Evaluation of two terpene-derived polymers as consolidants for archaeological wood

doi:10.21203/rs.3.rs-2238727/v1

Download PDF

Article

Evaluation of two terpene-derived polymers as consolidants for archaeological wood

https://doi.org/10.21203/rs.3.rs-2238727/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 04 Mar, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

The evaluation of two terpene-derived polymers, termed TPA6 and TPA7, as possible consolidants for archaeological wood was carried out. The overall objective of this work was to expand the non-aqueous treatment toolkit which is available for the conservation of the highly degraded Oseberg collection. The wood artefacts which were found on the Oseberg ship were treated with alum in the early 20th century, leading to the formation of sulfuric acid and to the precarious state that they are in today. Some of these artefacts cannot be treated with conventional aqueous consolidants, like polyethylene glycol, due to their highly degraded and/or reconstructed nature. This study sought to examine the level of penetration of the polymers in archaeological wood and to evaluate their consolidative effect. Both TPA6 and TPA7 were soluble in isopropanol and had a M_w of 3.9 and 4.2 kDa respectively. A number of archaeological wood specimens were immersed in solutions of these polymers. Their penetration and effects were evaluated using weight and dimensional change, colour change, infrared spectroscopy, scanning electron microscopy and hardness tests. Both polymers successfully penetrated the wood specimens, with a higher concentration found on the surface versus the core. Additionally, both polymers appeared to increase the hardness of the specimen surfaces. Increasing the polymer concentration and soaking time in future investigations could potentially facilitate the penetration to the wood cores.

Physical sciences/Materials science

Biological sciences/Biotechnology

Biological sciences/Biotechnology/Biomaterials

Treating wooden artefacts with hot alum solutions was a popular method amongst conservators from the mid-1800s all the way to the 1950s¹. This was the method that was used on the artefacts found in the Oseberg ship in the early 1900s. Unfortunately, these well-meaning conservators could not predict the damage that this treatment would eventually cause. We now know that heating alum solutions generates sulfuric acid, leading to the degraded state that these artefacts are in today².

In Northern Europe archaeological wood is usually found in a waterlogged state and therefore water-based conservation methods are the most effective to conserve it. The most popular conservation method usually involves treatment with polyethylene glycol (PEG) and subsequent freeze-drying. If the Oseberg artefacts are retreated using this method, they would first have to be immersed in water, to remove the alum salts and acid, before adding PEG³. Retreating the Oseberg artefacts is not such a simple matter however, due to the large variation of the wood’s preservation state. Some of the most highly degraded artefacts cannot withstand this type of aqueous treatment as it may result in the dissolution of the alum inside the wood, leading to total collapse. Additionally, some of the artefacts contain non-wood components which were added in past restorations such as metal rods, glue and plaster¹. For these objects, there is increased risk of damage if they are immersed in water.

The Saving Oseberg project was established by the University of Oslo to safeguard these artefacts and to investigate appropriate ways with which to treat them. One of the main aims of this endeavour was to develop new treatment materials and strategies in order to have alternatives to existing consolidants. A particular emphasis was put on the sustainability of the new materials, with the use of bioinspired polymers whenever possible. Examples of materials that have been investigated include lignin³, isoeugenol^4,5, chitosan^6,7 and aminocellulose⁸. Most recently, we have developed polymers derived from terpenes with functionalised hydroxylated moieties^9,10. In this work we wanted to further investigate the potential of two of these polymers, TPA6 and TPA7 (Fig. 1), as wood consolidants. The main aim of this investigation was to determine the extent to which these polymers penetrate wood after immersion in isopropanol and to determine if they provided any consolidative effect.

Preparation of wood samples

Cube specimens were made from waterlogged archaeological wood excavated in Oslo, Norway, and the propery of the Cultural History Museum. The wood, identified as Pinus (pine) with light microscopy, was initially in the form of a branch. This was sawn into six slices, with each slice then being cut into 2 x 2 x 2 cm³ cubes (Figure S1). These were freeze-dried and then acclimatised to 20°C and 50% RH. A total of 30 cubes of archaeological wood were used for these experiments: 20 for immersion tests with the polymer solutions and the remaining 10 for the two control groups. In addition, the leftover pieces of wood that remained after the cutting of the cubes were used for density and maximum moisture content measurements.

