- Email: [email protected]
New materials used for the consolidation of archaeological wood–past attempts, present struggles, and future requirements
Journal of Cultural Heritage 13S (2012) S183–S190
Available online at
www.sciencedirect.com
New materials used for the consolidation of archaeological wood–past attempts, present struggles, and future requirements Mikkel Christensen a,b,1 , Hartmut Kutzke a,∗,2 , Finn Knut Hansen b,3 a b
Museum for Cultural History, Department of Conservation, University of Oslo, Frederiks gate 3, P.O. Box 6762, St. Olavs plass, 0130 Oslo, Norway Department of Chemistry, University of Oslo, P.O. Box 1033, Blindern, 0315 Oslo, Norway
a r t i c l e
i n f o
Article history: Received 14 February 2012 Accepted 22 February 2012 Available online 31 March 2012 Keywords: Archaeological wood Consolidants Biomimetics New materials Cellulose Open structure
a b s t r a c t Given the perilous state of the Oseberg find from Norway, the Museum of Cultural History and the Department of Chemistry both at the University of Oslo, are looking into new methods for treating archaeological wood. While numerous polymers have been previously tested, most do not stabilise the wood sufficiently, penetrate far enough, or remain stable without producing toxic fumes. A few of the more common examples are: Alum salt, KAl(SO4 )2 ·12H2 O, which was used for treatment earlier but does not penetrate well and leaves the wood very acidic. Poly(oxy ethylene) (POE or Polyethylene glycol [PEG]) is widely used as a consolidant today but this material degrades over time and thus cannot support the finds for a very long time. Melamine-formaldehyde (Kauramin) has also been used and while it is fairly stable, it may also fill the wood and turn it into a ‘block’ of plastic. Since new consolidants would be advantageous, it is discussed what the requirements of such consolidants are and how material sciences may help procure them. It is proposed that an important requirement for a future stabilising agent is to leave an airy structure in order to allow retreatment in the future. This might be accomplished by foaming a polymer, or by combining nanoparticles with a polymer ‘spider web’ network to keep them in place. Such particles may help stabilise pH inside the wood by neutralising any acid generated inside treated artefacts. Special attention is given to the field of biomimetics–the discipline of constructing materials inspired by existing natural designs. It may be possible to construct a frame using bio-inspired materials (possibly an ‘artificial lignin’ mixed with other compounds optimise strength and flexibility) or through biomineralisation (an inorganic ‘skeleton’). Tests on biomimetic cellulose and chitosan have begun and the initial evaluation of these materials is given. Chitosan is made from modified chitin (primarily from shrimp and crabs) and may be dissolved in acidic solutions. Crystalline cellulose is interesting in conservation as the individual particles are resistant to acid and not as hygroscopic as the amorphous part of cellulose. The materials and the procedures used in testing are described. It is shown that crystalline cellulose particles usually flocculate when used to treat archaeological wood but that they may be treated with surfactants in order to improve penetration of archaeological finds. © 2012 Published by Elsevier Masson SAS.
1. Research aims
2. Introduction
Current methods for treating archaeological wood often have undesirable side-effects and the stability of previously treated artefacts may be threatened by these. This is an attempt to specify requirements of theoretical new consolidants by looking at fields such as materials science and biomimetics (materials inspired by natural designs).
Knowledge of previous treatments of archaeological wood and their consequences becomes more and more important in conservation. When problems and damaging processes caused by past conservation are recognised, possible retreatment procedures and preservation strategies have to be considered. However, materials must be chosen and the criteria for evaluating suitable re-conservation methods must be implemented. The historic treatments teach us that no method–no matter how well suited to the task it seems to be at the time–will preserve objects forever. For this reason, conservation ethics claim that every material used must be removable or, more generally, leave the object retreatable. The American Institute for Conservation of Historic and Artistic Works (AIC) code states that “The conservation professional must strive to select methods and materials that, to the best of current knowledge,
∗ Corresponding author. E-mail addresses: [email protected] (M. Christensen), [email protected] (H. Kutzke), [email protected] (F.K. Hansen). 1 Tel.: +47 22855693. 2 Tel.: +47 22859477. 3 Tel.: +47 22855554. 1296-2074/$ – see front matter © 2012 Published by Elsevier Masson SAS. doi:10.1016/j.culher.2012.02.013
S184
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
do not adversely affect cultural property or its future examination, scientific investigation, treatment, or function” [1]. Unfortunately, practice is usually more complicated than theory. This ethical dilemma is exemplified by the alum-conserved wooden objects from the Oseberg find, excavated in 1904 from a burial mound in Vestfold, Norway [2,3]. This uniquely richly carved Viking Age find was a burial of two high-standing women but besides their skeletons the mound also contained 15 horses and an ox, food, textiles, numerous tools for daily life, and ceremonial equipment–such as a wagon and four sledges. Many of the wooden objects have been alum-treated (a short description of the treatment is given in the section below). The collection also contains the famous Oseberg ship but this was never alum-treated since it is made of heartoak and was thus in a much better state than most of the other wooden items which were made from other types of wood. In the century which has passed since the treatment of the artefacts, they have become increasingly unstable [4]. Most are now unable to support their own weight and cannot be safely removed from their current display frames. The matter is further complicated by very low pH inside the objects. Metal ions from both treatments and the numerous metal screws and nails used to fasten the many wooden bits to the metal frames also catalyse degradation processes. If nothing is done, the most famous Viking Age find in Norway will be irrevocably lost. The complex problems concerning the development of a reconservation and preservation strategy for the Oseberg find may be a model case for the future handling of problems caused by historical conservation treatments. Even if the artefacts are stabilised, a future retreatment (possibly in a hundred years) must be taken into account. After all, the consequences of the alum-treatment teach us how vital it may be to have this option. Materials for wood conservation available today are removable only by washing out using water, organic solvents, or supercritical carbon dioxide. Indeed several polymers used in the conservation of waterlogged wood (such as poly(ethylene oxide) POE/PEG or melamine-formaldehyde) may not be completely removable at all. A future retreatment of objects of high complexity–like the ones from the Oseberg find–will be almost impossible given current conservation procedures. Additionally, previously used bulking materials typically have undesirable drawbacks such as darkening the wood, filling it so that no further work is possible, or releasing toxic reactants [5]. Thus, to secure wooden cultural heritage in the future, we must first look to the past and propose new strategies without the shortcomings of former materials.
3. Previously used materials Several materials have been used for the conservation of waterlogged wood. The most important of these as well as their consequences will be briefly discussed since it is essential to understand said consequences if new materials are to avoid known ‘pitfalls’. Restoration of historical finds have (until relatively few years ago) relied upon polymers which were not tailor-made to the object in question and eventually caused damage to the artefacts. This, for example, is the case with many resins which were thought to be reversible when applied to artefacts during the 1960s. Removal of polymer material is generally very difficult, resulting in damage to the finds [6]. Also, many of the treatments originally thought to be reversible are not so–as was for example studied with various polymers for the treatment of ceramics [7]. Polymers used for the conservation of wood must also be polymerised in situ or be very low molecular weight as anything with a diameter larger than about 0.55 nm cannot penetrate into the cell wall to replace adsorbed
water [5]. Depth of penetration becomes a vital factor when treating archaeological wood. A few of the more commonly used consolidants will be mentioned below. In addition to these, many different consolidants have been tested by conservators in the past. An overview of how to choose between the most common ones was recently presented at the 10th International ICOM-WOAM Conference [8]. A more complete list of treatments can be found in Ref. [5]. Alum is the salt KAl(SO4 )2· ·12H2 O (potassium aluminium sulphate dodecahydrate). During impregnation, objects are heated for about two hours in a supersaturated alum solution at more than 80 ◦ C. Alum crystals form on the surface of the objects [5]. Additionally, recent experiments indicate that sulphuric acid is released in the wood during the hot alum-treatment [9]. This means that the pH value of alum-treated artefacts becomes so low that even the lignin remaining in the objects is actively degraded. As metal ions may act as catalysts for further degradation, the presence of aluminium ions may also prove detrimental to the finds. PEG, also known as polyethylene oxide or by its IUPAC (the International Union of Pure and Applied Chemistry) name, polyoxy ethylene (POE), is a polymer with the repeating unit (OCH2 CH2 ). Long chains of PEG do not easily penetrate far into the wood, while short chains do not stabilise the surface in a satisfactory way [10]. Therefore, a two-step process for PEG impregnation has been recommended, using a low molecular weight chain (such as PEG 200) to penetrate far into the wood and hopefully stabilise the core. A longer-chain PEG (such as PEG 3000 or 4000) bath is then used to stabilise the fractured outer layers of the wood [5]. Problems with optimal freeze-drying temperature, when using mixtures of PEG200 and PEG2000, have led to the conclusion that it is probably sufficient to use only relatively longchained PEG in conservation [11]. Unfortunately, a degradation of PEG occurs inside the treated objects. Studies of the PEG in the Swedish warship Vasa seem to indicate a random cleavage of the chains [12]. Melamine-formaldehyde consolidation has been developed in Germany where the treatment is also referred to as the ‘Kauramin method’. Solubility in water and the small molecular size of the components allows for good penetration into wood. Treatment time is fairly short (from weeks to a year plus subsequent drying for weeks, months, or years in polyethylene wrapping), but the cured polymer is difficult to remove from the surfaces of treated objects and these must be treated to improve colour tone [13]. Despite the good chemical stability of melamine-formaldehyde, the method tends to fill all voids in the structure, leaving no room for later retreatment. This means that treated objects more or less become blocks of polymer with a limited lifetime and no way to retreat the finds. Silicon-containing polymers have been tested with relatively promising results. They are applicable to both wood and leather [14]. Consolidants such as dicarboxylic acids have been proposed since they can penetrate far into wood and have melting points in the range 50–100 ◦ C [15]. Unfortunately, more thorough data on stability and degradation is still needed before the effectiveness of these polymers can be properly evaluated.
