It is impossible to harmonize a real sample and routine analytical protocols, mostly due to small size and inner heterogeneity of the samples. Actually the microsample heterogeneity bears the most valuable information and must hence be retained. The detection limit, accuracy, correctness, and interferences are usually known in standard analytical techniques, but these features are usually unknown in analysis of microsamples. Mere availability of the modern instrumental device is then not sufficient, simultaneously, the analyst should also develop a method and test it for a given purpose, and this validation is not a trivial task. Additionally, it is never possible to use a single technique. In microanalysis of colour layer, there is no single and ready-to-use solution. The colour layer is a mixture of pigments and binders, and the identification of the organic and inorganic components of paintings requires at least two fundamentally different analytical techniques.
IR is a traditional analytical technique which is used for art work analysis in combination with light microscope (infrared microspectroscopy or microscopy). For the analysis of colour layers with a complex stratigraphy, it is necessary to prepare a cross section. For that the sample must first be fixed in an embedding resin, then cut and finally polished in a direction perpendicular to the layers. This sample preparation common in microscopy can worsen or even exclude IR measurement because of specular reflections. ALMA tests either the measurement of free fragments or thin (microtome) slices. The slicing can overcome the problem of the spectra acquisition. One more problem of IR spectroscopy is a qualified spectra interpretation, i.e. by examination of artificial samples approaching the real samples as much as possible. We found that even if properly measured, IR spectra can only bring information about the category of the binder but not exactly identify its kind. We are going to further define the applicability of IR, liberate it from non-realistic expectations, and change it to a reliable tool.
XRD is an indispensable and very reliable tool for the identification of mineral components of colour layer, i.e. inorganic pigments, binders, grounds, as well as secondary phases such as salt efflorescences or crusts and corrosion products. It is complementary to SEM/EDS, but in many cases XRD sees even more than SEM/EDS. Contrarily to Raman spectroscopy, XRD uses one universal and reliable database of crystal structures which makes the evaluation easier. The main direction of the development of the XRD analysis in ALMA is to decrease the amount of material to be analyzed ( < 1 mg) and sample size ( < 1 mm including cross sections) that is allowed due to specially equipped laboratory diffractometer X'Pert PRO with the primary beam diameter of 0.1 mm. In the development of XRD methodology for conservation science our laboratory cooperates with PANAnalytical Company, Almelo, Netherlands.
The current trend in many laboratories dealing with art work analysis is to use portable analytical instruments eliminating the necessity of taking samples. In 2005 ALMA co-operated with MOLAB mobile laboratory (Italy) and Moravian Gallery in Brno within an Eu-ARTech program project. .One of its aims was to compare the results of non-invasive methods with standard laboratory procedures. The results were very interesting. Non-invasive portable analytical methods can not fully replace materials examination of microsamples and the description of colour layers´ stratigraphy (e.g. technical construction of art work) but they can provide a very quick objective approximate information on colour layer composition and locate repaints without the need to invade the art work. Thus they can appropriately supplement imaging techniques and help make taking samples more effective. The quality of analytical data obtained by non-invasive methods is roughly comparable to microanalytical laboratory measurements (e.g. mobile XRF to SEM/EDX, mobile IR spectroscopy to IR microscopy). The ALMA Laboratory currently uses X-MET 3000 TXR equipment (Oxford instruments Company).