Influence of volatile degradation products from thermal wood modification on wood-water interactions

Influence of volatile degradation products from thermal wood modification on wood-water interactions | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Influence of volatile degradation products from thermal wood modification on wood-water interactions Liselotte De Ligne, Maria Fredriksson, Lisbeth G. Thygesen, Emil E. Thybring This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5191132/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Thermal modification is a widely used technology for enhancing the dimensional stability and durability of wood. However, thermal degradation reactions in wood are complex and depend on the process conditions applied. Volatile degradation products from thermal wood modification are expected to affect the wood chemistry and the wood-water interactions of the thermally modified wood. In this paper, we investigate the impact on wood chemistry and wood-water interactions of retaining or evaporating volatile degradation products while thermally modifying beech and Scots pine in a closed thermal treatment process. The presence of extractable degradation products was determined based on water and multi-solvent extraction. pH measurements and ATR-FTIR analyses were performed to determine differences in wood chemistry. Light microscopy images of xylem cross-sections were taken to determine the vessel and tracheid lumen areas, and LFNMR and DSC measurements were performed to determine wood-water interactions. We found that thermal treatment with a cooling step at atmospheric pressure indeed allowed volatile degradation products to evaporate, in the case of beech, resulting in a less hydrophobic end product. However, for Scots pine, evaporating volatile degradation products during the thermal modification process did not have an effect on the pH and the amount of extractives, and as the impact on wood-water interactions was not in line with the findings on wood chemistry, the results are inconclusive. Our results demonstrate that volatile degradation products can have an impact on the wood-water interactions of thermally modified wood in the cooling step and that the results are wood species dependent. thermal modification volatile degradation products wood-water interactions LFNMR ATR-FTIR beech Scots pine Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Introduction Wood protection used to focus mainly on selecting naturally durable wood species, often from tropical regions, or treating non-durable wood species with fungicidal wood preservatives. General awareness of the negative impact of biocidal products on the environment initiated a new way of thinking about wood protection. Modern methods to protect wood from degradation aim at increasing the material's service life by removing or reducing a prerequisite for fungal degradation: water within the material. This can, for instance, be done by thermal modification. During thermal modification, wood is heated at high temperatures, usually between 160 and 240°C, to improve its dimensional stability and durability (Hill et al. 2021; Zelinka et al. 2022). Several process parameters can be varies: the duration and temperature applied during treatment steps, the treatment atmosphere (inert gas, air, vacuum), the use of ‘wet’ or ‘dry’ systems and open or closed systems (Hill et al. 2021). Thermal degradation reactions in wood are complex and depend on the wood species and the process conditions applied (Hill et al. 2021; Zelinka et al. 2022). During thermal treatment, degradation products ensue from the breakdown of cell wall polymers and wood extractives. For instance, the amount of hemicellulose degradation products, such as acetic acid and furfurals, increases with increased treatment temperature (Candelier et al. 2011) or with increased vapour pressure in closed systems (Altgen et al. 2016). Within a closed system, degradation products accumulate within the treatment vessel, producing acidic conditions that catalyse the depolymerisation of polysaccharides (Sundqvist et al. 2006; Altgen et al. 2016). In an open system, volatile by-products of degradation and exhaust gases are expected to be expelled into the atmosphere. This is likely also the case in a closed system process if the vessel pressure is released before the cooling step. Since the presence of volatile degradation products can affect the chemical reactions taking place, the venting off of volatile degradation products in the cooling step might affect the wood chemistry and wood-water interactions of thermally modified wood. In this paper, we investigate the impact of retaining or evaporating volatile degradation products in a closed thermal treatment process on the wood chemistry and wood-water interactions of thermally modified beech and Scots pine. Material and methods Wood specimens Boards (25 x 100 x 1200 mm³) of Scots pine ( Pinus sylvestris L.) and beech ( Fagus sylvatica L.) were exposed to two thermal modification treatments based on the ThermoTreat 2.0 process. This closed thermal modification process modifies the wood by hydrolysis at high pressures but lower temperatures than typical open processes, whereby pressure acts as a substitute for temperature. In the high-pressure thermal treatment (TT), the wood was heated to 170°C over 4 hours and kept at this temperature for 2 hours at 15–18 bar pressure before cooling down over 8 hours while the pressure was kept high (> 12 bar). The procedure for the second thermal treatment follows the same steps, but the cooling steps occurred after the pressure had been lowered to 1–3 bar, allowing volatile degradation products to evaporate (TT evap ). In total, 60 untreated controls, 60 specimens of the higher and 60 of the lower pressure thermally modified per wood species were then cut to a size of 5 x 5 x 10 mm 3 , where 10 mm