4. Other fields and materials The enormous development of material sciences may provide us with completely new materials, strategies, and approaches when designing materials meeting special requirements. We may see terms like ‘re-conservation’ or ‘retreatment’ from a new angle. Ideas on how such a new generation of wood consolidants might be
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
designed were expressed (independently) by A. Truyen at a COST workshop in 20091 . The Museum of Cultural History and the Department of Chemistry, both part of the University of Oslo, have initiated a research project to help develop and test new materials for wood conservation. Above all, it is vital that the object has an open structure after treatment, leaving room for further re-conservation of the artefact in case the current consolidant proves to be unstable. Such an open structure could be realised by making consolidant ‘tentacles’ or applying a kind of ‘spider web’ inside the objects, leaving voids or pores through which a possible future consolidant may penetrate the artefact. Guaranteeing re-treatability due to ‘voids’ in the structure–possibly introduced through emulsion of the bulking agent and/or later freeze-drying of the treated object–is thus of much higher importance than finding a reversible treatment. Since the majority of the leftover wooden material consists of lignin [16], it is particularly interesting to look to lignin-like materials–either a kind of artificial lignin or actual lignin reused as part of the treatment. Having lignin-like properties means that it will be easy to get the bulking agent into the wood–and that even if it gets stuck in there it will not drastically alter the properties of the artefact (at least not from a chemical viewpoint). It has been proposed that waste lignin from for example the paper-making industry might be reused as a consolidant to prevent washout of sandy ground. Cheap and environmentally friendly products for this purpose already exist [17]. Wood-like lignin has been synthesised and it has been observed that both the amount of available monomers and the pH affect the structure of the final product [18]. Dehydrogenase polymer (DHP), a model compound made from enzymatic polymerisation of coniferyl alcohol, seems to form onion-like layers when polymerised at neutral pH and at room temperature [19]. In another test where lignin was extracted directly from wood to form lignophenols–possibly a future source of fuel–it was found that the polymerised matrices held powdery substances effectively [20]. Since some dry, degraded archaeological wood is powdery, this idea might be transferable to the field of cultural heritage. Lignophenol has been applied to archaeological wood where it overall gave better strengthening than PEG4000. Unfortunately, pure water may not be a suitable solvent for impregnation [21]. We propose that existing lignin–such as the surplus lignosulphonate from the paper industry–could be treated to adhere better to the archaeological wood. This may require acetylation or similar modification of the wood and/or lignin. Alternatively, lignin may be used as a filler material when impregnating the wood with a polymer. In this way, degraded wood can be strengthened by what is, chemically speaking, more wood. In this way, it would be ensured that dimensional change during fluctuations in relative humidity would be identical for the wood and the consolidant. Phenol-formaldehyde (PF) is currently being investigated at the University of Oslo. While the curing process is irreversible, the polymer has a structure very similar to lignin but is more resistant to degradation. The first commercial plastic, BakeliteTM , was a phenolformaldehyde, and items made from it are the only plastic objects which have endured a hundred years worth of real-time ageing [22]. Unfortunately, most phenolic compounds used are glues (for instance for making plywood) and it is important that these do not penetrate too far into the wooden material [23]. This makes it
1 Arnold Truyen, Conservation of wood - selected experiments, lecture at “Consolidation, reinforcement and stabilisation of decorated wooden artifacts”, WoodCultHer COST ACTION IE0601, Prague 30th–31st March 2009.
S185
hard to use the experience from the industry in the field of cultural heritage (where thorough penetration is necessary). Even though PF has been used for a century, the exact relationship between different reaction sites and ratio of reactants is still being discovered. It has been found that hexamethylene tetramine (HMTA), a common cross-linker, preferably reacts with a PF with a high ratio of ortholinkages. A slight surplus of phenol (1.2 to 1.3 phenol molecules per formaldehyde molecule) promotes this orientation [24]. Thus it may be possible to affect the structure in similar ways to promote greater compatibility with the wooden matrix. PF has been used in conservation of metals [25], fossils, and wood [26] but often at high temperature and with less satisfactory results than urea-formaldehyde [27]. It can be difficult to find references for PF treatment of wooden objects and problems with too fast curing times (despite excellent penetration), possible emission of formaldehyde, and above all the irreversibility of the curing process caused the method to be discontinued for treatment of wooden artefacts during the 1940s [5,26]. Thus it must probably be modified in order to ensure a better treatment. The lack of published information further makes it difficult to evaluate how effective the method is and how recent discoveries in phenol-formaldehyde research can be applied to the field of cultural heritage. Our own studies indicate that the PF penetrates wooden artefacts thoroughly and can be polymerised in situ at very low pH, resulting in open structures as long as the wood is fairly intact. The very degraded alum-treated wood from the Oseberg find, however, cannot be safely treated as the alum crystals dissolve in the watery pre-polymer mix. The finds are currently so fragile that they are held together by a thin ‘crust’ of alum crystals, lacquer and some remaining wood. If the alum salts will be dissolved the object may disintegrate. A more thorough report on PF for this use will be published elsewhere [28]. Other molecules than unmodified phenol can be incorporated into the structure of PFs, for example bisphenol A [29] or sucrose and similar carbohydrates (with xylitol giving the best result) [30]. Phenol can even be entirely substituted with furan [31], although our own tests have shown that pure furfuryl alcohol or furan blackens during self-polymerisation and is so extremely exothermic that it will likely damage the wood. It has also been proven that lignin can be incorporated into the PF structure [32] or even react directly with formaldehyde without the use of phenol [33]. The main goal is to introduce a polymer in such a way that it coats the remaining cell wall material or forms a ‘spider web structure’ which strengthens the lignin without filling the object completely. Various ‘fillers’ are also considered–including lignin as well as inorganic materials like calcium carbonate. Polymerisation in buffer solutions can be used to both control the rate of curing of the polymer and even neutralise the very acidic alum-treated wood from Oseberg and similarly affected artefacts. One possibility is to introduce aliphatic chains between oligomer PF molecules (for example about 6–10 phenol molecules per oligomer) to ensure some flexibility so the polymer can swell and shrink with the wood. Such a polymer would be similar to wood in structure and resistant to most solvents (due to the phenolic groups), while maintaining enough flexibility to swell with the wood and avoid being as brittle as traditional phenolformaldehyde. Some ideas from materials science have already been implemented in treatment of non-wooden cultural heritage–such as using nanoparticles for the conservation and restoration of wall paintings. The small size and large surface areas of these particles allow for good penetration. In addition, inorganic materials can be reinforced with the exact same material as the artefact is made from [6]. The successful integration of Si-containing polymers [14]
S186
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
means that Si-based particles will likely also adhere fairly well to a wooden matrix. 5. Bio-inspired materials Another possibility is to learn from nature. An important field for this is biomimetics, materials inspired by natural designs. Such materials are often ‘green’ (meaning their components and byproducts have very low toxicity) and easily form hybrid materials where individual parts have very different qualities. A key feature in most natural materials is a hierarchical structure which has been optimised to provide high strength and/or flexibility through porous structures. Several examples are discussed in Ref. [34]. Such qualities are very desirable when designing materials to be used for the conservation of wooden artefacts. It may even be possible to implement useful qualities of other plants. One of the most famous–and well-studied–examples is probably self-cleaning surfaces on plants which remain free from water and dirt due to its microstructure. Several examples are given in Ref. [35] (the most famous of which is probably the lotus plant). Biomineralisation, the process by which biological organisms produce minerals (such as the shells on molluscs and algae, or skeletal parts of vertebrates), has been used to grow artificial bone material [36]–and even used to create highly durable non-organic ‘copies’ of wood [37]. In the study in question, silicon carbide strings penetrated into wood cells to form a heat-resistant ‘copy’ upon calcination. For conservation of wooden items, however, biomimetic materials utilising wood-like components are also extremely interesting as they most likely interact well with degraded wood. This might make it possible to ‘grow’ a bionic self-organising ‘skeleton’ framework inside the degraded wood–possibly combined with inorganic particles for support or acid neutralisation. Hydroxides might be used to neutralise acid generated inside the conserved objects–this has already been tested on wood from the Vasa [38]. Initial experiments on in situ ‘growth’ of acidneutralising calcium carbonate on archaeological wood (at the Max Planck Institute for Colloids and Interfaces in Potsdam, Germany), seem promising–even though the tested technique might be too alkaline to be safely applied to unearthed objects. Most probably, nanoparticles will be incorporated into degraded wood in the future and thus help save wooden cultural heritage as either bulking agents, reinforcement points for a polymer network, or reservoirs of acid-neutralising material. 6. Cellulose and chitosan Cellulose is the most common renewable polymer in nature, making up 40–50% of wood and 90% of cotton fibre, with more than 7.5 × 1010 tons produced yearly [39,40]. In addition to wood and other plants (particularly hemp, flax, jute, ramie, and cotton) bacteria and tunicates may also produce cellulose [41]. Cellulose is very stable due to a network of hydrogen bonds which prevents melting and dissolution with many solvents [41]. Adjacent cellulose chains can thus fit together in crystalline regions. Crystallinity thus affects swelling and absorbtion of the fibres. The less dense amorphous regions, however, can be dissolved using acid (the type, temperature and treatment time affects the properties of the finished product). The remaining cellulose crystals are often referred to as ‘whiskers’ although crystalline cellulose is known by many names such as ‘cellulose nanocrystals’, ‘microcrystals’, ‘nanocrystals/particles’, ‘microcrystallites’, and ‘nanofibres’ [39,40]. Technically, ‘whiskers’ refer to rod-like nanoparticles while ‘nanofibrils’ should be used to describe long flexible nanoparticles with both amorphous and crystalline regions [41]. Most whiskers from plant sources have lengths of 100–300 nm while those from
tunicates or bacteria can be up to 3000 nm long. Diameters range from 3 to 70 nm with typical values around 15–20 nm [40]. While this makes the individual particles significantly bigger than the monomers used in the phenol-formaldehyde experiments, they are of comparable size to nanoparticles roughly 100 nm in diameter used to treat wood from the Vasa [38]. Cellulose was first isolated in 1832 and has been thoroughly studied since then. In the 1950s, it was reported that cellulose fibres can make stable suspensions in water after degradation with sulphuric acid [40]. This acid, in particular, can produce whiskers which do not flocculate due to electrostatic repulsion, and sulphur content can therefore be used as a measure of properties. H2 SO4 treated fibres contain some weak acids (sulphate ester and carboxyl groups) which are less common in fibres treated with HCl [39]. The residues are not acidic enough to threaten the durability of the wood and even allow for further chemical modification of the fibres (for example, allowing them to be incorporated into polymer films). The whiskers have varying degrees of polymerisation and lengths depending upon their origin. Those from wood are about 180–200 nm with those from cotton are 100–120 nm and those from tunicates can reach 1000 nm in length [39]. This means that whiskers from plants are better than those from other sources when treating archaeological wood as the dimensions of the fibres should be kept small to enhance penetration. The actual preparation is simple. Fibrous cellulose (SigmaAldrich) was treated in 67% H2 SO4 : 5 g cellulose in 100 g acid was heated to 60 ◦ C for 1 h under heavy stirring. The remaining mix was centrifuged at 10,000 RPM for 10 min. The supernatant was discarded and replaced with deionised water. This process was repeated at least five times until a turbid supernatant phase appeared. This was collected and ultrasonicated using a tip sonicator at 30% power output for five minutes to disperse the whiskers. It was observed that using metal (aluminium) spoons when handling the cellulose caused it to degrade upon immersion in acid, turning the mix dark brown and ruining the whiskers. After penetration, the samples were cut and observed using [environmental] scanning electron microscopy ([E]SEM). Imaging was initially performed on an FEI FE-ESEM, Quanta 600 (FEI Company, Oregon, USA) in low vacuum mode using a large field detector to minimise sample preparation. Later on, a 5 nm Au/Pd coating was applied to the samples before imaging on a JEOL JSM-7500F FE-SEM (Cryo-SEM) using a secondary electron in-lens detector (JEOL, Eching, Germany). Acceleration voltages used are noted on the individual images. Initial tests of such suspensions showed that the cellulose did not penetrate thoroughly into the wood–although it did adhere well to the surface of the wood (as seen in Fig. 1). Even though the appearance was unacceptable for display, it proved that cellulose can act as a gap filler and form the desired open structure. At high
Fig. 1. Freeze-dried Viking Age wood treated with cellulose whiskers. Note that the cellulose adheres to the surface of the wood and forms an open structure.
