Archaeological wood degradation at the site of Biskupin (Poland): Wet chemical analysis and evaluation of specific Py-GC/MS profiles

Highlights

• We investigated 3000 year old oak wood from the Biskupin site.

• We collected complementary results by wet chemical analyses and Py(HMDS)-GC/MS.

• Depletion in carbohydrates was quantitatively assessed in degraded samples.

• Short side chain lignin Py products and cyclopentenones are indexes of wood decay.

• Abundance of lignin monomers and levoglucosan indicate good preservation of wood.

Abstract

Eight samples of ca. 3000 year old oak wood from the Biskupin site and a piece of sound oak (Quercus sp.) wood were analysed. The degradation state of archaeological oak wood was investigated using two analytical approaches: classical wet chemical analysis and analytical pyrolysis coupled with gas chromatography and mass spectrometry (Py-GC/MS) with in situ silylation. The results were compared with those obtained for sound oak wood.

Chemical analysis provided information on the amount of wood components. Their alteration at a molecular level was investigated by Py-GC/MS, highlighting how degradation can affect the formation of primary and secondary wood pyrolysis products.

The results showed that the chemical changes in the wood material in the eight samples examined had different entities and extents with comparison to sound oak wood. Samples taken from the external parts of the fragments had undergone a significant loss in polysaccharide components, whereas the internal parts were in a relatively good state of preservation. Evaluation of the relative amounts of pyrolysis products deriving from holocellulose and lignin highlighted that specific categories of compounds, such as lignin monomers and anhydrosugars, can be taken as an index for good preservation of wood components.

These results will be used to evaluate the effectiveness of the in situ conservation strategy by repeating the analyses on these samples after some years and comparing the results.

Introduction

In 1933 a fortified settlement dating back to the Bronze and early Iron Ages (around mid 8^th century BC) was found in Biskupin (Poland). The site was situated on a marshy island, occupying an area of about 2 ha. Several excavation campaigns were conducted between 1934 and 1974 [1] which highlighted that the village was abandoned after around 150 years of occupation, probably due to gradual flooding caused by violent climate change and the subsequent rising of the lake-water level [2]. The settlement was surrounded by a breakwater, consisting of several rows of oak and pine wood piles. Behind the breakwater, there was a fortified embankment, made of wooden boxes filled with earth and sand. The inner part of the settlement was composed of a total of 104–106 houses in thirteen rows. The street pavement also made of wood. At the end of the archaeological campaign in 1974, the findings included a considerable amount of oak- and pine-wood, as well as troughs made of lime-wood, a birch-wood ladle, an ash-wood disk wheel, a canoe hollowed out of a spruce tree trunk, and an alder-wood structural element from one of the houses. It has been estimated that about 8000 m³ of wood were used to build the settlement.

After excavation, some of the timber construction elements were exposed in the trenches for a long time, undergoing quick decomposition. Thus in the 1970s, an attempt was made to conserve the wood with phenol resin, leading to the total destruction of considerable amount of the archaeological wood [1].

After this attempt, it was decided to adopt an in situ conservation strategy by leaving the wood remains in the environment in which they had been found, either in the ground or water. Today Biskupin is an open-air museum, where it is possible to visit a reconstruction of the ancient village, whereas the archaeological wood is still kept underground.

It is known that waterlogged conditions are ideal for the long term preservation of archaeological wood. Important archaeological wood findings such as the Vasa ship [3], the Oseberg ship [4], the Roman ships of San Rossore [5], etc. were found underwater in a good state of preservation. Anaerobic conditions and low temperatures prevent the biodegradation of wood by white rot fungi, brown rot fungi and insects. However soft rot fungi and bacteria are still active in waterlogged conditions and can slowly degrade wood, resulting in alterations in its physical, chemical and mechanical properties [6], [7], [8]. In addition, since excavations started in Biskupin, and because of climate changes, there have also been changes in the burial environment, such as the lowering of the ground water and aeration of the soil. These phenomena could have contributed to accelerate the decay of the wood and it is therefore necessary to understand the dynamics of degradation.