was in the longitudinal direction. Extraction For each type of extraction protocol and wood species, 20 specimens per heat-treated material and 30 specimens per untreated control material (5 x 5 x 10 mm 3 ) were oven-dried at 60°C in a BINDER VD23 vacuum oven (BINDER GmbH, Tuttlingen, Germany) for 24 hours, to determine the oven-dry mass of the specimens before extraction ( m 0 ). After each extraction protocol, the specimens were oven-dried once more at 60°C in the vacuum oven for 24 hours to determine the oven-dry weight of the specimens after extraction ( m 1 ) (0.1 mg resolution, QUINTIX224-1S, Sartorius Lab Instruments GmbH & Co. KG, Goettingen, Germany). Note that the 20 or 30 specimens of the same type were weighed together before and after extraction, as labelling was impossible. The average amount of extractives \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) removed by the extraction protocol was calculated as follows: $$\:{c}_{\text{e}\text{x}\text{t}\text{r}}=\:\frac{{m}_{0}-\:{m}_{1}}{{m}_{0}}\:.\:100\%$$ 1 Multi-solvent extraction The specimens were placed in cellulose thimbles, with all specimens of one treatment in one thimble. First, the specimens were subjected to multi-solvent extraction with a 9:1 acetone-water mixture for 24 hours. Then, the specimens were subjected to multi-solvent extraction with a 1:2 ethanol-toluene mixture for 24 hours (Holmbom 1999; Sluiter et al. 2005). While setting up the second extraction, the specimens were left to evaporate in the fumehood. Water extraction The specimens were leached/water-extracted in accordance with the procedure described in EN 84 (1997). The specimens were placed in two-neck round bottom flasks (volume of 100 ml), with the top neck connected to a vacuum pump RZ 2.5 (Vacuubrand, GMBH, Werthelm, Germany) and the side-neck closed off with a rubber stopper. A vacuum corresponding to 4 kPa was installed for 15 minutes, after which deionised water was injected into the flask through the rubber stopper with a syringe. After 5 minutes of immersion, the vacuum was lifted, and the immersed specimens were left for 2 hours, after which the water was removed, and the bottom flasks were refilled with fresh deionised water. The specimens remained submerged for 14 days in deionised water. The water was changed at the end of the first and second day of immersion and then seven more times during the remaining 12 days. pH measurements For each sample type (TT, TT evap , untreated) and extraction method (Non-, Multi-solvent- and Water extracted), a pH measurement was performed (1 replicate). Wood specimens were milled in a Retsch MM400 mill (Retsch, Germany). For each specimen, 0.7 g of wood powder was added to a 50 ml tube. Deionised water was added with a pipette at a weight ratio of 1:6 (wood:water). A magnet was placed inside each tube, and a rotation speed of 300 rpm was installed to ensure adequate mixing of the wood powder and water. The pH was measured with a pH sensor (827 pH lab, Metrohm, Switzerland). In addition, pH measurements were performed on the water in which the LFNMR wood specimens (see 2.6) had been immersed (1 replicate per thermal treatment and extraction method). It is assumed that the water contains leached volatile components from the wood. The water was transferred from the Eppendorf tubes containing the wood specimens into a broader glass vile and measured with the pH sensor. Attenuated total reflectance Fourier transform infrared spectroscopy Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) spectroscopy was applied to determine changes in the wood polymers due to the two thermal modification procedures. Two replicates of the non-extracted and multi-solvent-extracted beech and Scots pine specimens of both thermally modified wood and untreated controls were dried in a BINDER VD23 vacuum oven (BINDER GmbH, Tuttlingen, Germany) for 24 hours at 60°C. The specimens were placed on the ATR-FTIR platform and covered with a custom-made conditioning chamber with a continuous airflow of 0% RH. FTIR measurements were taken of each specimen using a Nicolet 6700 FT-IR, Pike Technologies GladiATR diamond spectrometer (Thermo Scientific, Waltham, MA, USA), with a working temperature of 25°C. The specimens were placed on the FTIR crystal using the openings on the side of the humidity chamber, to cause minimal disturbance to the targeted 0% RH conditions during measurement. The spectral range included was 3700–500 cm − 1 , and spectra were obtained using 64 scans (128 for the background) with a resolution of 2 cm − 1 . Atmospheric correction was applied in OMNIC (version 9.8.372, Thermo Fisher Scientific Inc.). Junk data between 1800 and 2700 cm − 1 was excluded. For each replicate, the borders of 13 regions of interest were identified and linear baseline correction was applied to these individual peaks. A baseline-corrected spectrum was obtained for each replicate and normalised by dividing all absorbance values by the absorbance value of the highest peak (around 1026 cm -1 ). A smoothed spectrum was obtained for each specimen after smoothing with a Savitzky-Golay filter (polynomial = 3, window = 9). Baseline corrected areas were taken (before normalisation) of four reference peaks: holocellulose (1107 cm -1 ), hemicellulose (1056 cm -1 ), lignin (1502 cm -1 ) and the carbonyl peak in hemicellulose (1739 cm -1 ), and peak ratios were determined. Principal component analysis (PCA) was applied to the smoothed FTIR spectra to determine the impact of thermal modification on the FTIR spectra of beech and Scots pine and to identify which wavenumbers in the FTIR spectra were related to thermal modification (Python, sklearn.decomposition.PCA package). Light microscopy Light microscopy images were taken from microtome slices to determine the cross-sectional tracheid cell lumen areas (Scots pine) and vessel areas (beech) per cell. Specimens of Scots pine