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
S187
Fig. 2. Cryo-SEM (scanning electron microscopy) image of freeze-dried gold/palladium-coated cellulose whiskers. Note that individual whiskers can be seen in the conglomerate.
magnification, it is possible to see the individual fibres in a SEM at low acceleration voltages (Fig. 2), and shows their dimensions to be roughly 15 nm in diameter and 200–300 nm in length. At a certain concentration, the whiskers form an ordered (chiral nematic) phase. The structure tends to be retained even if the fibres are covered in surfactants. Salt (NaCl) concentrations as low as 10−5 M may screen the electrostatic interactions [39]. The equilibrium between isotropic and anisotropic phases is sensitive to the presence of electrolytes and the kind of counterions used as well as the sulphate group content [40]. When treating archaeological wood, however, these things mean that the ions in the wood will cause the cellulose whiskers to flocculate–preventing proper penetration. An example is seen in Fig. 3, which shows uneven dispersion of the whiskers. Due to the hydroxyl groups on the surface of the whiskers, both chemical modification and various surfactants can affect the whiskers. Actual polymer grafting onto the surface of the whiskers has been successfully implemented [40]. In addition to ionic species
Fig. 3. Whiskers on the surface os an impregnated piece of Viking Age wood (right). Note the relatively poor dispersion in the wood structure itself (left).
Fig. 4. ESEM image of Viking Age wood treated with cellulose whiskers and PEG1000. Note the improved penetration due to the effects of the surfactant.
grafted during acid-treatment, whiskers may also be kept in suspension by steric repulsion between POE chains grafted onto the rods. Surfactants may be used to obtain dispersions in organic solvents as well [42]. The surfactant molecules adhere to the surface of the whiskers, leaving molecular chains ‘drifting’ in all directions, and thus physically prevent whiskers from sticking together. Use of non-ionic surfactants is especially attractive as they allow whiskers to remain suspended even in water with a high ion concentration where the whiskers otherwise flocculate. Initial tests with cellulose using PEG1000 as a stabiliser showed that it significantly improved penetration (Fig. 4). In the future, various other surfactants should be tested to see which ones provide the best effect. Cellulose also provides high axial stiffness, which is particularly advantageous when using fibres to reinforce other polymers. Even adding water can dramatically alter the modulus of cellulose composites [41]. This means that it is important to keep the cellulose in museums dry. Cellulose nanowhiskers have thus been incorporated as fillers into a significant number of polymers [40]. Upon successful introduction, cellulose whiskers prevented cracking in urea formaldehyde (UF) polymer bond lines which is frequent in UF without cellulose [41]. Cellulose whiskers have been combined with another tree component as part of a triblock copolymer. The plant matter used was xyloglucan oligosaccharides from tamarind kernels and the polymer a xyloglucan-oligosaccharide-polyethylene oxide-polystyrene (XGO-POE-PS) triblock copolymer. A diblock copolymer of XGO and PS was tested but precipitated and could not stabilise the whiskers in watery dispersions [43]. This is, to the best of our knowledge, the only attempt so far to approach a true wood structure by mixing cellulose whiskers with components which resemble hemicelluloses and attempts to incorporate other wood-like structures may be ideal for conservation purposes. Cellulose and cellulose whiskers have been modified for use in conservation science with the specific goal of creating a waterdispersible polymer which could stabilise archaeological wood [44]. Allylic groups were grafted onto the cellulose in order to allow the polymer to cross-link at 60 ◦ C. This stabilised the wood but made it difficult to remove the polymer. Carboxymethyl groups were introduced in order to make the cellulose dispersible in water.
S188
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
This also allowed the consolidant to be washed out of the treated wood. A test piece of archaeological wood was treated for 20 days but the dispersion did not fully penetrate the wood during this time. Chitin is the polysaccharide 2-acetamido-2-deoxy--d-glucose and is identical to cellulose with a hydroxyl group replaced by an acetamido group. Chitosan is a derivative of chitin produced by alkaline deacetylation and can be obtained from crab or shrimp shells as well as mycelia. It is more biocompatible and biodegradable than typical synthetic polymers [45]. Chitosan has been investigated for drug delivery due to its low toxicity and biodegradability and may promote tissue growth and wound healing [45,46]. The same qualities, however, are beneficial for conservation science as it allows chitosan to be handled with minimal precautions. Since chitin and chitosan are highly basic, they differ from most polysaccharides (like cellulose, dextran or pectin) in nature. Chitosan can chelate metal ions and this may be utilised in wastewater treatment (both industrial and otherwise) [45]. The degree of chelation is influenced by the pH of the solution and concentration of ions as well as the size of the chitosan particles. The efficiency increases with increasing pH but takes place to some extent at pH levels at least as low as 2 [47]. This makes chitosan potentially viable for immobilising metal ions in archaeological wood since treatment times are usually quite long. Ideally, it allows chitosan to both reinforce the wood and stop metal ions in it, thus reducing the catalytic effect which said ions have on many degradation processes. The pH of the dissolved chitosan affects its physical properties. The viscosity of chitosan solutions is at its lowest at pH 3–4 and rises on either side of this window [48]. As such this is the ideal pH for penetrating chitosan throughout archaeological wood. Chitosan is soluble in monovalent acids but not divalent ones. Acid concentrations of 0.1–0.15 M can dissolve 2%w /v chitosan but generally not 2.5%w /v (out of the tested acids, only chloroacetic acid could dissolve this higher concentration). Using HCl leads to a less viscous system than using carboxylic acids [46]. In order to test penetration, solutions of 2% (w/w) chitosan in 0.1 M acids were made. It was found that such chitosan solutions are very viscous and may have trouble penetrating archaeological wood. Concentrations of about 1% chitosan were much less viscous while still
offering acceptable support to the deteriorated wood. The chitosan had a foam-like structure after freeze-drying and treated pieces felt notably stronger than those treated with cellulose whiskers. Storage tests indicated that chitosan kept in hydrochloric acid may begin to aggregate in less than 2 weeks, whereas this did not happen to chitosan dissolved in acetic acid. Since acetic acid is likely to cause less damage than hydrochloric acid to archaeological wood, it is recommendable to use this weaker acid even it is not an ideal solvent in a museum environment as it reacts with metals. However, it should be possible to raise pH to about 6 after penetration, either through buffers or simply by diluting with water. This would not only help remove excess acid but might also force the chitosan to precipitate inside the objects and thus further experiments should be conducted to determine if this would be beneficial. Adding salt to chitosan solutions has much the same effect as increasing surfactant concentration. This is because salt reduces the hydrodynamic volume of the molecules for both the hydrophobically modified and pure chitosan. In fact surfactant becomes much less important in solutions which contain 10 mM NaCl [49]. Since it is effectively impossible to remove salts from archaeological wood prior to impregnation, this effect might help to further disperse the chitosan inside the degraded wood, rendering surfactants superfluous. A chitosan-cellulose film has been made by dipping a quartz plate into alternating solutions of the two polymers [50]. Note, however, that cellulose and chitosan stick together because they have opposite charges at pH values around 3. This means that the molecules will flocculate rapidly if mixed and thus prevents archaeological wood from being treated with both polymers simultaneously. Still, it might be possible to build composite films inside the wood by immersing it in alternating solutions/dispersions of chitosan/cellulose and such films may have better mechanical properties than a layer of one of the polymers. 7. Concluding remarks Advantages and disadvantages of various proposed methods are presented in Table 1.
Table 1 Overview of various possible methods for treating archaeological wood. Consolidant
Advantages
Disadvantages
Phenol- or melamine-formaldehyde
Phenol structure similar to lignin Penetrates easily in water Real-time ageing tests (PF) Very hard and durable Resistant to solvents/pH
Fills the wood as a ‘block’ of plastic Irreversible/Unremoveable Formaldehyde is toxic Literature only on high temperature/pressure Rather brittle
Lignin
Wooden structure Cheap Can be part of polymer
Large molecules cannot penetrate Probably irreversible Dark colour
Nanoparticles
Commercially available May neutralise acid Can be coated to adhere better Can be very durable Compatible with wood structure Resistant to solvents/pH
Difficult to disperse and penetrate May become too alkaline for wood Stabilising effect is unknown
Alifatic polymer
Flexible In situ polymerisation possible
Insufficient stabilisation Difficult penetration if not in situ Not completely reversible
Biomineralisation
Fits/fills structure “Built” in situ Re-treatable
Sensitive to pH Possibly incompatible with wood Not tested thoroughly
Biomimetics
Compatible with wood Usually non-toxic Water is an ideal solvent Easily forms open structures May entrap some ions (chitosan)
May require additional reinforcement Surfactants/acids required for dispersion Can be pH-sensitive
Nanotubes
Very hard to penetrate Black colouration
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
No matter which route is used when designing better consolidants for archaeological wood, there are many requirements for a stabilising agent: The consolidant must be workable in a museum environment (at normal temperature, pressure, and without emitting toxic fumes). The material cannot expand, contract or otherwise warp the wood. Likewise, it must have a slightly acidic pH value suitable for wood. It must be able to penetrate the wood thoroughly and last for a very long time. As mentioned, the most important aspect is the ‘open’ structure which must be possessed by the bulking material in order to ensure that retreatment is possible. When combined, these diverse disciplines may provide new ways of looking at the consolidation of archaeological wood. Such a combination might entail forming composite materials where nanoparticles fill the bulk of the wood while being held in place by polymer material which adheres to both the particles and the existing wooden structure. This may be done as a kind of ‘spider web’ which envelops the wood without reacting with it, or a ‘copolymer’ system which bonds chemically to the leftover lignin. Existing designs in nature may help as materials like cellulose and chitosan might both have beneficial properties–especially if combined with other polymers to add strength and durability as well as nanoparticles which serve to neutralise and acid generated in the wood. It is hoped that these ideas and requirements can help crossdisciplinary collaboration and encourage people working with wood–be they chemists, wood technologists, or conservators–to think along new lines in order to secure our cultural heritage. Acknowledgements We would like to thank Susan Braovac from the Museum for Cultural History, University of Oslo, for her help in elucidating the current state of the Oseberg wood and its previous treatments. Poul Jensen and Yvonne Shashoua, both from the National Museum of Denmark, Department of Conservation, are gratefully acknowledged for their information and advice on the PEG treatment and the degradation of polymers, respectively. Jaan Roots from the Department of Chemistry, University of Oslo, is thanked for his helpful discussions on polymer design. The COST Action FP0802, Experimental and Computational Micro-Characterization Techniques in Wood Mechanics, is thanked for a grant to perform a short-term scientific mission at the Max Planck Institute for Colloids and Interfaces in Potsdam, Germany, where Ingo Burgert, Friederike Saxe, Anayancy Osorio, and Susann Weichold from the Biomimetics Plant group were of invaluable help when initiating experiments on cellulose whiskers and recording SEM images while Kyu-Bock Lee initiated experiments on in situ growth of calcium carbonate. References [1] Code of Ethics and Guidelines for Practice, http://www.conservationus.org/index.cfm?fuseaction=page.viewPage&pageID;=858&nodeID=1 (14/82009). [2] A.M. Rosenqvist, The stabilizing of wood found in the Viking Ship of Oseberg–part I and II, Stud Conserv 4 (1) (1959) 13–22, 62-72. [3] A.V. Brøgger, H.G. Falk, H. Schetelig, Osebergfundet, I, Published by the Norwegian state, Kristiania, 1917. [4] The Oseberg find 100-year anniversary–the great adventure in Norwegian archaeology. http://www.khm.uio.no/utstillinger/oseberg/indexE.html (18/08-2009). [5] A. Unger, A.P. Schniewind, W. Unger, Conservation of wood artifacts, SpringerVerlag, Berlin Heidelberg, 2001. [6] P. Baglioni, R. Giorgi, Soft and hard nanomaterials for restoration and conservation of cultural heritage, Soft Matter 2 (2006) 293–303. [7] M. Favaro, R. Mendichi, F. Ossola, U. Russo, S. Simon, P. Tomasin, et al., Evaluation of polymers for conservation treatments of outdoor exposed stone monuments. Part I: photo-oxidative weathering, Polym Degrad Stabil 91 (2006) 3083–3096.