Naming System

The wood specimens were named in a way so that one could identify which slice they were cut from, and whether they were a surface or a core sample. Generally, the ID for each specimen was comprised of three characters: the first indicates their slice number (1–6), the second denotes whether they were cut from the surface or core (S = surface; C = core) and the last is a nominal specimen number. For each slice, eight cubes originated from the surface and two cubed from the core (Figure S1, right).

Density And Maximum Moisture Content Measurements

For these studies, 6 wedges from each of the six wood slices were used: 4 sawn from the surface from each slice, and 2 from the core. This was done to be able see whether the wood slices were degraded to different degrees and to observe whether there was any difference between the surface and the core of the wood. The density was determined by dividing the oven-dried weight by waterlogged volume. The waterlogged volume was determined from the weight of water displaced by the waterlogged samples. The maximum moisture content (MMC) percentage was also determined at the same time and is expressed as the weight of water contained in the sample relative to the oven-dried weight.

Immersion Tests With Polymer Solutions

The wood cubes were divided into two control groups and two treatment groups (Table 1). The cubes in control group 1 were freeze-dried only, while those in control group 2 were freeze-dried and then immersed in isopropanol. The specimens in treatment group 1 and 2 were immersed in polymer solutions of TPA6 and TPA7 respectively.

Table 1

The IDs of the wood specimens in each group used in this study
Control group 1 – freeze-dried only	Control group 2 – isopropanol	Treatment group 1 – TPA6	Treatment group 2 – TPA7
1.S.6	2.S.7	4.S.3	4.S.6
2.S.4	3.S.5	1.S.4	1.C.2
4.C.1	4.C.2	2.S.1	2.S.2
5.S.5	5.S.3	2.C.2	2.S.3
	6.S.6	3.S.1	3.S.2
		3.S.4	3.S.7
		4.S.2	4.S.5
		4.S.4	4.S.1
		5.S.1	5.S.2
		6.S.1	6.S.4

TPA6 is a homopolymer made from an α-pinene-derived monomer, with a M_w of 3.9 kDa. TPA7 is a copolymer derived from this same terpene monomer along with an oleic acid-based comonomer and has a M_w of 4.2 kDa. Both are yellow solids at room temperature. 50 g of each polymer was available for the wood testing. The two polymers were dissolved in isopropanol at a 15% w/w concentration. Both polymers appeared to dissolve completely in this concentration, therefore we assumed full dissolution. These solutions were then used to soak the assigned wood specimens. Control group 2, as well as both treatment groups were left immersed in their respective solutions for two weeks, based on a previous report of wood testing with non-aqueous treatments³. The wood cubes were then removed from the solutions and left to slowly air-dry in a fume-hood for one week, after which they were kept in climate chamber at 20°C and 50% RH.

Change in weight and dimensions

Weight was measured using a four decimal balance. The weight of the cubes was reported before (after freeze-drying and conditioning to 20°C and 50% RH) and after treatment (after drying and conditioning to 20°C and 50% RH).

Dimensions were measured with electronic callipers, using reference pins inserted into the cross-section of the cubes. Both the radial and the tangential directions were measured. The measurements were taken after freeze-drying and acclimatisation to 20.0°C and 50% RH and after the relevant treatment steps. The % weight change and % linear dimensional change were calculated for the treated groups using equations 1 and 2:

$$\text{\%} \text{w}\text{e}\text{i}\text{g}\text{h}\text{t} \text{c}\text{h}\text{a}\text{n}\text{g}\text{e} =\frac{(\text{w}\text{e}\text{i}\text{g}\text{h}\text{t} \text{a}\text{f}\text{t}\text{e}\text{r}-\text{w}\text{e}\text{i}\text{g}\text{h}\text{t} \text{b}\text{e}\text{f}\text{o}\text{r}\text{e})}{\text{w}\text{e}\text{i}\text{g}\text{h}\text{t} \text{b}\text{e}\text{f}\text{o}\text{r}\text{e}} \times 100\%$$

$$\text{\%} \text{l}\text{i}\text{n}\text{e}\text{a}\text{r} \text{d}\text{i}\text{m}\text{e}\text{n}\text{s}\text{i}\text{o}\text{n}\text{a}\text{l} \text{c}\text{h}\text{a}\text{n}\text{g}\text{e} = \frac{(\text{d}\text{i}\text{m}\text{e}\text{n}\text{s}\text{i}\text{o}\text{n} \text{a}\text{f}\text{t}\text{e}\text{r}-\text{d}\text{i}\text{m}\text{e}\text{n}\text{s}\text{i}\text{o}\text{n} \text{b}\text{e}\text{f}\text{o}\text{r}\text{e})}{\text{d}\text{i}\text{m}\text{e}\text{n}\text{s}\text{i}\text{o}\text{n} \text{b}\text{e}\text{f}\text{o}\text{r}\text{e}} \times 100\text{\%}$$