and beech were cut to a size of 2 x 2 x 10 mm³ for each treatment and extraction method with two replicates per material type. The wood specimens were placed in a Leica RM2255 automated rotary microtome (Leica Biosystems, United States) and slices were cut at 30 and 40 µm thickness for thermally treated Scots pine and beech, respectively. The untreated wood specimens were cut at 20 µm thickness and stained with Toluidine Blue. The thermally modified wood slices were not died as the darkened wood cells due to modification allowed sufficient visualisation. The wood slices were placed in a Leica compound microscope, and images were taken at 40x magnification with a Leica ICC50 camera (Leica Microsystems). For each specimen, an image was taken from an earlywood (EW) and a latewood zone (LW), taken at the beginning and end of the growth ring, respectively. Since the thermally modified wood was brittle, cell tissue ruptured easily, especially near the sides of the slice, and the images were taken where the tissue was most intact. Images were pre-processed in Fiji (Schindelin et al. 2012) to obtain binary images. Incomplete cell areas were manually coloured to be excluded from the measurements, and tracheid/lumen cell areas were determined in WinCELL (Regent Instruments Inc., Canada). Low-Field Nuclear Magnetic Resonance measurements Low-Field Nuclear Magnetic Resonance (LFNMR) spectroscopy was applied to gain knowledge on the water populations in the wood specimens in water-saturated state. Five replicates (5 x 5 x 10 mm³) for each treatment, extraction method and species were placed in two-neck round bottom flasks, with the top neck connected to a vacuum pump RZ 2.5 (Vacuubrand, GMBH, Werthelm, Germany) and the side-neck closed off with a rubber stopper. A vacuum corresponding to 4 kPa was installed for 15 minutes, after which deionised water was injected into the flask through the rubber stopper with a syringe. After 5 minutes of immersion, the vacuum was lifted, and the immersed specimens were transferred to water-filled specimen tubes and left overnight. Then, the specimens were put in a pre-weighed glass LFNMR tube, after the excess surface water was removed by dapping on a wet cloth (Fredriksson and Thygesen 2017). The wet mass of the specimen was measured with a resolution of 0.1 mg. A solid Teflon rod with an O-ring sealing was inserted into the glass LFNMR tube to fill the remaining space, thereby limiting water evaporation from the specimen during the experiment. The glass LFNMR tube containing the specimen was placed in a Bruker mq20 minispec with a 0.47 T permanent magnet (Bruker, Billerica, MA, USA) to perform LFNMR measurements. The temperature in the LFNMR around the specimen was kept constant at 22°C by water circulation in the probe using a refrigerated and heating circulator (Julabo GmbH, Seelbach, Germany). The specimen was allowed to thermally equilibrate in the instrument for two minutes before starting each measurement. The Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence was used to measure the spin-spin relaxation time ( T 2 ) of the specimens with a pulse separation ( τ ) of 0.1 ms, 8000 echoes, 32 scans and a recycle delay of 30 s. Exponential decay analysis (Istratov and Vyvenko 1999) was applied to the recorded LFNMR decay curves to give smooth, continuous distributions of T 2 relaxation times. 200 exponentials were fitted to the data with fixed, logarithmically spaced T2 relaxation times from the first to the last time point (0.2 to 2500 ms). For each peak, the relative sum of amplitudes of the exponential components and their T 2 values corresponding to maximum peak intensity were determined. After the experiment, the LFNMR specimens were weighed ( m w ), oven-dried at 60°C in a BINDER VD23 vacuum oven (BINDER GmbH, Tuttlingen, Germany) for 24 hours and weighed again ( m d ). When researching the impact of thermal modification, it is interesting to assess both total moisture concentration and moisture content of the wood specimens, as the oven-dry weight of thermally modified wood is lower than untreated wood. The moisture concentration ( c sat ) of the LFNMR specimens after water saturation was calculated as follows, with V the volume of the wood specimens: $$\:{c}_{sat}=\:\frac{{m}_{w}-\:{m}_{d}}{V}$$ 2 The moisture content (ω) of the LFNMR specimens after water saturation was calculated as follows: $$\:\omega\:=\:\frac{{m}_{w}-\:{m}_{d}}{{m}_{d}}$$ 3 To determine the influence of water-soluble components on LFNMR, three specimens per material and extraction type were water-saturated as described above and stored in Eppendorf tubes in a refrigerator. After 18 days of immersion, a 1 ml syringe was used to suck up water from the Eppendorf tube in which the specimen had been immersed. The liquid was passed through a Nylon syringe filter (MicroLab Scientific, diameter 13 mm, pore size 0.45 µm) while transferred to a new Eppendorf. 350 µl of liquid was transferred into a glass LFNMR tube. The Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence was used to measure the spin-spin relaxation time ( T 2 ) of the specimens with a pulse separation ( τ ) of 0.1 ms, 10000 echoes, 32 scans and a recycle delay of 30 s. Additionally, MilliQ water was assessed with LFNMR. For the MilliQ water and water from the multi-solvent extracted specimens, 32000 echoes were necessary. Characterisation of maximum cell wall moisture content with DSC The cell wall moisture content in water-saturated state was determined with Differential Scanning Calorimetry (DSC) as described by Fredriksson et al. (Fredriksson et al. 2023). Here, the maximum cell wall moisture content can be determined by measuring the heat required to thaw a frozen water-saturated specimen. Since the water outside the cell wall freezes, but the cell wall water does not, the total moisture content can be separated into capillary water and cell