ARTICLE pubs.acs.org/ac
3D Micro-XRF for Cultural Heritage Objects: New Analysis Strategies for the Investigation of the Dead Sea Scrolls Ioanna Mantouvalou,*,† Timo Wolff,†,‡ Oliver Hahn,‡ Ira Rabin,‡ Lars L€uhl,†,‡ Marcel Pagels,† Wolfgang Malzer,† and Birgit Kanngiesser† † ‡
Institute for Optics and Atomic Physics, Technische Universit€at Berlin, Hardenbergstrasse 36, 10623 Berlin, Germany BAM Federal Institute for Materials Research and Testing, Unter den Eichen 87, 12205 Berlin, Germany ABSTRACT: A combination of 3D micro X-ray fluorescence spectroscopy (3D micro-XRF) and micro-XRF was utilized for the investigation of a small collection of highly heterogeneous, partly degraded Dead Sea Scroll parchment samples from known excavation sites. The quantitative combination of the two techniques proves to be suitable for the identification of reliable marker elements which may be used for classification and provenance studies. With 3D micro-XRF, the three-dimensional nature, i.e. the depth-resolved elemental composition as well as density variations, of the samples was investigated and bromine could be identified as a suitable marker element. It is shown through a comparison of quantitative and semiquantitative values for the bromine content derived using both techniques that, for elements which are homogeneously distributed in the sample matrix, quantification with micro-XRF using a one-layer model is feasible. Thus, the possibility for routine provenance studies using portable micro-XRF instrumentation on a vast amount of samples, even on site, is obtained through this work.
D
uring the past decades, X-ray techniques proved to be useful tools for the nondestructive investigation of cultural heritage objects.1,2 With the advent of new X-ray optics, elemental microanalysis became a proper analytical tool for microanalysis, in particular due to the technological development of polycapillary lenses.35 Micro X-ray fluorescence spectroscopy (3D microXRF) utilizes a focusing X-ray optic in the excitation channel in order to make lateral micro imaging possible.6,7 Lateral resolutions of less than 100 μm have become feasible even with X-ray tube excitation,8 thus facilitating routine investigations on a micrometer scale. Depth resolution can be achieved, when utilizing a second X-ray lens in the detection channel. The invention of threedimensional micro-X-ray fluorescence spectroscopy (3D microXRF)9,10 allows for the first time nondestructive, laterally and depth resolved measurements for the elemental analysis of a wide variety of samples. The technique is based on the use of two polycapillary optics in a confocal arrangement. The overlap of the foci of the two X-ray optics forms a probing volume from which fluorescence and scattered radiation is exclusively derived. Through the development of a full quantification for stratified samples, 3D micro-XRF with synchrotron radiation has in recent years turned into a true analytical tool.11,12 3D micro-XRF with X-ray tube excitation on the other hand can only be used qualitatively13 since quantification for polychromatic excitation is still a research topic. Quantification of experimental 3D micro-XRF data is a laborious task. The depth profiles must first be evaluated qualitatively in order to obtain initial values for the quantification r 2011 American Chemical Society
such as the number of layers, their boundaries and the concentrations of the fluorescence elements. The composition of the dark matrix must be obtained with the help of additional techniques. With this knowledge, net peak fluorescence intensities as functions of depth position are calculated and fitted to the measured data with a self-developed least-squares algorithm.11 This fitting must be judged during the iteration processes. Meaningful restriction parameters have to be found and different fitting strategies have to be applied to each new analytical problem. Thus, because of the laborious quantification and the need for synchrotron radiation, quantitative 3D micro-XRF measurements cannot be used for the routine analysis of a large amount of samples. On the other hand micro-XRF measurements are well suited for such kind of investigations. The nondestructive nature of X-ray analytical techniques makes them very important for the investigation of art and archeological objects. Such samples are often highly inhomogeneous; thus, lateral and depth resolution is desirable for the characterization. The lack of suitable reference samples necessitates referencefree quantification and in many cases the use of specific elemental concentrations or ratios thereof as fingerprint values. Additionally, it is common in the field of cultural heritage that there is a vast amount of samples which have to be analyzed to obtain reliable conclusions. That implies that routine analysis has to be possible, preferably on site. These aspects show that specific Received: May 2, 2011 Accepted: June 29, 2011 Published: June 29, 2011 6308
dx.doi.org/10.1021/ac2011262 | Anal. Chem. 2011, 83, 6308–6315