Colour Change

Colour measurements were taken of all treatment groups and control groups. This was carried out with a Konica Minolta CM-700d handheld spectrophotometer, making use of the CIELAB (L*a*b*) colour space. The change in colour caused by TPA6 and TPA7 was calculated from the difference of the measurements between the two treatment groups and control group 2 (wood treated with just isopropanol). The absolute change, ΔE*, was finally calculated from the coordinate measurements using Eq. 3:

$${\Delta }{\text{E}}^{*} = {({\Delta }{\text{L}}^{*2} + {\Delta }{\text{a}}^{\text{*}2} + {\Delta }{\text{b}}^{\text{*}2})}^{\frac{1}{2}}$$

Attenuated Total Reflection-fourier Transform Infrared Spectroscopy (Atr-ftir)

Fourier transform infrared spectroscopy was carried out using an attenuated total reflection mode on a Thermo Fischer FTIR spectrometer (Nicolet iS50), with range 4000–400 cm^− 1. The samples used for this part of the study included those from sound pine, an isopropanol-treated control, the treatment groups, as well as pure polymer samples. A resolution of 4 cm^− 1 was used and each spectrum was derived from 32 scans. Three spectra from each sampling site were taken and averaged. The wood samples were analysed at four points: the surface, the core and between the surface and core both along and across the grain (Figure S2). Thermo Scientific OMNIC FTIR software was used to analyse the data. Spectra were baseline corrected and normalised to 1508 cm^− 1 (wood spectra) or 1725 cm^− 1 (pure polymer spectra).

Scanning Electron Microscope (Sem) Analyses

A FEI Quanta 450 Scanning Electron Microscope with a voltage of 8 kV on low vacuum was used. Spot size was 4.5, chamber pressure was 50–80 Pa with working distances of 7.0–10.0 mm.

Resistance To Indentation (Hardness Test)

A fruit penetrometer (Lutron FR-5120) motorised with a TRINAMIC Motion control engine (Model TMCS-40.6.35-10000-AT-01) with 10000 P/R resolution was utilised (Figure S3). The penetrometer was fitted with a 3 mm radius tip which was used to apply a force on the tangential face of the wood specimens. The motor was operated through a hardwired controller. It was ensured that the wood specimens were lying on a flat surface and that the penetrometer’s tip touched the wood’s testing face perpendicularly. The testing faces of the wood specimens were shaved with a razor in order to ensure that they were flat and parallel. This was not always possible however, as a balance had to be struck between how much material was shaved off and the flatness of the testing faces. A specimen of sound pine was first measured. The wood specimens in the control groups were all measured. Eight specimens were measured from each treatment group. Each wood specimen was measured five times on its tangential surface and the values were then averaged (Figure S3). Additionally, each wood specimen from the treatment groups were split in two and the core was measured once on each half and averaged (Figure S3). Prior to each measurement, a preload was applied by lowering the penetrometer tip on to the testing face of the wood specimen until the force read by the sensor was between 6–10 N. For the actual measurement, the tip was lowered down 1 mm and the force in N was recorded. The net force was calculated by subtracting the preload force from the measured force. IBM® SPSS® software was used to analyse the data.

Density and maximum moisture content measurements

The density and MMC of wood are both important indicators of its degradation. Knowing the density of the wood can give us an idea of the amount of consolidant required to conserve it¹¹. The moisture content of waterlogged wood is referred to as MMC since it is assumed that wood which has been preserved in such conditions would possess the highest possible volume of water in its pores¹². The MMC increases and the basic density decreases with increasing wood degradation, which results from the loss of mass and volume of cell wall material due to bacterial degradation taking place during burial¹³.

From the density and MMC measurements, it appeared that the state of preservation of our wood branch was approximately even throughout. All the samples showed similar results, except for one outlier (sample 24) (Fig. 2). The density measurements ranged from 0.132 to 0.179 g/mL and the MMC from 492 to 653%. Wood with a high degree of degradation has a density of ~ 0.1 g/mL while sound wood usually has a density of ~ 0.4–0.5 g/mL¹¹. Well-preserved pine specifically is reported to have a density of around 0.5 g/mL³. With regards to MMC, a value of above 400% indicates that the wood is highly degraded¹⁴. Based on this, our wood samples seemed to have a high degree of degradation, with the means of the density and MMC being 0.146 g/mL and 610% respectively (Fig. 2). Moreover, the samples taken from the surface and the core of our wood branch appeared to show little difference in both density and MMC, meaning that they had similar preservation states.