S189
[8] P. Hoffmann, On the efficiency of stabilisation methods for large waterlogged wooden objects, and on how to choose a method, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 323–350. [9] S. Braovac, H. Kutzke, The presence of sulphuric acid in alum-conserved wood – Origin and consequences, contribution at “Wood Science for Conservation of Cultural Heritage”, WoodCultHer COST Action IE0601, Hamburg, 7th–9th October, 2009 (see also the paper by Braovac and Kutze in current issue). [10] R.M. Rowell, R.J. Barbour (Eds.), Archaeological wood: properties, chemistry, and preservation, American Chemical Society, Washington DC, USA, 1990. [11] P. Jensen, S. Schnell, The implication of using low molecular weight PEG in impregnation of waterlogged archaeological wood prior to freeze-drying, in: P. Hoffmann, K. Strætkvern, J.A. Spriggs, D. Gregory (Eds.), Proceedings of the 9th ICOM Group on Wet Organic Archaeological Materials Conference, Copenhagen, 2004, pp. 279–307. [12] G. Almkvist, I. Persson, Iron catalysed degradation processes in the Vasa, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 499–505. [13] M. Wittköpper, Der aktuelle Stand der Konservierung archäologischer Naßhölzer mit Melamin/Aminoharzen am Römisch-Germanischen Zentralmuseum, Arbeitsblätter für Restauratoren Gruppe 8: Holz Heft 2, (1998), 277–282. [14] C. Wayne Smith, Archaeological conservation using polymers: practical applications for organic artifact stabilization, Texas A&M University Press, College Station, USA, 2003. [15] G. Chaumat, C. Albino, L. Blanc, Development of new consolidation treatments from fatty acid resin solution, in: Program and abstracts of papers and posters ICOM–WOAM Conference, Amsterdam, 2007. [16] M.P. Colombini, M. Orlandi, F. Modugno, E.-L. Tolppa, M. Sardelli, L. Zoila, et al., Archaeological wood characterisation by PY/GC/MS, GC/MS, NMR and GPC techniques, Microchem J 85 (2007) 164–173. [17] G. Shulga, T. Betkers, V. Shakels, B. Neiberte, A. Verovkins, J. Brovkina, et al., Effect of the modification of lignocellulostic materials with a lignin-polymer complex on their mulching properties, BioResources 2 (4) (2007) 572–582. [18] N. Terashima, R.H. Atalla, S.A. Ralph, L.L. Landucci, C. Litpierre, B. Monties, New preparations of lignin polymer models under conditions that approximate cell wall lignification, Holtzforschung 50 (1996) 9–14. [19] M. Micic, M. Jeremic, K. Radotic, M. Mavers, R.M. Leblanc, Visualization of artificial lignin supramolecular structures, Scanning 22 (2000) 288–294. [20] Y. Nagamatsu, M. Funaoka, Design of recyclable matrixes from lignin-based polymer, Green Chem 5 (2003) 595–601. [21] T. Kataoka, Y. Kurimoto, Conservation of archaeological waterlogged wood by lignophenol, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 315–322. [22] D. Crespy, M. Bozonnet, M. Meier, 100 years of BakeliteTM , the material of a 1000 uses, Angew Chem Int Edit 47 (2008) 3322–3328. [23] J. Monni, L. Alvila, T.T. Pakkanen, Structural and physical changes in phenolformaldehyde resol resin, as a function of the degree of condensation of the resol solution, Ind Eng Chem Res 46 (2007) 6916–6924. [24] J. Huang, M. Xu, Q. Ge, M. Lin, Q. Lin, Y. Chen, et al., Controlled synthesis of highortho-substitution phenol-formaldehyde resins, J Appl Polym Sci 97 (2005) 652–658. [25] H.M. Sleicher, J.L. Everhart, Mounting of small metallographic specimens and metal powders in BakeliteTM , Metals Alloys 5 (1934) 59–60. [26] M. Ventikou, Old treatment, new problem: BakeliteTM as a consolidant, V & A Conserv. J. 32 (1999) 5–7, The paper is accessible at the time of writing at: http://www.vam.ac.uk/res cons/conservation/journal/journal32/ oldtreatment32/index.html. [27] L. Biek, J.W. Anstee, E.S. Cripps, A wooden bucket restored, Museums J. 57 (1958) 257–261. [28] M. Christensen, F.K. Hansen, H. Kutzke, Phenol-formaldehyde revisited–novolac resins for the treatment of degraded archaeological wood. To be published. [29] Y. Liu, C. Zhang, Z. Du, H. Li, Preparation and curing kinetics of bisphenol A Type novolac epoxy resins, J Appl Polym Sci 99 (2006) 858–868. [30] A.H. Conner, B.H. River, L.F. Lorenz, Carbohydrate modified phenolformaldehyde resins, J Wood Chem Technol 6 (4) (1986) 519–613. [31] E. Kumpinsky, pH effects on phenol-formaldehyde runaway reactions, Ind Eng Chem Res 34 (9) (1995) 3096–3101. [32] M. Turunen, L. Alvila, T.T. Pakkanen, J. Rainio, Modification of phenolformaldehyde resol resins by lignin, starch, and urea, J Appl Polym Sci 88 (2003) 582–588. [33] A.Y. Kharade, D.D. Kale, Effect of lignin on phenolic novolak resins and moulding powder, Eur Polym J 34 (2) (1998) 201–205. [34] P. Fratzl, R. Weinkamer, Nature’s hierarchical materials, Prog Mater Sci 52 (2007) 1263–1334. [35] K. Koch, B. Bhushan, W. Barthlott, Diversity of structure, morphology and wetting of plant surfaces, Soft Matter 4 (2008) 1943–1963. [36] A.C. Thackray, R.L. Sammons, L.E. Macaskie, P. Yong, H. Lugg, P.M. Marquis, Bacterial biosynthesis of a calcium phosphate bone-substitute material, J Mater Sci-Mater M 15 (2004) 403–406. [37] Y. Shin, G.J. Exarhos, Conversion of cellulose materials into nanostructured ceramics by biomineralization, Cellulose 14 (2007) 269–279.
S190
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
[38] R. Giorgi, D. Chelazzi, P. Bagleoni, Nanoscience contribution to the preservation of acidic shipwrecks, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 525–534. [39] M.M. de Sousa Lima, R. Borsali, Rodlike cellulose microcrystals: structure, properties, and applications, Macromol Rapid Comm 25 (2004) 771–787. [40] Y. Habibi, L.A. Lucia, O.J. Rojas, Cellulose nanocrystals: chemistry self-assembly, and applications, Chem Rev 110 (2010) 3479–3500. [41] S.J. Eichhorn, A. Dufresne, M. Aranguren, N.E. Marcovich, J.R. Capadona, S.J. Rowan, et al., Review: current international research into cellulose nanofibres and nanocomposites, J Mater Sci 45 (2010) 1–33. [42] C. Bonini, L. Heux, J.-Y. Cavaillé, P. Lindner, C. Dewhurst, P. Terech, Rodlike cellulose whiskers coated with surfactant: a small-angle neutron scattering characterization, Langmuir 18 (2002) 3311–3314. [43] Q. Zhou, H. Brumer, T.T. Teeri, Self-organization of cellulose nanocrystals adsorbed with xyloglucan oligosaccharide-poly(ethylene glycol)-polystyrene triblock copolymer, Macromolecules 42 (2009) 5430–5432. [44] G. Cipriani, A. Salvini, P. Baglioni, E. Bucciarelli, Cellulose as a renewable resource for the synthesis of wood consolidants, J Appl Polym Sci 118 (2010) 2939–2950.
[45] M.N.V.R. Kumar, A review of chitosan applications, React Funct Polym 46 (2000) 1–27. [46] Q.S. Zhao, X.J. Cheng, Q.X. Ji, C.Z. Kang, X.G. Chen, Effect of organic and inorganic acids on chitosan/glycerophosphate thermosensitive hydrogel, J Sol-Gel Sci Techn 50 (2009) 111–118. [47] A. Burke, E. Yilmaz, N. Hasirci, O. Yilmaz, Iron(III) ion removal from solution through adsorption on chitosan, J Appl Polym Sci 84 (2002) 1185–1192. [48] A.-L. Kjøniksen, B. Nyström, T. Nakken, O. Palmgren, T. Tande, Effect of surfactant concentration, pH, and shear rate on the rheological properties of aqueous systems of a hydrophobically modified chitosan and its unmodified analogue, Polym Bull 38 (1997) 71–79. [49] A.L. Kjøniksen, B. Nyström, C. Iversen, T. Nakken, O. Palmgren, T. Tande, Viscosity of dilute aqueous solutions of hydrophobically modified chitosan and its unmodified analogue at different conditions of salt and surfactant concentrations, Langmuir 13 (1997) 4948–4952. [50] G.H. Vikhoreva, N.R. Kil’deeva, I.N. Gorbacheva, E.A. Shablykova, S.Z. Rogovina, T.A. Akopova, Study of cellulose-chitosan composites. Solid-phase modification, rheology, films, Fibre Chem+ 32 (6) (2000) 402–406.