Analytical Chemistry quantification strategies are required for investigations in the archaeometric field. One very important art historian or archeological concern when dealing with a set of samples is the question of provenance. For this purpose, the samples are often analyzed and grouped according to the concentration values of a few marker elements. The most important prerequisite for these marker elements is that their concentration has not been altered during storage or restoration treatments and, thus, originates from the manufacturing process. Within the Qumran project coordinated by the BAM Federal Institute for Material Research and Testing, we investigated fragments of ancient parchment from the Dead Sea Scrolls as well as reference specimens from other archeological sites close to the western shore of the Dead Sea. The Dead Sea Scrolls are a collection of hand-written manuscripts found in caves along the western shore of the Dead Sea and are dated to the period between the end of the second century BCE and the first century CE. Fragments of more than 800 parchment, leather, and papyrus scrolls were discovered in diverse states of conservation ranging from excellent to very poor. Within the 60 years that passed after the discovery, the fragments have undergone further deterioration. Parchment is a writing material produced by drying a dehaired, wet animal skin under tension. Since the Middle Ages, the skins are dehaired with the help of liming whereas in antiquity, one employed allegedly only enzymatic methods to facilitate hair removal. Thus, ancient parchment should consist mainly of collagen, a triple-helix protein molecule composed primarily of glycine, proline, and hydroxyproline while medieval and modern parchment possesses a sizable CaCO3 fraction distributed evenly in the material. Typical densities of medieval parchment range between 1.4 and 2 g/cm3 as described in ref 14. For ancient parchment, no general indication was found. The question of archeological provenance and origin of the individual documents from the large collection of about 19 000 fragments has been one of the main objectives of the research concerning the Dead Sea Scrolls. On the basis of text analysis, scholars argue whether all, many or some scrolls were copied in Qumran.15 During parchment manufacturing the animal skin is immersed in water, so that different water fingerprints from different regions may be reflected in the parchment matrix composition. If this assumption is correct and changes in the composition due to storage or treatment can be excluded, provenance investigations are rendered feasible. Therefore, it has been suggested to seek for reliable elemental markers that would allow identification of the parchment that was manufactured locally. Water from the Dead Sea region has a very specific composition of salts, resulting from a rich history of geochemical processes starting in Oligocene times.16 Especially the content of the two major anions Cl (∼200 g/L) and Br (∼5 g/L)17 results in a Cl/Br ratio of 40 and less, due to possible precipitation of halite. In comparison, the Cl/Br ratio of seawater has a mean value of about 290.16,18 The principle goal of the 3D micro-XRF measurements at the synchrotron in this work was to render vast provenance studies possible using portable micro-XRF instrumentation. The threedimensional nature of the fragments was investigated with 3D micro-XRF in order to find marker elements, preferably Br and Cl. A homogeneous distribution of an element within a sample may indicate its presence rather due to the manufacturing process than to the influences of the environment or
ARTICLE
Figure 1. (a) Scheme of micro-XRF: Radiation is focused with a polycapillary lens onto a sample; emitted radiation is collected with an energy-dispersive detector. The use of the X-ray optic enables lateral microanalysis. (b) Scheme of 3D micro-XRF: the overlap of the foci of two polycapillary optics forms a probing volume from which information is exclusively collected. Three-dimensional elemental maps can be obtained.
postdiscovery intrusions. Such elements classify in this work as marker elements, and their concentration can be measured with an integral method. With the gained knowledge about the existence of such elements, routine investigations with micro-XRF can then be launched, even on site. A detailed interpretation of the microXRF results obtained during the investigations of the Dead Sea Scroll fragments is beyond the scope of this work and will be published elsewhere.19 There, the hypothesis that a chemical element which is homogeneously distributed may be a marker for the origin of the parchment is examined and archeological conclusions are drawn.
’ EXPERIMENTAL METHODS In this work we combined micro-XRF with X-ray tube excitation and 3D micro-XRF with synchrotron excitation for lateral and in-depth investigation of the parchment samples. In the following, the two techniques as well as the experimental details and quantification schemes will be briefly introduced. Micro-XRF. In micro-XRF spectroscopy, the incident X-ray radiation is focused by an X-ray optic, e.g., a polycapillary lens, onto a sample, and the characteristic fluorescence is collected with an energy-dispersive detector, see Figure 1 (left). In the case of thin samples such as parchment (approximately 300 μm thick), the incident radiation can excite atoms throughout the whole thickness of the sample. The intensity of the detected fluorescence radiation is dependent on its energy due to absorption in the sample and in air. For elements with low fluorescence energy such as Cl (KR radiation at 2.621 keV, information depth