Wood is a very complex material from a chemical point of view and its complexity is enhanced due to degradation [9]. Degradation has not a univocal meaning and the causes can be different (biological, chemical, mechanical, etc) [10], [11]. For instance, a wood can be considered degraded from a chemical point of view when chemical changes involving its components have occurred, but this does not necessarily imply a decrease in its mechanical properties [12]. This is why complementary approaches need to be used in order to assess the preservation state of wood from different points of view. Physical and mechanical properties are very important, because they are related to the usability and the general strength of wood [13]. Studying the wood morphology is also important, because it provides information on biological attacks and the structural integrity [7]. Finally, the assessment of the state of degradation from a chemical point of view provides information on the components of the wood at the molecular level and offers the possibility to understand the causes of decay, in order to establish preventive conservation conditions [14], [15], [16]. In this work degradation from a chemical point of view is considered with attention to the chemical changes undergone by archaeological wood components with respect to sound wood.

Many analytical techniques have been used to evaluate wood deterioration, starting with microscopic techniques, such as SEM and TEM, which are essential to investigate wood morphology [17]. Analytical methods have been applied to archaeological and historical wood [16], [18], [19], [20]. They are widely used in the pulp and paper industry, and are based on the determination of wood components by isolating and quantifying them using gravimetric techniques [21], [22].

To investigate the functionalities and the chemical bonds, various instrumental techniques have also been applied, such as NMR [23], [24], FT-IR [25], [26], [27] and thermal analysis [28]. Analytical pyrolysis is also a very powerful approach [29], [30], [31] coupled with GC, MS or GC/MS. Py-GC/MS enables complex macromolecules to be studied by observing smaller and simpler molecules [32]. It only requires very small amount of sample, and provides semi-quantitative results and information at a molecular level [31].

The pyrolysis of cellulose and hemicelluloses involves chain scission and water elimination as primary reactions, leading to the formation of anhydrosugars as the most abundant pyrolysis products. Secondary pyrolysis reactions involve further decomposition and the rearrangement of anhydrosugars, which produce smaller molecules, such as furans, pyrans and cyclopentenones [33], [34], [35]. During the pyrolysis of lignin, the formation of the monomers (coniferyl and sinapyl alcohols) is the first pyrolytic reaction, due to the predominant initial cleavage of the β-ether bonds between phenylpropane units. Reactions involving conversion/alteration of the side-chains and of the methoxy substituents on the aromatic ring are, on the other hand, secondary reactions, which lead to the formation of guaiacyl and syringyl units with shorter side chains and different functionalities [36], [37], [38], [39]. The use of in situ silylation enhances the detection of non-volatile compounds and protects alcoholic functionalities, particularly primary alcohol [40], which can easily undergo radical oxidation reactions. For example, the pyrolysis of lignin without derivatisation, produces coniferylaldehyde and sinapylaldehyde as major products [36], whereas the in situ derivatisation allows the TMS derivatives of coniferyl and sinapyl alcohols to be detected.

Some studies on Biskupin wood have already been carried out. An archaeological wood fragment of an oak trunk was investigated. The degree of degradation of sapwood and heartwood was determined on the basis of selected wood physical properties, of the concentration of major chemical constituents (TAPPI methods), and of microscopic observations, concluding that the wooden constructions deposited in the soil of the Biskupin site are generally in a good state of preservation [19]. In addition, within the framework of a monitoring program, results were obtained for sound oak wood deposited in the water of the lake and underground, and recovered after two years of deposition. The results showed that the wood had been colonized by aerobic and anaerobic bacterial and fungal microflora [41]. FT-IR was used to analyse pine and oak wood after 2, 4, 6 and 8 years of deposition [26]. Some selected water and soil parameters, such as groundwater level, pH and water conductivity, as well as redox potential and soil temperature, were measured, finding periodic fluctuations. Archaeological material situated in the layer of wet peat, where the samples investigated in this work were taken, lied in reducing conditions, which were proven not to favor degradation processes of wood [1].

In a previous work the different potentials of wet chemical analysis and Py(HMDS)-GC/MS analysis of archaeological wood were tested and compared [20]. In this paper we exploit the information given by the two approaches in a complementary way, using classical analysis to quantitatively determine the amounts of lignin and holocellulose, and Py(HMDS)-GC/MS to evaluate the chemical changes occurred at molecular level in the two components of archaeological oak wood from the Biskupin site. Wood pyrolysis products were grouped into categories. Changes in their relative abundances were evaluated and related to wood degradation, highlighting differences in the yields of primary and secondary pyrolysis reactions of archaeological wood. The analyses will be repeated after some years in order to evaluate the effectiveness of the in situ conservation strategy.