wall water, respectively (Fredriksson and Thygesen 2017; Thybring et al. 2020). Specimens (5 x 5 x 10 mm 3 ) from the LFNMR experiments of all material types were vacuum-saturated with water. From these, circular discs with a diameter of approximately 5 mm and a height of 1.4 mm were cut to fit DSC-pans of 40 µl. The specimens were cut while keeping them in water to avoid drying. Excess water was then removed by wiping the surfaces of the specimen with a moist cloth (Wettex, Freudenberg Home and Cleaning Solutions GmbH, Malmö, Sweden). Five replicates were cut for each material type, so 90 specimens were measured. After cutting, each specimen was placed in a pre-weighed DSC pan, which was then hermetically sealed. The water-saturated mass, m sat (g), was determined. Then, the specimens were placed in the sample robot of the DSC (DSC 3+, Mettler Toledo AG, Schwerzenbach, Switzerland) where the following method was applied: − 20°C for 5 min, a ramp from − 20 to 20°C with 2°C min − 1 , and finally a temperature stabilisation period of 5 min at 20°C. After the measurements, the lids of the DSC pans were pierced with a needle and placed in a vacuum oven at 60°C for 24 h. Each specimen's dry mass, m d , was determined after temperature stabilisation in a desiccator with molecular sieves. The cell wall moisture content, ω cw, was then evaluated (Fredriksson and Thybring 2019; Fredriksson et al. 2023) by: $$\:{\omega\:}_{\text{c}\text{w}}=\frac{{m}_{\text{s}\text{a}\text{t}}-{m}_{\text{d}}-\left(\frac{\text{Q}\bullet\:{m}_{\text{s}\text{a}\text{t}}}{{H}_{\text{f}}}\right)}{{m}_{\text{d}}}$$ 3 where H f (J g − 1 ) is the enthalpy of fusion of water of 333.7 J g − 1 and Q (J) is the heat of melting of the capillary water, which was determined by integration of the melting peak. This integration was performed in STARe Evaluation Software (version 17.00, Mettler Toledo AG, Schwerzenbach, Switzerland). Calibration of the DSC for enthalpy of fusion was done with deionised water (melting point 0°C, enthalpy of fusion 333.7 J g − 1 ). Statistical data analysis The Kruskal–Wallis H-test, a nonparametric statistical test, was applied to test whether the median values of untreated, TT and TT evap datasets differed. When the difference between medians was significant, Dunn’s multiple comparison test (Dunn 1961) was applied to pinpoint for which materials the medians differed. Benjamini-Hochberg correction (Benjamini and Hochberg 1995) was performed to control the false discovery rate (Type I error). Results Extractive content and pH The thermally modified wood was clearly more acidic for both species than the non-modified wood (Fig. 1 b,c). The cooling down step at atmospheric pressure allowed volatile degradation products to evaporate in the case of beech (Fig. 1 a). For Scots pine, the results are inconclusive, as no significant difference was observed in amount of extractives between TT and TT evap . Thermally modified beech contained higher amounts of extractives than the untreated beech, both for multi-solvent and water-extracted specimens (Fig. 1 a). For beech, more extractives were removed from TT as compared to TT evap , indicating that less volatile components were present in TT evap as compared to TT (Fig. 1 a) and that the cooling step at lowered pressure indeed allowed volatile degradation products to evaporate. Likewise, the pH of TT was slightly more acidic than TT evap , confirming the lower presence of acidic, volatile degradation products in TT evap (Fig. 1 b and c). However, the magnitude of the pH difference was different for the two methods used. For pH measurements in water in which the wood had been immersed, the difference in pH between TT and TT evap was higher than for the measurements made on wood powder. This is likely because of the high acidity of the wood compounds that remained in the wood cell wall (volatile and non-volatile) and came free due to the milling. The multi-solvent-extracted specimens all had a high pH, but there was still a difference between TT and TT evap for the pH measurements on wood powder. This high pH is not unexpected, as less volatile components remain in the wood after multi-solvent extraction, and the relative amount of acid components relative to water (pH 7) is less. Note that the pH of the thermally modified wood powder was more acidic than the non-modified wood powder both for beech and Scots pine, which is expected for thermally modified wood (Boonstra et al. 2007). For Scots pine, only 2–3 wt% was removed after multi-solvent and water extraction and only small (< 1%) and no consistent differences were observed in \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) between TT, TT evap and the untreated wood. The extractive content in untreated Scots pine wood is generally around 3–5% (Verkasalo et al. 2022). Likewise, the pH of the water in which the wood specimens had been immersed was similar between thermally modified and untreated Scots pine, with only minor differences of 0.2 (water extracted, non-extracted) in pH value (Fig. 1 c). In contrast, the pH of the thermally modified, non-extracted wood powder was significantly lower than the that of the untreated wood, indicating that the wood components themselves had become more acidic after the thermal modification (Fig. 1 b). The pH of the multi-solvent-extracted specimens was very high, both for the wood powder and the leachate. Wood composition assessed with ATR-FTIR ATR-FTIR spectra The ATR-FTIR spectra of non-extracted beech show clear differences among TT, TT evap and untreated (Fig. 2 ). In the OH range, there was a clear decrease in the region 3600 − 3400 cm -1 (peak 1 in Fig. 2 ), assigned to OH-stretching in alcohols (Esteves et al. 2013) for thermally modified beech compared to untreated beech. For TT evap there was an increase in the region 3350 − 3000 cm -1 (peak 2), assigned to OH-stretching of carboxyl acids (Esteves et al. 2013). The intensity of the carbonyl peak at 1739 cm -1 (peak 3) had decreased for thermally modified beech as compared to untreated beech, which is an indication of the degradation of hemicelluloses, such as xylan and glucomannan (Tjeerdsma and Militz 2005a; Guo et al. 2018a). The aromatic rings in hardwood lignin (Guaiacyl - G and Syringyl - S) usually exhibit benzene ring stretching vibrations at approximately 1505 cm -1 (Esteves et al. 2013), which is indeed the case for the untreated beech wood (1502 cm -1 ) (peak 4). For thermally modified beech, the peak maximum shifted to 1515 cm -1 , which can be attributed to the breaking of aliphatic side chains in lignin and/or cross-linking by condensation reactions (Windeisen et al. 2007; Kocaefe et al. 2008; Esteves et al. 2013). There is a clear decrease at 1228 cm -1 (peak 5), tentatively assigned to C-O absorption (Windeisen et al. 2007; Pedersen 2015; Huang et al. 2019), for thermally modified wood compared to untreated wood, indicating alteration and cleavage of carbonyl groups (Windeisen et al. 2007). The bands at 1161 cm -1 , 1109 cm -1 and 1057 cm -1 (peaks 6,7 and 8), related to cellulose (Lupoi et al. 2015; Özgenç et al. 2017), increased after thermal modification. Similar observations could be made based on the ATR-FTIR spectra of the multi-solvent extracted beech samples (Suppl. Fig. S1). The ATR-FTIR spectra of non-extracted Scots pine show a higher standard deviation than of beech, especially for untreated and TT evap . For TT, a clear decrease in the region 3600 − 3400 cm -1 (peak 1), assigned to OH-stretching in alcohols (Esteves et al. 2013) could be observed, but not for TT evap, while the latter could be observed in the ATR-FTIR spectra of the multi-solvent extracted samples (Suppl. Fig. S1). Similarly, the intensity of the carbonyl peak at 1739 cm -1 (peak 3) had decreased for thermally modified Scots pine as compared to untreated Scots pine, indicating the degradation of hemicelluloses, but not for TT evap . The bands at 1161 cm -1 and 1109 cm -1 (peaks 6 and 7), related to cellulose (Lupoi et al. 2015; Özgenç et al. 2017) increased for both thermal modifications (TT and TT evap ). Peak ratios Peak ratios were calculated for specific FTIR wavenumber regions representing holocellulose, hemicellulose and lignin, as well as the carbonyl peak in hemicellulose (Fig. 3 ). Overall, the standard deviation of the non-extracted peak ratios was higher than the multi-solvent-extracted peak ratios, especially for TT evap . For beech, no significant differences were found for the 1056/1107 peak ratio representing hemicellulose relative to holocellulose. Indeed, from the ATR-FTIR spectra (Fig. 2 ), it could be observed that both peaks increased for thermally modified beech. Similarly, no significant difference was found for the 1508/1107 peak ratio representing lignin relative to holocellulose. While there was a clear peak shift from 1502 to 1515 cm -1 , the overall peak area taken for calculating the peak ratio did not change (area taken from wavenumber region 1480–1530 cm -1 ). In contrast, the 1739/1502 peak ratio for the TT evap beech was significantly lower than the untreated wood for multi-solvent extracted specimens (Suppl. Table 1). Likewise, the non-extracted specimens followed the same pattern, but no significant difference was found with Kruskal-Wallice, due to the high standard deviation of TT evap . For Scots pine, similar results were found, except the non-extracted 1056/1107 peak ratio, where TT was significantly higher than TT evap and untreated wood (Suppl. Table 1). The 1739/1502 peak ratio for TT Scots pine was significantly lower than the untreated wood for non-extracted specimens (Suppl. Table 1). Interestingly, this was not the case for the multi-solvent extracted specimens, for which the median peak ratios were similar. PCA of ATR-FTIR spectra For beech, untreated (untr) and thermally modified wood (TT and TT evap ) were clearly distinguishable based on PC1 (56.1% of variance) (Fig. 4 ). The spectral ranges that were most important for separating thermally modified beech wood from untreated beech wood based on the ATR-FTIR spectra (Fig. 5 ) were peak 1a, 3600 − 3300 cm -1 OH assigned to stretching from alcohols (Esteves et al. 2013), peak 3 representing the 1739 cm -1 carbonyl peak, peak 5 at 1228 cm -1 tentatively assigned to C-O absorption (Windeisen et al. 2007; Pedersen 2015; Huang et al. 2019) and peak 7 and 8, related to cellulose (Lupoi et al. 2015; Özgenç et al. 2017). The increase in the region 3350 − 3000 cm -1 (peak 2), assigned to OH-stretching of carboxyl acids (Esteves et al. 2013) is more clearly visible in the loading profiles (Fig. 5 ) than in the average ATR-FTIR spectra (Fig. 3 ). For Scots pine, a distinction between thermally modified and untreated wood could also be made, based on a combination of PC3 (2.2%) and PC4 (4.6%). The PCA score plots also distinguished between multi-solvent-extracted and non-extracted beech samples, based on PC4, with mostly negative values for multi-solvent-extracted specimens. Light microscopy The light microscopy images showed significantly lower cross-sectional vessel areas per cell for thermally modified beech (TT) than untreated beech for earlywood and latewood in non-extracted samples (Fig. 5 , Fig. 7 , Suppl. Tables 2 and 3). Interestingly, larger cross-sectional vessel cell areas were found for TT earlywood and TT evap latewood for water-extracted beech specimens, but not for TT beech latewood with light microscopy (Fig. 6 , Fig. 7 ). No significant difference in earlywood cross-sectional cell lumen areas was found between TT, TT evap and untreated Scots pine for non-extracted and water-extracted specimens (Suppl. Table 3). The latewood cell lumen areas of TT evap were significantly smaller than untreated wood (100–130 µm²) and TT (50–150 µm²), independent of extraction