Available online at
www.sciencedirect.com
New materials used for the consolidation of archaeological wood–past attempts, present struggles, and future requirements Mikkel Christensen a,b,1 , Hartmut Kutzke a,∗,2 , Finn Knut Hansen b,3 a b
Museum for Cultural History, Department of Conservation, University of Oslo, Frederiks gate 3, P.O. Box 6762, St. Olavs plass, 0130 Oslo, Norway Department of Chemistry, University of Oslo, P.O. Box 1033, Blindern, 0315 Oslo, Norway
a r t i c l e
i n f o
Article history: Received 14 February 2012 Accepted 22 February 2012 Available online 31 March 2012 Keywords: Archaeological wood Consolidants Biomimetics New materials Cellulose Open structure
a b s t r a c t Given the perilous state of the Oseberg find from Norway, the Museum of Cultural History and the Department of Chemistry both at the University of Oslo, are looking into new methods for treating archaeological wood. While numerous polymers have been previously tested, most do not stabilise the wood sufficiently, penetrate far enough, or remain stable without producing toxic fumes. A few of the more common examples are: Alum salt, KAl(SO4 )2 ·12H2 O, which was used for treatment earlier but does not penetrate well and leaves the wood very acidic. Poly(oxy ethylene) (POE or Polyethylene glycol [PEG]) is widely used as a consolidant today but this material degrades over time and thus cannot support the finds for a very long time. Melamine-formaldehyde (Kauramin) has also been used and while it is fairly stable, it may also fill the wood and turn it into a ‘block’ of plastic. Since new consolidants would be advantageous, it is discussed what the requirements of such consolidants are and how material sciences may help procure them. It is proposed that an important requirement for a future stabilising agent is to leave an airy structure in order to allow retreatment in the future. This might be accomplished by foaming a polymer, or by combining nanoparticles with a polymer ‘spider web’ network to keep them in place. Such particles may help stabilise pH inside the wood by neutralising any acid generated inside treated artefacts. Special attention is given to the field of biomimetics–the discipline of constructing materials inspired by existing natural designs. It may be possible to construct a frame using bio-inspired materials (possibly an ‘artificial lignin’ mixed with other compounds optimise strength and flexibility) or through biomineralisation (an inorganic ‘skeleton’). Tests on biomimetic cellulose and chitosan have begun and the initial evaluation of these materials is given. Chitosan is made from modified chitin (primarily from shrimp and crabs) and may be dissolved in acidic solutions. Crystalline cellulose is interesting in conservation as the individual particles are resistant to acid and not as hygroscopic as the amorphous part of cellulose. The materials and the procedures used in testing are described. It is shown that crystalline cellulose particles usually flocculate when used to treat archaeological wood but that they may be treated with surfactants in order to improve penetration of archaeological finds. © 2012 Published by Elsevier Masson SAS.
1. Research aims
2. Introduction
Current methods for treating archaeological wood often have undesirable side-effects and the stability of previously treated artefacts may be threatened by these. This is an attempt to specify requirements of theoretical new consolidants by looking at fields such as materials science and biomimetics (materials inspired by natural designs).
Knowledge of previous treatments of archaeological wood and their consequences becomes more and more important in conservation. When problems and damaging processes caused by past conservation are recognised, possible retreatment procedures and preservation strategies have to be considered. However, materials must be chosen and the criteria for evaluating suitable re-conservation methods must be implemented. The historic treatments teach us that no method–no matter how well suited to the task it seems to be at the time–will preserve objects forever. For this reason, conservation ethics claim that every material used must be removable or, more generally, leave the object retreatable. The American Institute for Conservation of Historic and Artistic Works (AIC) code states that “The conservation professional must strive to select methods and materials that, to the best of current knowledge,
∗ Corresponding author. E-mail addresses: [email protected] (M. Christensen), [email protected] (H. Kutzke), [email protected] (F.K. Hansen). 1 Tel.: +47 22855693. 2 Tel.: +47 22859477. 3 Tel.: +47 22855554. 1296-2074/$ – see front matter © 2012 Published by Elsevier Masson SAS. doi:10.1016/j.culher.2012.02.013
S184
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
do not adversely affect cultural property or its future examination, scientific investigation, treatment, or function” [1]. Unfortunately, practice is usually more complicated than theory. This ethical dilemma is exemplified by the alum-conserved wooden objects from the Oseberg find, excavated in 1904 from a burial mound in Vestfold, Norway [2,3]. This uniquely richly carved Viking Age find was a burial of two high-standing women but besides their skeletons the mound also contained 15 horses and an ox, food, textiles, numerous tools for daily life, and ceremonial equipment–such as a wagon and four sledges. Many of the wooden objects have been alum-treated (a short description of the treatment is given in the section below). The collection also contains the famous Oseberg ship but this was never alum-treated since it is made of heartoak and was thus in a much better state than most of the other wooden items which were made from other types of wood. In the century which has passed since the treatment of the artefacts, they have become increasingly unstable [4]. Most are now unable to support their own weight and cannot be safely removed from their current display frames. The matter is further complicated by very low pH inside the objects. Metal ions from both treatments and the numerous metal screws and nails used to fasten the many wooden bits to the metal frames also catalyse degradation processes. If nothing is done, the most famous Viking Age find in Norway will be irrevocably lost. The complex problems concerning the development of a reconservation and preservation strategy for the Oseberg find may be a model case for the future handling of problems caused by historical conservation treatments. Even if the artefacts are stabilised, a future retreatment (possibly in a hundred years) must be taken into account. After all, the consequences of the alum-treatment teach us how vital it may be to have this option. Materials for wood conservation available today are removable only by washing out using water, organic solvents, or supercritical carbon dioxide. Indeed several polymers used in the conservation of waterlogged wood (such as poly(ethylene oxide) POE/PEG or melamine-formaldehyde) may not be completely removable at all. A future retreatment of objects of high complexity–like the ones from the Oseberg find–will be almost impossible given current conservation procedures. Additionally, previously used bulking materials typically have undesirable drawbacks such as darkening the wood, filling it so that no further work is possible, or releasing toxic reactants [5]. Thus, to secure wooden cultural heritage in the future, we must first look to the past and propose new strategies without the shortcomings of former materials.
3. Previously used materials Several materials have been used for the conservation of waterlogged wood. The most important of these as well as their consequences will be briefly discussed since it is essential to understand said consequences if new materials are to avoid known ‘pitfalls’. Restoration of historical finds have (until relatively few years ago) relied upon polymers which were not tailor-made to the object in question and eventually caused damage to the artefacts. This, for example, is the case with many resins which were thought to be reversible when applied to artefacts during the 1960s. Removal of polymer material is generally very difficult, resulting in damage to the finds [6]. Also, many of the treatments originally thought to be reversible are not so–as was for example studied with various polymers for the treatment of ceramics [7]. Polymers used for the conservation of wood must also be polymerised in situ or be very low molecular weight as anything with a diameter larger than about 0.55 nm cannot penetrate into the cell wall to replace adsorbed
water [5]. Depth of penetration becomes a vital factor when treating archaeological wood. A few of the more commonly used consolidants will be mentioned below. In addition to these, many different consolidants have been tested by conservators in the past. An overview of how to choose between the most common ones was recently presented at the 10th International ICOM-WOAM Conference [8]. A more complete list of treatments can be found in Ref. [5]. Alum is the salt KAl(SO4 )2· ·12H2 O (potassium aluminium sulphate dodecahydrate). During impregnation, objects are heated for about two hours in a supersaturated alum solution at more than 80 ◦ C. Alum crystals form on the surface of the objects [5]. Additionally, recent experiments indicate that sulphuric acid is released in the wood during the hot alum-treatment [9]. This means that the pH value of alum-treated artefacts becomes so low that even the lignin remaining in the objects is actively degraded. As metal ions may act as catalysts for further degradation, the presence of aluminium ions may also prove detrimental to the finds. PEG, also known as polyethylene oxide or by its IUPAC (the International Union of Pure and Applied Chemistry) name, polyoxy ethylene (POE), is a polymer with the repeating unit (OCH2 CH2 ). Long chains of PEG do not easily penetrate far into the wood, while short chains do not stabilise the surface in a satisfactory way [10]. Therefore, a two-step process for PEG impregnation has been recommended, using a low molecular weight chain (such as PEG 200) to penetrate far into the wood and hopefully stabilise the core. A longer-chain PEG (such as PEG 3000 or 4000) bath is then used to stabilise the fractured outer layers of the wood [5]. Problems with optimal freeze-drying temperature, when using mixtures of PEG200 and PEG2000, have led to the conclusion that it is probably sufficient to use only relatively longchained PEG in conservation [11]. Unfortunately, a degradation of PEG occurs inside the treated objects. Studies of the PEG in the Swedish warship Vasa seem to indicate a random cleavage of the chains [12]. Melamine-formaldehyde consolidation has been developed in Germany where the treatment is also referred to as the ‘Kauramin method’. Solubility in water and the small molecular size of the components allows for good penetration into wood. Treatment time is fairly short (from weeks to a year plus subsequent drying for weeks, months, or years in polyethylene wrapping), but the cured polymer is difficult to remove from the surfaces of treated objects and these must be treated to improve colour tone [13]. Despite the good chemical stability of melamine-formaldehyde, the method tends to fill all voids in the structure, leaving no room for later retreatment. This means that treated objects more or less become blocks of polymer with a limited lifetime and no way to retreat the finds. Silicon-containing polymers have been tested with relatively promising results. They are applicable to both wood and leather [14]. Consolidants such as dicarboxylic acids have been proposed since they can penetrate far into wood and have melting points in the range 50–100 ◦ C [15]. Unfortunately, more thorough data on stability and degradation is still needed before the effectiveness of these polymers can be properly evaluated.