Weight Change

The specimen groups treated with TPA6 and TPA7 were observed to substantially increase in weight (Fig. 3a). The control group treated with isopropanol (referred to as IPA in Fig. 3a) was expected to decrease in weight due to the dissolution of some pine resin in the solvent. During immersion, the solvent was observed to turn from transparent to a pale yellow, possibly indicating that some resin may have exited the wood. This was not found to be the case however since the specimens actually increased in weight, albeit very slightly. It is possible that these cubes were not conditioned to the same extent before and after isopropanol immersion, thus leading to a slight irregularity in the weight measurements.

Weight gain appeared to be uniform for the specimens treated with TPA6 and TPA7. TPA6 had a % weight change ranging from 40.5–45.9% while TPA7 ranged from 39.5–50.4%. The specimen with the most weight gain overall (50.4%) was a piece from the core treated with TPA7 (1.C.2). The sample showing the least weight gain (39.5%) was also from the TPA7 group (3.S.2). It is unclear why the polymer deviated in its penetration for these two particular specimens. Taking this into consideration, TPA6’s penetration into the wood appeared to be slightly more regular with no extremes. It is uncertain whether this is due to the polymer’s properties, or if it is an effect of the wood’s inherent variability.

Dimensional Change

For all groups, the dimensional change for both radial and tangential faces primarily ranged at ± 2%, except for a few outliers (Fig. 3b). Both radial and tangential faces seemed to change to the same degree, that is, one face did not predominate over the other. Control group 2 appeared to shrink after isopropanol treatment. The treatment groups had more varied results, showing both shrinkage (negative values) and swelling (positive values). TPA6 seemed to mainly cause shrinkage. TPA7 seemed to have a more variable effect, with swelling slightly predominating over shrinkage.

As mentioned, there were a few outliers in these measurements. The most extreme was 2.S.7 (control group 2), which showed a radial dimensional change of -8.7%. Other outliers include 4.C.2 (control group 2; tangential change of 3.1%), 6.S.1 (treatment group 1; tangential change of -3.2%) and 1.C.2 (treatment group 2; radial change of 4.2%). These measurements may potentially be attributed to the high variability of the wood or they may also be due to experimental error. The ‘pin method’ of measuring dimensional changes has limitations when it comes to the accuracy of the measurements. This is because the measurement values are highly dependent on the angle of the calliper relative to the pins. Therefore, if this angle is not exactly replicated during the ‘before’ and ‘after’ measurements, it will be difficult to get results that truly represent the dimensional change that has taken place.

With regards to control group 1 (freeze-dried only), the reference pins were inserted and measured directly after freeze-drying. These cubes were then kept in a desiccator at ~ 20.0°C and 50% RH during the soaking of the other groups. Its measurements were then repeated after the treatment of the other specimens was completed. This means that theoretically the specimens in control group 1 should have exhibited no dimensional changes since they were not subjected to any treatment after freeze-drying. Nevertheless, slight dimensional changes were still recorded. This demonstrates the inherent inaccuracy of this measuring method. The dimensional changes recorded in control group 2 may therefore be considered as the error range that can be expected for the other measurements ( ~ ± 2%). If we take this into consideration, then this means that most of wood specimens did not exhibit significant dimensional changes after treatment.

Figure S4 compares the % dimensional change to the % weight change in all groups and confirms the observations that were noted previously. The treatment groups gained weight after treatment due to polymer uptake. Conversely, the % dimensional changes did not appear to have varied greatly after treatment. Since isopropanol has a low surface tension, it was not surprising that we did not observe too much shrinkage during drying after the treatment steps.

Figure 3a. The % weight change for all the measured groups. Control group 1 was freeze-dried and kept in a desiccator at ~ 20°C and 50% RH during the immersion of control group 2 (isopropanol-immersed) and treatment groups 1 and 2 (TPA6- and TPA7-immersed respectively).

Figure 3b. A comparison of the radial and tangential dimensional % changes for control groups 1 and 2 and treatment groups 1 and 2.