method (Fig. 6 , Suppl. Table 3). The cell lumen areas clearly differed between water extracted specimens compared to the other extraction methods for TT and TT evap , except for TT beech latewood (Suppl. Tables 4 and 5). Interestingly, a significant difference was only found twice for the untreated specimens (Beech EW: Water x Non and Scots pine EW: Water x Multi-solvent). Wood-water interactions in water-saturated state Water populations assessed with LFNMR The water populations of each material were assessed with LFNMR in water-saturated state. Continuous T 2 distributions for non-extracted beech and Scots pine are shown in Fig. 8 . All continuous T 2 distributions can be found in Supplementary materials, plotted per extraction method (Fig. S3) and per treatment method (Fig. S4). Thermal treatment decreased the mean T 2 value of the cell wall peak, both for beech and Scots pine ( Fig. 8 , Fig. 9 a), with the decrease statistically different between TT evap and untr. (Suppl. Table 6–7). TT evap had a lower T 2 value for the cell wall peak as compared to TT for beech (Fig. 9 a). While this difference was not significant, it did occur consistently for all extraction methods (Suppl. Table 7, Suppl. Figures 3 and 4). A similar effect could be observed to a lesser extent for Scots pine, but the standard deviation of TT evap was high. For the non-extracted and multi-solvent-extracted specimens, the T 2 value for the vessel peak of beech was slightly higher for TT than the untreated specimens, and slightly lower for TT evap , with the difference between TT and TT evap significant (Fig. 9 b, Suppl. Figure 3, Suppl. Table 7). Interestingly, the T 2 values for the water-extracted beech specimens were much higher than those of the non- and multi-solvent-extracted specimens, but only for the thermally modified wood. For Scots pine, the T 2 value of the lumen peak increased after thermal modification, with a higher increase for TT evap than TT, for all extractions (Fig. 9 b, Suppl. Figure 3, Suppl. Table 7). The volatiles peak shows the T 2 values for liquid water in which a wood specimen had been immersed for 18 days (Fig. 9 c). The presence of volatiles caused the T 2 value to decrease, with non-extracted wood having lower T 2 values than extracted wood, thermally modified wood having lower T 2 values than untreated wood and deionized water having the highest T 2 values. Cell wall moisture content at saturation The maximum cell wall moisture content of the thermally modified beech and Scots pine was significantly lower than for untreated wood (Suppl. Tables 8–9), for both LFNMR and DSC (Fig. 10 ). No significant difference was found between TT and TT evap, though for beech the cell wall moisture content for TT evap was consistently lower than TT. While both techniques gave similar results, the absolute values for moisture content were lower for LFNMR than for DSC measurements, which has been shown before (Thybring et al. 2020). Interestingly, the cell wall moisture content for the thermally modified samples was higher for multi-solvent extracted specimens, as shown with DSC (both beech and Scots pine) and LFNMR (only beech). Total moisture concentration and moisture content in water-saturated state For thermally modified wood, it is interesting to assess both total moisture concentration and moisture content of the specimens are provided in Fig. 11 . Both for beech and Scots pine, the moisture concentration of the thermally modified specimens in the water-saturated state was lower than that of the untreated samples (Fig. 11 ), with TT evap (but not TT) significantly lower than the untreated wood for beech and both thermal treatments significantly lower for Scots pine (Suppl. Table 9). The moisture concentration of TT evap was lower than TT for beech, but not significantly. While the total moisture concentration of the thermally modified beech specimens was lower, the total moisture content values were similar to untreated wood, as the oven-dry weight of the thermally modified wood specimens was lower (Suppl. Fig. S5). Interestingly, for Scots pine, the oven-dry mass of TT evap was lower than the untreated wood, but not that of TT, which explains the lower moisture content (Suppl. Fig. S5). Discussion Impact of thermal treatment on wood chemistry Thermal modification leads to changes in the wood cell wall polymers. The decreased pH of the thermally treated beech and Scots pine samples is a clear indication of this (Hofmann et al. 2008; Cai et al. 2018; Dzurenda and Dudiak 2021), as (acetic) degradation products are formed during the thermal degradation of hemicellulose and lignin (Hill et al. 2021). Thermally modified beech contained higher amounts of extractives than the untreated beech, which was not the case for Scots pine. A study on steam thermal treatment under N 2 gas also showed a clear increase in \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) after thermal treatment for beech (from 2.4–3.6 to 11.9–13.7%) for ethanol-toluene extracts (Hofmann et al. 2008). For Scots pine, only small differences (< 1%) were observed for multi-solvent extraction, in line with two other studies on multi-solvent extraction of thermally modified Scots pine (0.5-2%) (Hofmann et al. 2008; Källbom et al. 2018). Hardwoods have been shown to be more sensitive to thermodegradation than softwoods, for which degradation products appear at slightly higher temperature (Candelier et al. 2011). For beech, the lesser amount of extractives and the less acidic pH value of TT evap confirm that cooling at lowered pressure allows volatile degradation products to evaporate. In contrast, for Scots pine, there was no discernible difference in pH and extractives among the two thermal treatments. Do note that there was no mass decrease after multi-solvent extraction in untreated beech, while 2–3 wt% would be expected (Hofmann et al. 2008). During thermal