4. Other fields and materials The enormous development of material sciences may provide us with completely new materials, strategies, and approaches when designing materials meeting special requirements. We may see terms like ‘re-conservation’ or ‘retreatment’ from a new angle. Ideas on how such a new generation of wood consolidants might be
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
designed were expressed (independently) by A. Truyen at a COST workshop in 20091 . The Museum of Cultural History and the Department of Chemistry, both part of the University of Oslo, have initiated a research project to help develop and test new materials for wood conservation. Above all, it is vital that the object has an open structure after treatment, leaving room for further re-conservation of the artefact in case the current consolidant proves to be unstable. Such an open structure could be realised by making consolidant ‘tentacles’ or applying a kind of ‘spider web’ inside the objects, leaving voids or pores through which a possible future consolidant may penetrate the artefact. Guaranteeing re-treatability due to ‘voids’ in the structure–possibly introduced through emulsion of the bulking agent and/or later freeze-drying of the treated object–is thus of much higher importance than finding a reversible treatment. Since the majority of the leftover wooden material consists of lignin [16], it is particularly interesting to look to lignin-like materials–either a kind of artificial lignin or actual lignin reused as part of the treatment. Having lignin-like properties means that it will be easy to get the bulking agent into the wood–and that even if it gets stuck in there it will not drastically alter the properties of the artefact (at least not from a chemical viewpoint). It has been proposed that waste lignin from for example the paper-making industry might be reused as a consolidant to prevent washout of sandy ground. Cheap and environmentally friendly products for this purpose already exist [17]. Wood-like lignin has been synthesised and it has been observed that both the amount of available monomers and the pH affect the structure of the final product [18]. Dehydrogenase polymer (DHP), a model compound made from enzymatic polymerisation of coniferyl alcohol, seems to form onion-like layers when polymerised at neutral pH and at room temperature [19]. In another test where lignin was extracted directly from wood to form lignophenols–possibly a future source of fuel–it was found that the polymerised matrices held powdery substances effectively [20]. Since some dry, degraded archaeological wood is powdery, this idea might be transferable to the field of cultural heritage. Lignophenol has been applied to archaeological wood where it overall gave better strengthening than PEG4000. Unfortunately, pure water may not be a suitable solvent for impregnation [21]. We propose that existing lignin–such as the surplus lignosulphonate from the paper industry–could be treated to adhere better to the archaeological wood. This may require acetylation or similar modification of the wood and/or lignin. Alternatively, lignin may be used as a filler material when impregnating the wood with a polymer. In this way, degraded wood can be strengthened by what is, chemically speaking, more wood. In this way, it would be ensured that dimensional change during fluctuations in relative humidity would be identical for the wood and the consolidant. Phenol-formaldehyde (PF) is currently being investigated at the University of Oslo. While the curing process is irreversible, the polymer has a structure very similar to lignin but is more resistant to degradation. The first commercial plastic, BakeliteTM , was a phenolformaldehyde, and items made from it are the only plastic objects which have endured a hundred years worth of real-time ageing [22]. Unfortunately, most phenolic compounds used are glues (for instance for making plywood) and it is important that these do not penetrate too far into the wooden material [23]. This makes it
1 Arnold Truyen, Conservation of wood - selected experiments, lecture at “Consolidation, reinforcement and stabilisation of decorated wooden artifacts”, WoodCultHer COST ACTION IE0601, Prague 30th–31st March 2009.
S185
hard to use the experience from the industry in the field of cultural heritage (where thorough penetration is necessary). Even though PF has been used for a century, the exact relationship between different reaction sites and ratio of reactants is still being discovered. It has been found that hexamethylene tetramine (HMTA), a common cross-linker, preferably reacts with a PF with a high ratio of ortholinkages. A slight surplus of phenol (1.2 to 1.3 phenol molecules per formaldehyde molecule) promotes this orientation [24]. Thus it may be possible to affect the structure in similar ways to promote greater compatibility with the wooden matrix. PF has been used in conservation of metals [25], fossils, and wood [26] but often at high temperature and with less satisfactory results than urea-formaldehyde [27]. It can be difficult to find references for PF treatment of wooden objects and problems with too fast curing times (despite excellent penetration), possible emission of formaldehyde, and above all the irreversibility of the curing process caused the method to be discontinued for treatment of wooden artefacts during the 1940s [5,26]. Thus it must probably be modified in order to ensure a better treatment. The lack of published information further makes it difficult to evaluate how effective the method is and how recent discoveries in phenol-formaldehyde research can be applied to the field of cultural heritage. Our own studies indicate that the PF penetrates wooden artefacts thoroughly and can be polymerised in situ at very low pH, resulting in open structures as long as the wood is fairly intact. The very degraded alum-treated wood from the Oseberg find, however, cannot be safely treated as the alum crystals dissolve in the watery pre-polymer mix. The finds are currently so fragile that they are held together by a thin ‘crust’ of alum crystals, lacquer and some remaining wood. If the alum salts will be dissolved the object may disintegrate. A more thorough report on PF for this use will be published elsewhere [28]. Other molecules than unmodified phenol can be incorporated into the structure of PFs, for example bisphenol A [29] or sucrose and similar carbohydrates (with xylitol giving the best result) [30]. Phenol can even be entirely substituted with furan [31], although our own tests have shown that pure furfuryl alcohol or furan blackens during self-polymerisation and is so extremely exothermic that it will likely damage the wood. It has also been proven that lignin can be incorporated into the PF structure [32] or even react directly with formaldehyde without the use of phenol [33]. The main goal is to introduce a polymer in such a way that it coats the remaining cell wall material or forms a ‘spider web structure’ which strengthens the lignin without filling the object completely. Various ‘fillers’ are also considered–including lignin as well as inorganic materials like calcium carbonate. Polymerisation in buffer solutions can be used to both control the rate of curing of the polymer and even neutralise the very acidic alum-treated wood from Oseberg and similarly affected artefacts. One possibility is to introduce aliphatic chains between oligomer PF molecules (for example about 6–10 phenol molecules per oligomer) to ensure some flexibility so the polymer can swell and shrink with the wood. Such a polymer would be similar to wood in structure and resistant to most solvents (due to the phenolic groups), while maintaining enough flexibility to swell with the wood and avoid being as brittle as traditional phenolformaldehyde. Some ideas from materials science have already been implemented in treatment of non-wooden cultural heritage–such as using nanoparticles for the conservation and restoration of wall paintings. The small size and large surface areas of these particles allow for good penetration. In addition, inorganic materials can be reinforced with the exact same material as the artefact is made from [6]. The successful integration of Si-containing polymers [14]
S186
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
means that Si-based particles will likely also adhere fairly well to a wooden matrix. 5. Bio-inspired materials Another possibility is to learn from nature. An important field for this is biomimetics, materials inspired by natural designs. Such materials are often ‘green’ (meaning their components and byproducts have very low toxicity) and easily form hybrid materials where individual parts have very different qualities. A key feature in most natural materials is a hierarchical structure which has been optimised to provide high strength and/or flexibility through porous structures. Several examples are discussed in Ref. [34]. Such qualities are very desirable when designing materials to be used for the conservation of wooden artefacts. It may even be possible to implement useful qualities of other plants. One of the most famous–and well-studied–examples is probably self-cleaning surfaces on plants which remain free from water and dirt due to its microstructure. Several examples are given in Ref. [35] (the most famous of which is probably the lotus plant). Biomineralisation, the process by which biological organisms produce minerals (such as the shells on molluscs and algae, or skeletal parts of vertebrates), has been used to grow artificial bone material [36]–and even used to create highly durable non-organic ‘copies’ of wood [37]. In the study in question, silicon carbide strings penetrated into wood cells to form a heat-resistant ‘copy’ upon calcination. For conservation of wooden items, however, biomimetic materials utilising wood-like components are also extremely interesting as they most likely interact well with degraded wood. This might make it possible to ‘grow’ a bionic self-organising ‘skeleton’ framework inside the degraded wood–possibly combined with inorganic particles for support or acid neutralisation. Hydroxides might be used to neutralise acid generated inside the conserved objects–this has already been tested on wood from the Vasa [38]. Initial experiments on in situ ‘growth’ of acidneutralising calcium carbonate on archaeological wood (at the Max Planck Institute for Colloids and Interfaces in Potsdam, Germany), seem promising–even though the tested technique might be too alkaline to be safely applied to unearthed objects. Most probably, nanoparticles will be incorporated into degraded wood in the future and thus help save wooden cultural heritage as either bulking agents, reinforcement points for a polymer network, or reservoirs of acid-neutralising material. 6. Cellulose and chitosan Cellulose is the most common renewable polymer in nature, making up 40–50% of wood and 90% of cotton fibre, with more than 7.5 × 1010 tons produced yearly [39,40]. In addition to wood and other plants (particularly hemp, flax, jute, ramie, and cotton) bacteria and tunicates may also produce cellulose [41]. Cellulose is very stable due to a network of hydrogen bonds which prevents melting and dissolution with many solvents [41]. Adjacent cellulose chains can thus fit together in crystalline regions. Crystallinity thus affects swelling and absorbtion of the fibres. The less dense amorphous regions, however, can be dissolved using acid (the type, temperature and treatment time affects the properties of the finished product). The remaining cellulose crystals are often referred to as ‘whiskers’ although crystalline cellulose is known by many names such as ‘cellulose nanocrystals’, ‘microcrystals’, ‘nanocrystals/particles’, ‘microcrystallites’, and ‘nanofibres’ [39,40]. Technically, ‘whiskers’ refer to rod-like nanoparticles while ‘nanofibrils’ should be used to describe long flexible nanoparticles with both amorphous and crystalline regions [41]. Most whiskers from plant sources have lengths of 100–300 nm while those from
tunicates or bacteria can be up to 3000 nm long. Diameters range from 3 to 70 nm with typical values around 15–20 nm [40]. While this makes the individual particles significantly bigger than the monomers used in the phenol-formaldehyde experiments, they are of comparable size to nanoparticles roughly 100 nm in diameter used to treat wood from the Vasa [38]. Cellulose was first isolated in 1832 and has been thoroughly studied since then. In the 1950s, it was reported that cellulose fibres can make stable suspensions in water after degradation with sulphuric acid [40]. This acid, in particular, can produce whiskers which do not flocculate due to electrostatic repulsion, and sulphur content can therefore be used as a measure of properties. H2 SO4 treated fibres contain some weak acids (sulphate ester and carboxyl groups) which are less common in fibres treated with HCl [39]. The residues are not acidic enough to threaten the durability of the wood and even allow for further chemical modification of the fibres (for example, allowing them to be incorporated into polymer films). The whiskers have varying degrees of polymerisation and lengths depending upon their origin. Those from wood are about 180–200 nm with those from cotton are 100–120 nm and those from tunicates can reach 1000 nm in length [39]. This means that whiskers from plants are better than those from other sources when treating archaeological wood as the dimensions of the fibres should be kept small to enhance penetration. The actual preparation is simple. Fibrous cellulose (SigmaAldrich) was treated in 67% H2 SO4 : 5 g cellulose in 100 g acid was heated to 60 ◦ C for 1 h under heavy stirring. The remaining mix was centrifuged at 10,000 RPM for 10 min. The supernatant was discarded and replaced with deionised water. This process was repeated at least five times until a turbid supernatant phase appeared. This was collected and ultrasonicated using a tip sonicator at 30% power output for five minutes to disperse the whiskers. It was observed that using metal (aluminium) spoons when handling the cellulose caused it to degrade upon immersion in acid, turning the mix dark brown and ruining the whiskers. After penetration, the samples were cut and observed using [environmental] scanning electron microscopy ([E]SEM). Imaging was initially performed on an FEI FE-ESEM, Quanta 600 (FEI Company, Oregon, USA) in low vacuum mode using a large field detector to minimise sample preparation. Later on, a 5 nm Au/Pd coating was applied to the samples before imaging on a JEOL JSM-7500F FE-SEM (Cryo-SEM) using a secondary electron in-lens detector (JEOL, Eching, Germany). Acceleration voltages used are noted on the individual images. Initial tests of such suspensions showed that the cellulose did not penetrate thoroughly into the wood–although it did adhere well to the surface of the wood (as seen in Fig. 1). Even though the appearance was unacceptable for display, it proved that cellulose can act as a gap filler and form the desired open structure. At high
Fig. 1. Freeze-dried Viking Age wood treated with cellulose whiskers. Note that the cellulose adheres to the surface of the wood and forms an open structure.