Colour Change

Based on visual inspection alone, it was obvious that the polymers had caused a colour change in the wood specimens, mostly darkening them and imparting a yellow tinge (Fig. 4). This is possibly a result of the polymers imparting their colour to the wood. The spectrophotometer colour measurements of both treatment groups were compared to those of control group 2, which was treated with isopropanol (Fig. 5). Both polymers seemed to change the colour to a similar degree, as was observed visually. The measurements show that TPA7 seemed to cause slightly less colour change than TPA6, although this could not be seen by the naked eye. The largest degree of change was observed in the L* coordinate which decreased after the treatments, meaning that the specimens became darker. There was also a substantial difference in the b* axes, indicating a shift from blue to yellow. The measurements also showed a positive difference in the a* axes (green to red), although at a smaller degree. Figure 5 shows the ΔE values for the treatment groups, in relation to control group 2. Treatment groups 1 (TPA6) and 2 (TPA7) had a ΔE value of 11.55 and 10.55 respectively.

An ΔE above 3 can be seen by the human eye¹⁵. However, the colour change caused by these polymers may possibly not be detected by laypeople if applied to the alum-treated Oseberg collection, since the artefacts are already dark³.

Figure 4a. Photos of treatment group 1 (TPA6) before (top) and after (bottom) treatment. Specimens became noticeably darker, with a yellow tinge.

Figure 4b. Photos of treatment group 2 (TPA7) before (top) and after (bottom) treatment. Specimens were similar to those treated with TPA6.

Atr-ftir Analyses

The archaeological wood controls were first compared to sound wood, in order to better understand their state of degradation. Figure 6a shows such a comparison, with the use of sample 6.S.6 (isopropanol-immersed control) as the archaeological wood. Normalisation was carried out at the lignin peak (1508 cm^− 1). One could see that the cellulose peaks (1369, 1159 and 896 cm^− 1)^16,17 were muted in the archaeological wood, indicating its state of degradation. Similarly, the hemicellulose peaks at 1731 and 1239 cm^− 1 were barely visible¹⁶. Conversely, the lignin signals at 1595, 1263 and 1220 cm^− 1 were prominent in the archaeological wood, mainly due to their amplification as a result of the loss of cellulose and hemicellulose¹⁶. The band at 1024 cm^− 1 is due to a combination of several wood components (cellulose, hemicellulose and lignin)¹⁷ and as such, it shows the greatest absorption in both sound and archaeological pine.

Figure S5 shows a comparison of the IR absorbances of pure TPA6 and TPA7, normalised at 1725 cm^− 1 (C = O bond). The spectra of the two polymers were almost identical to one another. This was expected as they are structurally very similar. The major difference between the polymers was at 2856 cm^− 1, where TPA7 had a more pronounced peak. This could be attributed to a C-H stretching vibration, possibly due to the oleic acid component of the copolymer.

The absorbances of the two pure polymers were then compared to those of a control specimen. The spectra showed that both polymers were easily distinguishable from the wood, although some signals overlapped, such as at the 3500–3000 cm^− 1 region (Figure S6). The polymers had identifiable signals at 3000–2900 cm^− 1, as well as at 1716 and 1154 cm^− 1 (assigned as C = O and C-O respectively). Due to these differences, it was expected that the polymers would be easily distinguished from the wood in the treated samples.

Figure 6b & c. IR spectra comparison (fingerprint region) between the isopropanol-immersed control (red; 6.S.6) and the treated samples. (b) shows the samples treated with TPA6 (2.C.2) and (c) shows the TPA7-treated samples (2.S.2). The treated samples were measured at different depths: at the surface (purple), core (blue), between the surface and the core along the grain (green) and between the surface and the core across the grain (yellow). The signals depicting polymer penetration are annotated. Spectra were normalised at 1508 cm^− 1.

The wood specimens treated with the polymers were analysed at different depth points. This was done to be able to get an indication of the level of distribution of the polymer through the entire specimen. Samples were taken from the surface, core, between the surface and the core along the grain and between the surface and the core across the grain. Their spectra were then compared to those of an untreated, isopropanol-immersed control (Figs. 6b and 6c). The spectra were normalised at the lignin peak at 1508 cm^− 1 as this was shared by all the samples.

