X-Ray Fluorescence for Cultural Heritage: Scanning Biochemical Fingerprints in Archaeological Shipwrecks

VOL. 27 NO. 1 (2015) ARTICLE

X-ray fluorescence for cultural heritage: scanning biochemical fingerprints in archaeological shipwrecks

Yvonne Fors,a Håkan Grudd,b Anders Rindbyc and Lennart Bornmalmd aDepartment of Conservation, University of Gothenburg, Sweden bBolin Centre for Climate Research, Department of Physical Geography, Stockholm University, Sweden cCox Analytical Systems, Mölndal, Sweden dSciences and Department of Biology and Environmental Sciences, University of Gothenburg, Sweden

The challenges of archaeological wooden treasures recovered from the seabed Archaeological wood constitutes an important part of our cultural heritage, and the finds recovered from excava- tion sites may represent a broad range of different objects. This is particularly true for well-preserved shipwrecks, such as the famous Vasa (1628) (Figure 1) of Sweden and the British Mary Rose (1545). The wreck site context of such vessels offers a rich source of archaeological information, and the shipwrecks themselves usually constitute strong 1,2 historical symbols. Figure 1. The Swedish warship Vasa lost in 1628, salvaged in 1961. Photo: Victor Maschek/ Waterlogged archaeological wood is Shutterstock.com often found relatively well-preserved in reduced aquatic environments, or for example, Figure 2) and the develop- material that occurs at the seabed. 5 embedded in sediments or in bogs. In ment of sulfuric and organic acids in the The biological degradation in the form particular, the Baltic Sea is well-known wood. These might endanger not only of fungi and bacteria weaken the wood for the many shipwrecks preserved in its the delicate archaeological traces on the cells and is associated with accumula- low salinity water, which prevents inva- wood surface but also the stability of the tion of inorganic elements, which may sion by the infamous Teredo navalis construction over a long term.1,4 cause further preservation challenges (shipworm).3 However, the conservation in conserved wood on museum display procedures of these magnificent finds Biogeochemical and in storage.6 from the seabed usually consume a lot processes at the seabed Embedded in sediments at the seabed, of resources in terms of time, money and with concerns for the the waterlogged wood material is in many commitment. Moreover, the long-term preservation of the wood ways protected from mechanical and preservation of waterlogged wood and The main reason behind undertaking the biological erosion. However, the reduced especially that from shipwrecks can turn big task of conserving marine archaeo- anaerobic environment, which is generally out to be even more challenging, espe- logical wood and applying conservation considered to protect the timbers, actu- cially when facing the formation of acidic agents can be traced to the microbial ally initiates biochemical processes with sulfate salts on the wood surface (see, invasion of the waterlogged wooden consequences for the long-term pres- www.spectroscopyeurope.com SPECTROSCOPYEUROPE 11 VOL. 27 NO. 1 (2015) ARTICLE

Figure 2. Sulfuric acid and sulfate salt precipitates on the wood surfaces of objects from the Swedish warship Vasa. This is mainly the result of oxidis- ing iron sulfides (pyrite; FeS2) originally formed in the archaeological wood at the seabed. Photo: National Maritime Museums in Sweden. ervation of the conserved wood. 6,7 In different shipwrecks. The inhomogeneous Baltic Sea and the Mary Rose lost outside anaerobic environments, erosion bacte- sulfur and iron distribution profiles can Portsmouth in 1545.1,4 ria act as the primary wood degraders differ significantly even between neigh- The Vasa and the Kronan samples and consume cellulose in the wood cell bouring objects from the same wreck.4 show similar accumulation profiles, with walls. Associated with the erosion bacteria higher concentrations of sulfur and iron activity are the scavenging sulfate reduc- Mapping the accumulation close to the surface of the wood. Please ing bacteria. In a low-oxygen environment of inorganic elements in note that the Kronan sample analysed the sulfate reducing bacteria use seawa- archaeological wood with was cored completely through the wood ter sulfate as an electron acceptor when x-ray fluorescence to show two outermost surface ends, degrading the organic debris, and as a The distribution of sulfur and iron both of which indicated high sulfur result hydrogen sulfide (HS–) is formed. throughout the wood is of significant concentrations. The Vasa was only cored In the presence of iron ions from the interest since it may have a strong influ- to the centre of the hull. The wood seawater or from corroding iron objects at ence on both its conservation and its sample from the Mary Rose was bacteri- the wreck site, iron sulfides such as pyrite long-term preservation. This motivates a ally degraded in pockets throughout the

(FeS2) are formed in the wood matrix. In detailed study of the penetration profile full length of the core, accompanied by a secondary scenario, in the absence of of elements in wood samples from the sulfate reducing bacteria producing iron ions, organosulfur such as thiols (R– such archaeological objects of inter - hydrogen sulfide, which is present in iron SH) are formed in lignin-rich areas of the est. Scanning wood cores with X-ray from iron sulfides in the wood. wood.6 If the excavated wood is placed fluorescence (XRF) enables the non- These variations in degradation and in an environment with high and unsta- destructive mapping of the distribution accumulation patterns in the wood ble relative humidity, the iron sulfides are of selected inorganic elements at narrow samples from the shipwrecks can only easily oxidised to sulfuric acid at the wood steps (0.05 mm) throughout the wood be explained by differences in the local surfaces of the drying wood. This develop- sample. This offers a strong advantage to wreck site environment.2 A key question ment can be followed with the increasing the early applied elemental analyses of is, can the state of preservation in terms number of acidic sulfate salt precipitates.3,8 separate segments from the wood. The of biological degradation and inorganic The iron ions may themselves catalyse XRF-scanning technique has revealed a accumulation be predicted by improved oxidation and decomposition of cellulose strong correlation between the sulfur and knowledge about the wreck site envi- and hemicellulose in the wood, and may iron profiles, suggesting these elements ronment from basic geochemical data endanger the long-term stability of the are “chemically related” through the and XRF-scanning of other accumulated wooden object. There are also indications simultaneous accumulation by the elements in the wood? that the iron ions may have a degradation formation of iron sulfides in the wood.2 effect on the conservation agent polyeth- The accumulation profiles of sulfur Outlook ylene glycol.9 and iron have been mapped in wood Submerged wood provides ecosys- The accumulation of sulfur and iron samples from several different shipwrecks tems of high microbial diversity, which seems to be associated with bacte - of which three are presented here, see together result in a successive degrada- rial degradation of the wood.6 However, Figure 3. The Swedish Vasa (1628) from tion of wood and circulation of a number the accumulation profile of total sulfur Stockholm harbour, the Kronan wrecked of redox elements. The microbial activity and iron may differ between wood from in 1676 close to the island of Öland in the and therefore the degradation of water-