treatment, the amorphous polysaccharide content degrades, predominantly affecting hemicelluloses. Hardwoods have indeed been shown to experience a greater degree of hemicellulose degradation compared to softwoods when subjected to the same conditions (Hill et al. 2021). Hardwoods degrade at lower temperatures due to the higher acetyl group content in the hemicelluloses, with the xylans of hardwoods being acetylated, in contrast to those of softwoods (Hill et al. 2021). In dry thermal treatment conditions, degradation of carbohydrates is likely to lead to dehydration products, such as furfurals (Tjeerdsma and Militz 2005b), which have the potential to polymerise and form water-insoluble hydrophobic materials in the cell wall. While not removable with water extraction, furfurals are easily extractable with toluene (Nhien et al. 2017), possibly accounting for the high \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) for multi-solvent extracted thermally modified beech. Non-modified beech wood contains 1.0% lipophilic and 3.7% hydrophilic extractives, with saturated fatty acids, fatty alcohols, and free sterols as the main extractive components in the heartwood (Vek et al. 2016). Besides hemicellulose degradation during thermal modification, degradation of wood extractives occurs, as well as migration of wood extractives towards the wood surface, if these are mobile under the thermal treatment conditions (Hill et al. 2021). However, Mecca et al. (2021) reported that there were no substantial changes in the type of compounds in beech extractives after thermal modification. The non-volatile extractive compounds in Scots pine are mainly free fatty acids, resin acids and phenolic compounds. Estevez et al. show an increase in ethanol-soluble extractives between 1-2.5% (Esteves et al. 2011). This increase in extractives was mainly attributed to phenolic compounds and anhydrosugars, with the appearance of new compounds after thermal modification, such as vanillin, vanillic acid and 3-vanillyl propanol. Interestingly, the pH of the multi-solvent-extracted specimens was very high, both for beech and Scots pine. This could simply indicate that the multi-solvent extraction removes acidic products better than water extraction. That would indicate that although the \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) in Scots pine was low, a certain amount of (volatile) acidic compounds were still removed during multi-solvent extraction. Another hypothesis is that part of the solvents remained in the wood after multi-solvent extraction and contributed to the higher pH, with ethanol and acetone (pH 7) the most likely solvents as toluene is a non-polar molecule. The wood composition was assessed with ATR-FTIR to determine the modification of wood cell wall polymers. Several indicators of thermal modification were present when comparing the ATR-FTIR spectra of TT and TT evap to the untreated wood, such as a decrease in the region 3600 − 3400 cm -1 , assigned to OH-stretching in alcohols (Esteves et al. 2013), a decrease in intensity of the carbonyl peak (Tjeerdsma and Militz 2005a; Guo et al. 2018a) at 1739 cm -1 and an increase in intensity for bands related to cellulose (Lupoi et al. 2015; Özgenç et al. 2017). Both for beech and Scots pine a decrease in the hemicellulose/lignin ratio could be observed, which is expected after thermal modification (Källbom et al. 2018), though this was not the case for the multi-solvent extracted specimens of the latter. Interestingly, for thermally modified beech, the peak maximum shifted to 1515 cm -1 , which can be attributed to the breaking of aliphatic side chains in lignin and/or cross-linking by condensation reactions (Windeisen et al. 2007; Kocaefe et al. 2008; Esteves et al. 2013). The chemical composition of the thermally modified beech wood differed slightly between TT and TT evap based on the PCA of the FTIR spectra, with TT having more negative values than TT evap for PC1. This difference in chemical composition was also confirmed by the difference in pH and \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) . We would expect the hemicellulose/lignin ratio to be higher for TT evap than TT, which is indeed the case for the non-extracted wood, but not for the multi-solvent extracted wood. For Scots pine, the ATR-FTIR peak ratio of 1739/1502 was significantly lower for TT as TT evap , which might indicate a higher level of thermal modification, while the 1056/1107 (hemicellulose/holocellulose) peak ratio was lower for TT evap than TT. In addition, only a small part of the variance in ATR-FTIR spectra was assignable to a difference between thermally modified and untreated Scots pine wood in the PCA. Wood-water interactions in water-saturated state Capillary water interactions The T 2 value of the lumen peak increased after the thermal modification of Scots pine, with a higher increase for TT evap than TT for all extractions. This increase in T 2 values for the lumen peak typically occurs in thermally modified wood, and has been shown for Scots pine (Cai et al. 2020), Norway spruce (Cai et al. 2020) and radiata pine (Guo et al. 2018b). Our results show that this increase in T 2 value was not caused by an increase in cell lumen diameter, as no increase in tracheid lumen area was observed for thermally modified Scots pine with light microscopy imaging, and even a decrease in cell lumen area occurred in the latewood. Such decreases in lumen dimensions have been shown for thermally modified Scots pine before, in a study applying pulsed-field-gradient stimulated-echo NMR and field emission scanning electron microscopy (Kekkonen et al. 2010). Wood volume typically decreases after thermal modification, mainly due to the loss of degraded cell wall material. Additionally, there is a strong possibility that molecular restructuring within the cell wall contributes to this volume decrease (Hill et