Figures 6b and 6c show that the specimen surface spectra were very different from all the other signals, clearly showing a high concentration of polymer. The rest of the treated wood specimens were more similar to the control, however they still showed a clear indication of polymer penetration. This was seen at 1725 and 914 cm^− 1 in particular. The spectra taken from the core of the treated specimen had the least absorbance. The sample taken from in between the surface and core absorbed less than the surface but more than the core. The sample taken along the grain had more absorbance than the one across the grain, as can be seen at both 1725 and 914 cm^− 1. This was somehow expected since it would be easier for the polymer to penetrate along the grain of the wood rather than across it. The decreasing degree of polymer penetration (both TPA6 and TPA7) could therefore be described as follows:

surface > between surface and core (along the grain) > between surface and core (across the grain) > core

Sem Analyses

The SEM samples were carefully shaved with a razor in order to get a surface which was flat, enabling better visualisation of the wood cells. It was ensured that not too much material was shaved off from the surface, in order not to compromise the results. Latewood was primarily looked at, as opposed to earlywood. This is because generally, morphological changes appear more prominently in latewood due to its thicker cell walls. Sound pine was first compared to an archaeological wood control. The sound pine image showed that the cells clearly had a more robust shape, with regular and thick cell walls. In contrast, the archaeological wood had cells which appeared ‘feathery’ and fragile, with very thin cell walls appearing to be mostly dislodged (Figure S7) This may be due to the loss of holocellulose (the term for combined cellulose and hemicellulose) previously depicted by the FTIR analyses.

Samples from wood specimens treated with both TPA6 and TPA7 (Fig. 7) were taken. For each specimen, a sample was taken from the surface and the core. It should be noted that generally, it is difficult to identify organic polymers such as TPA6 and TPA7 from other organic materials such as wood with SEM. The specimens treated with both polymers showed similar results. The polymer was clearly visible in the surface samples, in some cases clogging the wood pores. The polymer was more difficult to discern in the samples taken from the cores. It could be noted however that the treated cores appeared to be less airy than the control, with some cells having slightly thicker cell walls. They also seemed to be less deformed and appeared to be able to withstand the shaving that was carried out on the samples more than the control, possibly indicating an increased resilience. It was however difficult to say for certain whether this was due to the presence of the polymer.

In general, the SEM analyses appeared to correlate with the FTIR results. Both techniques showed that the highest concentration of polymer was on the surface of the wood specimens. FTIR indicated that the polymers managed to penetrate the cores of the specimens, albeit at a much lower concentration. This was difficult to observe with the SEM, but there were possible morphological changes that could perhaps indicate the presence of the polymers.

Hardness Test

The hardness tests were carried out in order to get an indication of the extent of consolidation that the polymers conferred. The fruit penetrometer measures the wood’s resistance to indentation. Specimens from control group 1 (freeze-dried only) and control group 2 (freeze-dried and immersed in isopropanol) were measured five times on the tangential surface. Specimens from both polymer treatment groups were measured five times on the tangential surface. In addition, each of the treated specimens was split in half and two measurements were taken from the core. The multiple measurements taken from every specimen were then averaged (Fig. 8).

From the averaged values, one could see that there seemed to be an increase in surface hardness compared to the controls. Conversely, the core hardness did not appear to change in the treated groups. Statistical analysis with the one-way ANOVA was carried out on the surface measurements of the treated groups in order to determine whether the observed change in hardness was significant. This was done using the individual measurements instead of the averaged values.

The control group used for the statistical analysis was made up of both the freeze-dried and isopropanol-treated controls. This was justified since the difference in means between the two groups was not significant (Figure S8). There were 40 values in the control group, 40 values in the TPA6 treatment group and 40 values in the TPA7 treatment group. There were two outliers in the TPA6 treatment group, as assessed by a boxplot. It was decided to not remove these outliers from the dataset. The data was normally distributed except for the TPA7 group, as assessed by the Shapiro-Wilk test (p > 0.05). However, the one-way ANOVA is known to be robust to deviations from normality, especially if the sample sizes are equal (as was the case here)¹⁸. The Levene's Test indicated that homogeneity of variances was violated (p = 0.025), therefore the Welch’s one-way ANOVA was used for this analysis. The surface hardness score was found to be statistically significantly different for the two groups treated with the different polymers (Welch's F(2, 1641.693) = 33.258, p < 0.001) (Figure S9). The surface hardness increased from the control (58.7 ± 6.8) to the TPA6 treatment group (69.8 ± 6.4) and the TPA7 treatment group (69.8 ± 8.9). Games-Howell post hoc analysis revealed that the increase from the control to the TPA6 treatment group (11.0875, 95% CI (7.561 to 14.614)) was statistically significant (p < 0.001), as well as the increase in surface hardness for the TPA7 treatment group (11.1050, 95% CI (6.855 to 15.355), p < 0.001). The difference between the groups treated with the different polymers was not statistically significant (0.0175, 95% CI (0.0 to 4.174), p = 1.00).