12 SPECTROSCOPYEUROPE www.spectroscopyeurope.com VOL. 27 NO. 1 (2015) ARTICLE

solid-state X-ray detector. The analyt- (a) ical information depth in the wood was about 1 mm. Quantitative calibra- tion was achieved by analysing stand- ard reference material 1547 from NIST (National Institute of Standards and Technology).

Acknowledgements The authors gratefully acknowledge the National Maritime Museums in Sweden, the Mary Rose Trust and Kalmar Läns Museum for providing the wood samples. (b) References 1. Y. Fors and M. Sandström, “Sulfur and iron in shipwrecks cause conservation concerns”, Chem. Soc. Rev. 35, 399 (2006). doi: http://dx.doi.org/10.1039/b507010b 2. Y. Fors and C. Björdal, in Interpretation of Shipwrecks, Ed by J. Adams and J. Rönnby. Södertörn Academic Studies 56, Southampton Archaeology Monographs New Series No. 4, The Highfield Press Southampton, Ch. 4, pp. 36–45 (2014). (c) 3. Y. Fors, Sulfur-Related Conservation Concerns for Marine Archaeological Wood. The Origin, Speciation and Distribution of Accumulated Sulfur with some Remedies for the Vasa. Doctoral thesis, Stockholm University (2008). 4. Y. Fors, M. Sandström, E. Damian Risberg, F. Jalilehvand and E. Philips, “Sulfur and iron analyses of marine archaeological wood in shipwrecks from the Baltic Sea and Scandinavian waters”, J. Archaeol. Sci. 39, 2521 (2012). doi: http://dx.doi. org/10.1016/j.jas.2012.03.006 Figure 3. XRF analysis from (a) the Kronan, (b) the Mary Rose and (c) the Vasa showing the 5. C. Björdal, T. Nilsson and G. Daniel, accumulation of total sulfur and iron in wood cores drilled from the shipwrecks. Please note “Microbial decay of waterlogged archaeological wood found in Sweden applicable to the increased concentration of the inorganic elements in the outer surface wood from the Vasa archaeology and conservation”, Int. Biodet. and the Kronan. The Mary Rose hull was bacterially degraded throughout the hull with subse- Biodeg.43, 63–71 (1999). doi: http://dx.doi. quent accumulation of sulfur and iron. Images reproduced by permission of The Royal Society of org/10.1016/S0964-8305(98)00070-5 Chemistry and by permission of Elsevier. 6. Y. Fors, T. Nilsson, E. Damian Risberg, M. Sandström and P. Torssander, “Sulfur accumulation in pinewood (Pinus sylvestris) logged wood appear strongly environ- ing sulfur and iron, were obtained by induced by bacteria in a simulated seabed 2 environment: Implications for marine archae- ment-adaptive. By mapping different means of an Itrax Multiscanner (Cox ological wood and fossil fuels”, Int. Biodet. elements in the wood and how these Analytical Systems, Mölndal, Sweden; Biodeg. 62, 336–347 (2008). doi: http:// correlate with sulfur and iron may provide http://coxsys.se) at the Department dx.doi.org/10.1016/j.ibiod.2007.11.008 7. Y. Fors, H. Grudd, A. Rindby, F. Jalilehvand, information about how the biodegrada- of Physical Geography, Stockholm M. Sandström, I. Cato and L. Bornmalm, tion of wood is regulated by symbiotic University. A conventional X-ray aggre- “Sulfur and iron accumulation in three biogeochemical interactions and circu- gate equipped with a special collimator marine-archaeological shipwrecks in the Baltic Sea: The Ghost, the Crown and the lation of redox elements under different utilising total reflection provided focused Sword”, Scientific Reports 4, 4222 (2014). wreck site conditions. This information Cr Kα X-rays (1.5420 Å) in a high inten- doi: http://dx.doi.org/10.1038/srep04222 enables prerequisites in predicting wood sity 2 × 0.05 mm beam. The element 8. M. Sandström, F. Jalilehvand, I. Persson, U. Gelius, P. Frank and I. Hall-Roth, preservation status and accumulation of specific X-ray fluorescence (XRF) was “Deterioration of the seventeenth-century inorganic contaminations in the wood. excited along a line from the outer to warship Vasa by internal formation of sulphu- the inner surface in the mid-section of ric acid”, Nature 415, 893–897 (2002). doi: http://dx.doi.org/10.1038/415893a XRF scanning the wood samples, see plots of Figures 9. G. Almkvist and I. Persson, “Degradation of Concentration profiles of a range of 3(a–c), recorded with 0.05 mm resolu- polyethylene glycol and hemicellulose in the elements in the wood samples, includ- tion by means of an energy dispersive Vasa”, Holzforschung 62, 64–70 (2008). www.spectroscopyeurope.com SPECTROSCOPYEUROPE 13