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
S187
Fig. 2. Cryo-SEM (scanning electron microscopy) image of freeze-dried gold/palladium-coated cellulose whiskers. Note that individual whiskers can be seen in the conglomerate.
magnification, it is possible to see the individual fibres in a SEM at low acceleration voltages (Fig. 2), and shows their dimensions to be roughly 15 nm in diameter and 200–300 nm in length. At a certain concentration, the whiskers form an ordered (chiral nematic) phase. The structure tends to be retained even if the fibres are covered in surfactants. Salt (NaCl) concentrations as low as 10−5 M may screen the electrostatic interactions [39]. The equilibrium between isotropic and anisotropic phases is sensitive to the presence of electrolytes and the kind of counterions used as well as the sulphate group content [40]. When treating archaeological wood, however, these things mean that the ions in the wood will cause the cellulose whiskers to flocculate–preventing proper penetration. An example is seen in Fig. 3, which shows uneven dispersion of the whiskers. Due to the hydroxyl groups on the surface of the whiskers, both chemical modification and various surfactants can affect the whiskers. Actual polymer grafting onto the surface of the whiskers has been successfully implemented [40]. In addition to ionic species
Fig. 3. Whiskers on the surface os an impregnated piece of Viking Age wood (right). Note the relatively poor dispersion in the wood structure itself (left).
Fig. 4. ESEM image of Viking Age wood treated with cellulose whiskers and PEG1000. Note the improved penetration due to the effects of the surfactant.
grafted during acid-treatment, whiskers may also be kept in suspension by steric repulsion between POE chains grafted onto the rods. Surfactants may be used to obtain dispersions in organic solvents as well [42]. The surfactant molecules adhere to the surface of the whiskers, leaving molecular chains ‘drifting’ in all directions, and thus physically prevent whiskers from sticking together. Use of non-ionic surfactants is especially attractive as they allow whiskers to remain suspended even in water with a high ion concentration where the whiskers otherwise flocculate. Initial tests with cellulose using PEG1000 as a stabiliser showed that it significantly improved penetration (Fig. 4). In the future, various other surfactants should be tested to see which ones provide the best effect. Cellulose also provides high axial stiffness, which is particularly advantageous when using fibres to reinforce other polymers. Even adding water can dramatically alter the modulus of cellulose composites [41]. This means that it is important to keep the cellulose in museums dry. Cellulose nanowhiskers have thus been incorporated as fillers into a significant number of polymers [40]. Upon successful introduction, cellulose whiskers prevented cracking in urea formaldehyde (UF) polymer bond lines which is frequent in UF without cellulose [41]. Cellulose whiskers have been combined with another tree component as part of a triblock copolymer. The plant matter used was xyloglucan oligosaccharides from tamarind kernels and the polymer a xyloglucan-oligosaccharide-polyethylene oxide-polystyrene (XGO-POE-PS) triblock copolymer. A diblock copolymer of XGO and PS was tested but precipitated and could not stabilise the whiskers in watery dispersions [43]. This is, to the best of our knowledge, the only attempt so far to approach a true wood structure by mixing cellulose whiskers with components which resemble hemicelluloses and attempts to incorporate other wood-like structures may be ideal for conservation purposes. Cellulose and cellulose whiskers have been modified for use in conservation science with the specific goal of creating a waterdispersible polymer which could stabilise archaeological wood [44]. Allylic groups were grafted onto the cellulose in order to allow the polymer to cross-link at 60 ◦ C. This stabilised the wood but made it difficult to remove the polymer. Carboxymethyl groups were introduced in order to make the cellulose dispersible in water.
S188
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
This also allowed the consolidant to be washed out of the treated wood. A test piece of archaeological wood was treated for 20 days but the dispersion did not fully penetrate the wood during this time. Chitin is the polysaccharide 2-acetamido-2-deoxy--d-glucose and is identical to cellulose with a hydroxyl group replaced by an acetamido group. Chitosan is a derivative of chitin produced by alkaline deacetylation and can be obtained from crab or shrimp shells as well as mycelia. It is more biocompatible and biodegradable than typical synthetic polymers [45]. Chitosan has been investigated for drug delivery due to its low toxicity and biodegradability and may promote tissue growth and wound healing [45,46]. The same qualities, however, are beneficial for conservation science as it allows chitosan to be handled with minimal precautions. Since chitin and chitosan are highly basic, they differ from most polysaccharides (like cellulose, dextran or pectin) in nature. Chitosan can chelate metal ions and this may be utilised in wastewater treatment (both industrial and otherwise) [45]. The degree of chelation is influenced by the pH of the solution and concentration of ions as well as the size of the chitosan particles. The efficiency increases with increasing pH but takes place to some extent at pH levels at least as low as 2 [47]. This makes chitosan potentially viable for immobilising metal ions in archaeological wood since treatment times are usually quite long. Ideally, it allows chitosan to both reinforce the wood and stop metal ions in it, thus reducing the catalytic effect which said ions have on many degradation processes. The pH of the dissolved chitosan affects its physical properties. The viscosity of chitosan solutions is at its lowest at pH 3–4 and rises on either side of this window [48]. As such this is the ideal pH for penetrating chitosan throughout archaeological wood. Chitosan is soluble in monovalent acids but not divalent ones. Acid concentrations of 0.1–0.15 M can dissolve 2%w /v chitosan but generally not 2.5%w /v (out of the tested acids, only chloroacetic acid could dissolve this higher concentration). Using HCl leads to a less viscous system than using carboxylic acids [46]. In order to test penetration, solutions of 2% (w/w) chitosan in 0.1 M acids were made. It was found that such chitosan solutions are very viscous and may have trouble penetrating archaeological wood. Concentrations of about 1% chitosan were much less viscous while still
offering acceptable support to the deteriorated wood. The chitosan had a foam-like structure after freeze-drying and treated pieces felt notably stronger than those treated with cellulose whiskers. Storage tests indicated that chitosan kept in hydrochloric acid may begin to aggregate in less than 2 weeks, whereas this did not happen to chitosan dissolved in acetic acid. Since acetic acid is likely to cause less damage than hydrochloric acid to archaeological wood, it is recommendable to use this weaker acid even it is not an ideal solvent in a museum environment as it reacts with metals. However, it should be possible to raise pH to about 6 after penetration, either through buffers or simply by diluting with water. This would not only help remove excess acid but might also force the chitosan to precipitate inside the objects and thus further experiments should be conducted to determine if this would be beneficial. Adding salt to chitosan solutions has much the same effect as increasing surfactant concentration. This is because salt reduces the hydrodynamic volume of the molecules for both the hydrophobically modified and pure chitosan. In fact surfactant becomes much less important in solutions which contain 10 mM NaCl [49]. Since it is effectively impossible to remove salts from archaeological wood prior to impregnation, this effect might help to further disperse the chitosan inside the degraded wood, rendering surfactants superfluous. A chitosan-cellulose film has been made by dipping a quartz plate into alternating solutions of the two polymers [50]. Note, however, that cellulose and chitosan stick together because they have opposite charges at pH values around 3. This means that the molecules will flocculate rapidly if mixed and thus prevents archaeological wood from being treated with both polymers simultaneously. Still, it might be possible to build composite films inside the wood by immersing it in alternating solutions/dispersions of chitosan/cellulose and such films may have better mechanical properties than a layer of one of the polymers. 7. Concluding remarks Advantages and disadvantages of various proposed methods are presented in Table 1.
Table 1 Overview of various possible methods for treating archaeological wood. Consolidant
Advantages
Disadvantages
Phenol- or melamine-formaldehyde
Phenol structure similar to lignin Penetrates easily in water Real-time ageing tests (PF) Very hard and durable Resistant to solvents/pH
Fills the wood as a ‘block’ of plastic Irreversible/Unremoveable Formaldehyde is toxic Literature only on high temperature/pressure Rather brittle
Lignin
Wooden structure Cheap Can be part of polymer
Large molecules cannot penetrate Probably irreversible Dark colour
Nanoparticles
Commercially available May neutralise acid Can be coated to adhere better Can be very durable Compatible with wood structure Resistant to solvents/pH
Difficult to disperse and penetrate May become too alkaline for wood Stabilising effect is unknown
Alifatic polymer
Flexible In situ polymerisation possible
Insufficient stabilisation Difficult penetration if not in situ Not completely reversible
Biomineralisation
Fits/fills structure “Built” in situ Re-treatable
Sensitive to pH Possibly incompatible with wood Not tested thoroughly
Biomimetics
Compatible with wood Usually non-toxic Water is an ideal solvent Easily forms open structures May entrap some ions (chitosan)
May require additional reinforcement Surfactants/acids required for dispersion Can be pH-sensitive
Nanotubes
Very hard to penetrate Black colouration
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
No matter which route is used when designing better consolidants for archaeological wood, there are many requirements for a stabilising agent: The consolidant must be workable in a museum environment (at normal temperature, pressure, and without emitting toxic fumes). The material cannot expand, contract or otherwise warp the wood. Likewise, it must have a slightly acidic pH value suitable for wood. It must be able to penetrate the wood thoroughly and last for a very long time. As mentioned, the most important aspect is the ‘open’ structure which must be possessed by the bulking material in order to ensure that retreatment is possible. When combined, these diverse disciplines may provide new ways of looking at the consolidation of archaeological wood. Such a combination might entail forming composite materials where nanoparticles fill the bulk of the wood while being held in place by polymer material which adheres to both the particles and the existing wooden structure. This may be done as a kind of ‘spider web’ which envelops the wood without reacting with it, or a ‘copolymer’ system which bonds chemically to the leftover lignin. Existing designs in nature may help as materials like cellulose and chitosan might both have beneficial properties–especially if combined with other polymers to add strength and durability as well as nanoparticles which serve to neutralise and acid generated in the wood. It is hoped that these ideas and requirements can help crossdisciplinary collaboration and encourage people working with wood–be they chemists, wood technologists, or conservators–to think along new lines in order to secure our cultural heritage. Acknowledgements We would like to thank Susan Braovac from the Museum for Cultural History, University of Oslo, for her help in elucidating the current state of the Oseberg wood and its previous treatments. Poul Jensen and Yvonne Shashoua, both from the National Museum of Denmark, Department of Conservation, are gratefully acknowledged for their information and advice on the PEG treatment and the degradation of polymers, respectively. Jaan Roots from the Department of Chemistry, University of Oslo, is thanked for his helpful discussions on polymer design. The COST Action FP0802, Experimental and Computational Micro-Characterization Techniques in Wood Mechanics, is thanked for a grant to perform a short-term scientific mission at the Max Planck Institute for Colloids and Interfaces in Potsdam, Germany, where Ingo Burgert, Friederike Saxe, Anayancy Osorio, and Susann Weichold from the Biomimetics Plant group were of invaluable help when initiating experiments on cellulose whiskers and recording SEM images while Kyu-Bock Lee initiated experiments on in situ growth of calcium carbonate. References [1] Code of Ethics and Guidelines for Practice, http://www.conservationus.org/index.cfm?fuseaction=page.viewPage&pageID;=858&nodeID=1 (14/82009). [2] A.M. Rosenqvist, The stabilizing of wood found in the Viking Ship of Oseberg–part I and II, Stud Conserv 4 (1) (1959) 13–22, 62-72. [3] A.V. Brøgger, H.G. Falk, H. Schetelig, Osebergfundet, I, Published by the Norwegian state, Kristiania, 1917. [4] The Oseberg find 100-year anniversary–the great adventure in Norwegian archaeology. http://www.khm.uio.no/utstillinger/oseberg/indexE.html (18/08-2009). [5] A. Unger, A.P. Schniewind, W. Unger, Conservation of wood artifacts, SpringerVerlag, Berlin Heidelberg, 2001. [6] P. Baglioni, R. Giorgi, Soft and hard nanomaterials for restoration and conservation of cultural heritage, Soft Matter 2 (2006) 293–303. [7] M. Favaro, R. Mendichi, F. Ossola, U. Russo, S. Simon, P. Tomasin, et al., Evaluation of polymers for conservation treatments of outdoor exposed stone monuments. Part I: photo-oxidative weathering, Polym Degrad Stabil 91 (2006) 3083–3096.