This analysis indicated that the polymers imparted a clear increase in surface hardness, possibly making the wood more resistant to stresses caused by handling or the environment. Both TPA6 and TPA7 seemed to increase the hardness to the same extent, as there was no statistical difference between them. Conversely, the cores of the treated samples did not appear to increase in hardness but this was expected, as there was only a small amount of polymer present compared to the surfaces.

This study aimed to determine the extent of penetration of two terpene-derived polymers, TPA6 and TPA7, in archaeological pine. This was carried out by monitoring weight and dimensional change, alongside the use of ATR-FTIR supplemented with SEM. Other tests were carried out to observe the effects that the polymers had on the wood, such as colour change measurements and hardness tests.

TPA6 and TPA7 showed very similar results and they seemed to affect the wood in mostly the same way. Generally, all the treated specimens increased in weight, indicating polymer uptake. Slight dimensional changes were also recorded for all specimens. The colour of the polymer-treated wood was slightly darker, with a yellowish tinge. FTIR showed that the polymers successfully penetrated the wood, with the highest concentration on the surface. This correlated well with the SEM analyses, which showed an abundance of polymer on the surface of the wood specimens. Both polymers also penetrated the core of the wood specimens, although to a lesser extent, making them difficult to pinpoint in the SEM images. Hardness tests proved that the surface of the treated specimens was significantly harder than the controls and therefore more resistant to indentation. The differences between the two polymer treatment groups were not significant, as was observed for all the tests that were carried out. Because of this, TPA6 should be favoured over TPA7 since its synthesis possesses fewer steps and is therefore less time-consuming.

The general scope of this study can be expanded in future work. A new, more accurate method of measuring dimensional change should be developed. A possible idea is to make markings on the wood with a pen before treatment, and then measure the dimensional changes with a technical ruler as opposed to callipers. Higher concentrations of polymer can be used to determine whether this would make a difference in the extent of penetration to the wood core. The immersion time can also be prolonged as this may also potentially increase the amount of polymer that manages to penetrate to the core. Other different methods of polymer administration can also be explored, such as injection. Additionally, these tests should be run using higher sample numbers in order to increase the robustness of these studies. Long-term studies should also be carried out to determine the degree of stability of the polymers over time. If the results from these tests prove to be promising then the polymers may pass into the next phase of investigation, where they’ll be tested on alum-treated Oseberg samples.

Acknowledgements

This work was supported by the UK Engineering & Physical Sciences Research Council [grant number EP/L015633/1], the Norwegian Ministry of Education and Research, and the University of Oslo as part of the Saving Oseberg project. M.C. is supported by a UK EPSRC Centre for Doctoral Training award.

Author Contributions

M.C. synthesised and characterised the polymers TPA6 and TPA7; the wood cubes were cut and prepared by S.B.; the evaluation and analyses of the wood was carried out by M.C. and S.B.; S.B., S.E.H, R.A.S. and S.M.H. provided supervision for the research; the paper was written by M.C. and edited by S.B. and S.H.; all other authors reviewed and agreed to the published article.

Additional Information

All procedures were conducted in accordance with the relevant guidelines. Archaeological Pinus wood was supplied by the Museum of Cultural History in Oslo and had been stored in a waterlogged state until use. Collection and use of this wood did not require permission.

The authors declare no competing financial interests.

Data availability

Raw data is available in Supplementary Information and on request from the Corresponding Authors.

