the one in which the sample is well distributed) is used for the measurement. A stereomicroscopy is used to select the area of interest.
Mid IR is particularly interesting in the identification of the binding media in the painting layers, which is one of the most important aspects of the study. The small quantity of material and its heterogeneous nature hamper the effectiveness of conventional methods in determining the presence of certain substances, in particular, the binding medium. Mid IR allowes us to discriminate and identify the presence of oil and/or egg in the Gothic altarpieces. It is important to keep in mind that the presence of other materials pigments And impurities, for example lead basic carbonate, and the corresponding reaction products, lead carboxylates, result in a strong overlapping of the IR bands that difficult the detection of the binding media. It is precisely the high signal to noise ratio of the SR FTIR and small spot size used that result in an enhanced resolution of the spectra which is essential for the binding media identification.
They are quite an important amount of pigments such as oxides and sulphides, that do not show bands above 700cm-1, but that have bands in the far infrared region. Moreover, the identification of some of the substances in a mixture of compounds is heavily limited due to the strong overlapping of bands present. In the far-infrared region the interferences between the pigments compounds with the binding media are less significant. The use of SR-FT-IR microspectroscopy with a bolometer detector allowed us to reach the far-infrared at the same time that a small spot size is used (overcoming the diffraction limit) and thus expand the possibilities of this technique in the study of ancient paintings. We have to remark that the use of a small spot size is essential with these samples. The high intensity and high brilliance of synchrotron radiation in the far infrared is fundamental for these measurements. In this study pigments from 15th century paintings, such as, hematite, cinnabar, minium and lead tin oxide and which have no Mid IR bands have been identified. Moreover, the study of the variability of the composition of natural lapis lazuli used as pigment has also been carried out.
In parallel, the spectra from a set of reference compounds were also obtained with the same conditions. The reference compounds are selected to have a broad representation of the materials used in medieval painting.
Finally, some measurements were carried out in the region between 400-100 cm-1. In this case, a bigger spot size was found necessary and this limited the separation of the regions of interest. This would require a new sample preparation methodology. It is necessary to keep in mind that the quantity of material available strongly limits the sample preparation methods.