S189
[8] P. Hoffmann, On the efficiency of stabilisation methods for large waterlogged wooden objects, and on how to choose a method, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 323–350. [9] S. Braovac, H. Kutzke, The presence of sulphuric acid in alum-conserved wood – Origin and consequences, contribution at “Wood Science for Conservation of Cultural Heritage”, WoodCultHer COST Action IE0601, Hamburg, 7th–9th October, 2009 (see also the paper by Braovac and Kutze in current issue). [10] R.M. Rowell, R.J. Barbour (Eds.), Archaeological wood: properties, chemistry, and preservation, American Chemical Society, Washington DC, USA, 1990. [11] P. Jensen, S. Schnell, The implication of using low molecular weight PEG in impregnation of waterlogged archaeological wood prior to freeze-drying, in: P. Hoffmann, K. Strætkvern, J.A. Spriggs, D. Gregory (Eds.), Proceedings of the 9th ICOM Group on Wet Organic Archaeological Materials Conference, Copenhagen, 2004, pp. 279–307. [12] G. Almkvist, I. Persson, Iron catalysed degradation processes in the Vasa, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 499–505. [13] M. Wittköpper, Der aktuelle Stand der Konservierung archäologischer Naßhölzer mit Melamin/Aminoharzen am Römisch-Germanischen Zentralmuseum, Arbeitsblätter für Restauratoren Gruppe 8: Holz Heft 2, (1998), 277–282. [14] C. Wayne Smith, Archaeological conservation using polymers: practical applications for organic artifact stabilization, Texas A&M University Press, College Station, USA, 2003. [15] G. Chaumat, C. Albino, L. Blanc, Development of new consolidation treatments from fatty acid resin solution, in: Program and abstracts of papers and posters ICOM–WOAM Conference, Amsterdam, 2007. [16] M.P. Colombini, M. Orlandi, F. Modugno, E.-L. Tolppa, M. Sardelli, L. Zoila, et al., Archaeological wood characterisation by PY/GC/MS, GC/MS, NMR and GPC techniques, Microchem J 85 (2007) 164–173. [17] G. Shulga, T. Betkers, V. Shakels, B. Neiberte, A. Verovkins, J. Brovkina, et al., Effect of the modification of lignocellulostic materials with a lignin-polymer complex on their mulching properties, BioResources 2 (4) (2007) 572–582. [18] N. Terashima, R.H. Atalla, S.A. Ralph, L.L. Landucci, C. Litpierre, B. Monties, New preparations of lignin polymer models under conditions that approximate cell wall lignification, Holtzforschung 50 (1996) 9–14. [19] M. Micic, M. Jeremic, K. Radotic, M. Mavers, R.M. Leblanc, Visualization of artificial lignin supramolecular structures, Scanning 22 (2000) 288–294. [20] Y. Nagamatsu, M. Funaoka, Design of recyclable matrixes from lignin-based polymer, Green Chem 5 (2003) 595–601. [21] T. Kataoka, Y. Kurimoto, Conservation of archaeological waterlogged wood by lignophenol, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 315–322. [22] D. Crespy, M. Bozonnet, M. Meier, 100 years of BakeliteTM , the material of a 1000 uses, Angew Chem Int Edit 47 (2008) 3322–3328. [23] J. Monni, L. Alvila, T.T. Pakkanen, Structural and physical changes in phenolformaldehyde resol resin, as a function of the degree of condensation of the resol solution, Ind Eng Chem Res 46 (2007) 6916–6924. [24] J. Huang, M. Xu, Q. Ge, M. Lin, Q. Lin, Y. Chen, et al., Controlled synthesis of highortho-substitution phenol-formaldehyde resins, J Appl Polym Sci 97 (2005) 652–658. [25] H.M. Sleicher, J.L. Everhart, Mounting of small metallographic specimens and metal powders in BakeliteTM , Metals Alloys 5 (1934) 59–60. [26] M. Ventikou, Old treatment, new problem: BakeliteTM as a consolidant, V & A Conserv. J. 32 (1999) 5–7, The paper is accessible at the time of writing at: http://www.vam.ac.uk/res cons/conservation/journal/journal32/ oldtreatment32/index.html. [27] L. Biek, J.W. Anstee, E.S. Cripps, A wooden bucket restored, Museums J. 57 (1958) 257–261. [28] M. Christensen, F.K. Hansen, H. Kutzke, Phenol-formaldehyde revisited–novolac resins for the treatment of degraded archaeological wood. To be published. [29] Y. Liu, C. Zhang, Z. Du, H. Li, Preparation and curing kinetics of bisphenol A Type novolac epoxy resins, J Appl Polym Sci 99 (2006) 858–868. [30] A.H. Conner, B.H. River, L.F. Lorenz, Carbohydrate modified phenolformaldehyde resins, J Wood Chem Technol 6 (4) (1986) 519–613. [31] E. Kumpinsky, pH effects on phenol-formaldehyde runaway reactions, Ind Eng Chem Res 34 (9) (1995) 3096–3101. [32] M. Turunen, L. Alvila, T.T. Pakkanen, J. Rainio, Modification of phenolformaldehyde resol resins by lignin, starch, and urea, J Appl Polym Sci 88 (2003) 582–588. [33] A.Y. Kharade, D.D. Kale, Effect of lignin on phenolic novolak resins and moulding powder, Eur Polym J 34 (2) (1998) 201–205. [34] P. Fratzl, R. Weinkamer, Nature’s hierarchical materials, Prog Mater Sci 52 (2007) 1263–1334. [35] K. Koch, B. Bhushan, W. Barthlott, Diversity of structure, morphology and wetting of plant surfaces, Soft Matter 4 (2008) 1943–1963. [36] A.C. Thackray, R.L. Sammons, L.E. Macaskie, P. Yong, H. Lugg, P.M. Marquis, Bacterial biosynthesis of a calcium phosphate bone-substitute material, J Mater Sci-Mater M 15 (2004) 403–406. [37] Y. Shin, G.J. Exarhos, Conversion of cellulose materials into nanostructured ceramics by biomineralization, Cellulose 14 (2007) 269–279.
S190
M. Christensen et al. / Journal of Cultural Heritage 13S (2012) S183–S190
[38] R. Giorgi, D. Chelazzi, P. Bagleoni, Nanoscience contribution to the preservation of acidic shipwrecks, in: D.J. Huisman, K. Strætkvern (Eds.), Proceedings of the 10th ICOM Group on Wet Organic Archaeological Materials Conference, Drukkeri Stampij, Amersfoot, Amsterdam, 2009, pp. 525–534. [39] M.M. de Sousa Lima, R. Borsali, Rodlike cellulose microcrystals: structure, properties, and applications, Macromol Rapid Comm 25 (2004) 771–787. [40] Y. Habibi, L.A. Lucia, O.J. Rojas, Cellulose nanocrystals: chemistry self-assembly, and applications, Chem Rev 110 (2010) 3479–3500. [41] S.J. Eichhorn, A. Dufresne, M. Aranguren, N.E. Marcovich, J.R. Capadona, S.J. Rowan, et al., Review: current international research into cellulose nanofibres and nanocomposites, J Mater Sci 45 (2010) 1–33. [42] C. Bonini, L. Heux, J.-Y. Cavaillé, P. Lindner, C. Dewhurst, P. Terech, Rodlike cellulose whiskers coated with surfactant: a small-angle neutron scattering characterization, Langmuir 18 (2002) 3311–3314. [43] Q. Zhou, H. Brumer, T.T. Teeri, Self-organization of cellulose nanocrystals adsorbed with xyloglucan oligosaccharide-poly(ethylene glycol)-polystyrene triblock copolymer, Macromolecules 42 (2009) 5430–5432. [44] G. Cipriani, A. Salvini, P. Baglioni, E. Bucciarelli, Cellulose as a renewable resource for the synthesis of wood consolidants, J Appl Polym Sci 118 (2010) 2939–2950.
[45] M.N.V.R. Kumar, A review of chitosan applications, React Funct Polym 46 (2000) 1–27. [46] Q.S. Zhao, X.J. Cheng, Q.X. Ji, C.Z. Kang, X.G. Chen, Effect of organic and inorganic acids on chitosan/glycerophosphate thermosensitive hydrogel, J Sol-Gel Sci Techn 50 (2009) 111–118. [47] A. Burke, E. Yilmaz, N. Hasirci, O. Yilmaz, Iron(III) ion removal from solution through adsorption on chitosan, J Appl Polym Sci 84 (2002) 1185–1192. [48] A.-L. Kjøniksen, B. Nyström, T. Nakken, O. Palmgren, T. Tande, Effect of surfactant concentration, pH, and shear rate on the rheological properties of aqueous systems of a hydrophobically modified chitosan and its unmodified analogue, Polym Bull 38 (1997) 71–79. [49] A.L. Kjøniksen, B. Nyström, C. Iversen, T. Nakken, O. Palmgren, T. Tande, Viscosity of dilute aqueous solutions of hydrophobically modified chitosan and its unmodified analogue at different conditions of salt and surfactant concentrations, Langmuir 13 (1997) 4948–4952. [50] G.H. Vikhoreva, N.R. Kil’deeva, I.N. Gorbacheva, E.A. Shablykova, S.Z. Rogovina, T.A. Akopova, Study of cellulose-chitosan composites. Solid-phase modification, rheology, films, Fibre Chem+ 32 (6) (2000) 402–406.