Braovac, S. et al. Navigating conservation strategies: linking material research on alum-treated wood from the Oseberg collection to conservation decisions. Herit. Sci. 6, 1–16 (2018).
Braovac, S. & Kutzke, H. The presence of sulfuric acid in alum-conserved wood - Origin and consequences. J. Cult. Herit. 13, S203–S208 (2012).
Łucejko, J. J. et al. Evaluation of soda lignin from wheat straw/sarkanda grass as a potential future consolidant for archaeological wood. Forests 12, 1–28 (2021).
McHale, E. et al. Synthesis and characterisation of lignin-like oligomers as a bio-inspired consolidant for waterlogged archaeological wood. Pure Appl. Chem. 969–977 (2016). doi:10.1515/pac-2016-0814
McHale, E., Steindal, C. C., Kutzke, H., Benneche, T. & Harding, S. E. In situ polymerisation of isoeugenol as a green consolidation method for waterlogged archaeological wood. Sci. Rep. 7, 1–9 (2017).
Wakefield, J. M. K., Gillis, R. B., Adams, G. G., McQueen, C. M. A. & Harding, S. E. Controlled depolymerisation assessed by analytical ultracentrifugation of low molecular weight chitosan for use in archaeological conservation. Eur. Biophys. J. 47, 769–775 (2018).
Wakefield, J. M. K., Braovac, S., Kutzke, H., Stockman, R. A. & Harding, S. E. Tert-butyldimethylsilyl chitosan synthesis and characterization by analytical ultracentrifugation, for archaeological wood conservation. Eur. Biophys. J. 49, 781–789 (2020).
Wakefield, J. M. K. et al. Aminoethyl substitution enhances the self-assembly properties of an aminocellulose as a potential archaeological wood consolidant. Eur. Biophys. J. 49, 791–798 (2020).
Cutajar, M. et al. Terpene polyacrylate TPA5 shows favorable molecular hydrodynamic properties as a potential bioinspired archaeological wood consolidant. Sci. Rep. 11, 7343 (2021).
Cutajar, M. et al. Comparative hydrodynamic characterisation of two hydroxylated polymers based on α-pinene- or oleic acid-derived monomers for potential use as archaeological consolidants. Sci. Rep. 12, 18411 (2022).
Jensen, P. & Gregory, D. J. Selected physical parameters to characterize the state of preservation of waterlogged archaeological wood: A practical guide for their determination. J. Archaeol. Sci. 33, 551–559 (2006).
Macchioni, N., Pecoraro, E. & Pizzo, B. The measurement of maximum water content (MWC) on waterlogged archaeological wood: A comparison between three different methodologies. J. Cult. Herit. 30, 51–56 (2018).
Babiński, L., Izdebska-Mucha, D. & Waliszewska, B. Evaluation of the state of preservation of waterlogged archaeological wood based on its physical properties: Basic density vs. wood substance density. J. Archaeol. Sci. 46, 372–383 (2014).
Jingran, G., Jian, L., Jian, Q. & Menglin, G. Degradation assessment of waterlogged wood at Haimenkou site. Frat. ed Integrita Strutt. 30, 495–501 (2014).
Argyropoulos, D.S. · Wood and Cellulosic Chemistry 2nd Edition, edited by. Hon, D.N.S & Shiraishi, N. Marcel Dekker, New York (2021).
Schwanninger, M., Rodrigues, J. C., Pereira, H. & Hinterstoisser, B. Effects of short-time vibratory ball milling on the shape of FT-IR spectra of wood and cellulose. Vib. Spectrosc. 36, 23–40 (2004).
Mohebby, B. Attenuated total reflection infrared spectroscopy of white-rot decayed beech wood. Int. Biodeterior. Biodegrad. 55, 247–251 (2005).
Lix, L. M., Keselman, J. C. & Keselman, H.. J. Consequences of Assumption Violations Revisited : A Quantitative Review of Alternatives to the One-Way Analysis of Variance ‘F’ Test Author (s): Lisa M. Lix, Joanne C. Keselman and H. J. Keselman Published by : American Educational Research Assoc. Rev. Educ. Res. 66, 579–619 (1996).

No competing interests reported.

SupplementaryMaterial.docx

Download PDF

Journal Publication

published 04 Mar, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
06 Dec, 2022
Reviews received at journal
04 Dec, 2022
Reviewers agreed at journal
29 Nov, 2022
Reviewers invited by journal
21 Nov, 2022
Editor assigned by journal
14 Nov, 2022
Editor invited by journal
14 Nov, 2022
Submission checks completed at journal
14 Nov, 2022
First submitted to journal
04 Nov, 2022

You are reading this latest preprint version

Evaluation of two terpene-derived polymers as consolidants for archaeological wood

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials And Methods

Preparation of wood samples

Naming System

Density And Maximum Moisture Content Measurements

Immersion Tests With Polymer Solutions

Colour Change

Attenuated Total Reflection-fourier Transform Infrared Spectroscopy (Atr-ftir)

Scanning Electron Microscope (Sem) Analyses

Resistance To Indentation (Hardness Test)

Results And Discussion

Density and maximum moisture content measurements

Weight Change

Dimensional Change

Colour Change

Atr-ftir Analyses

Sem Analyses

Hardness Test

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1