al. 2021). Furthermore, the increase in T 2 value for the lumen peak was not linked to the presence of volatile compounds. Volatiles caused the T 2 value to decrease, with non-extracted wood having lower T 2 values than extracted wood, non-extracted thermally modified wood having lower T 2 values than non-extracted untreated wood and deionised water having the highest T 2 values. T 2 values have indeed been shown to decrease with increasing concentration of dissolved molecules, for sugars for example (Hsieh et al. 2014), because the average molecular distance between the solute becomes shorter. As the cell lumen areas had not increased and volatile compounds have a decreasing effect on T 2 values, the increase in T 2 values for the lumen peak is a clear indication of the increased hydrophobicity of the modified Scots pine. The differences in T 2 values and pH confirmed the presence of volatile components in the leached water, supporting the hypothesis that (part of) the volatiles remained in the water as assumed before the start of the experiments. For beech, the T 2 value for the lumen peak increased for TT and decreased for TT evap . However, like for Scots pine, the light microscopy images showed significantly lower vessel cell areas for thermally modified beech (TT) as compared to untreated beech, both for earlywood and latewood in non-extracted samples. Likewise, less volatile compounds, which have a decreasing effect on T 2 values, are expected in TT evap . as compared to TT. The higher T 2 value for the lumen peak for TT as compared to TT evap is, therefore, likely an indication of an increased hydrophobicity. Indeed, more hydrophobic compounds were present in TT, as confirmed by the higher \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) after multi-solvent extraction. Note that in the case of beech, the light microscopy data might be slightly skewed towards smaller vessel areas, as cross-sections that had smaller vessels were less likely to rupture and provide an undamaged imaging region for data analysis. Interestingly, the T 2 values for the water-extracted beech specimens were much higher than those of the non- and multi-solvent-extracted specimens, but only for the thermally modified wood. Similarly, larger vessel areas were found for TT earlywood and TT evap latewood, but not for TT latewood with light microscopy. This is unexpected as thermal modification is generally expected to decrease lumen size (Kekkonen et al. 2010). However, beech is known for the formation of tyloses (De Micco et al. 2016), which could indeed be observed (Fig. 7 , non-extracted beech earlywood TT evap ). It could be that the water extraction procedure, in which the samples are water extracted under vacuum pressure and kept submerged in water for 14 days, caused the tyloses in the thermally modified wood to rupture, enlarging the space for capillary water. Cell wall moisture content at saturation Thermal treatment decreased the mean T 2 value of the cell wall peak, both for beech and Scots pine. However, the removal of hemicelluloses due to thermal degradation is expected to increase the available space for moisture in the cell wall (Thybring and Fredriksson 2021). Presumably, the decrease in T 2 value is linked to a decrease in pore size in the wood cell wall in water-saturated state, due to increased crosslinking within the cell wall polymers, changes in the mobility of the polymer network and bulking of thermally degraded components that have remained in the cell wall (Thybring and Fredriksson 2021; Hill et al. 2021). Such a decrease in the cell wall peak has also been shown for thermally modified radiata pine (200°C, 24 h, N 2 ) (Guo et al. 2018b). However, no decrease in T 2 value of the cell wall peak was observed for thermally modified Scots pine under steam (180°C, 2 h, steam-treatment) (Cai et al. 2020). For beech, the TT evap had a lower T 2 value for the cell wall peak as compared to TT. Since less extractives were present in TT evap as compared to TT beech, this is not likely linked to increased bulking of thermally degraded components. Like for the increased T 2 value of the lumen peak, this could be an indication of increased hydrophobicity, related to the higher content of hydrophobic compounds in TT, as compared to TT evap . The retaining of volatile degradation products during the thermal modification process could also have affected other parameters, such as crosslinking within the cell wall polymers and changes in the mobility of the polymer network. Water holding capacity The cell wall moisture content of Scots pine reduced from around 30% to around 20% after thermal modification, as assessed with LFNMR. Similar values were found for Radiata pine in a study by Guo et al. (2018b). The cell wall moisture content for the thermally modified samples was higher for multi-solvent extracted specimens, as shown with DSC (both beech and Scots pine) and LFNMR (only beech). Multi-solvent extraction removes hydrophobic compounds and is consequently expected to increase the hydrophilicity (Källbom et al. 2018). In addition to that, mass loss from extraction typically results in an increase in cell wall moisture content, as soluble components that have leached out of the cell wall free up space for water molecules. Interestingly, the cell wall moisture content of water-extracted Scots pine specimens was similar to that of untreated wood. This phenomenon has also been shown in earlier studies demonstrating that although thermal modification in dry conditions results in a reduction of (equilibrium) moisture content, the effects can be partially reversible when exposed to high RH levels or liquid water. This reversible effect has been shown before for Scots pine (Hill et al. 2012), though also for beech (Cermak et al. 2015), where the water extraction procedure in our study did not have the same effect. The maximum cell wall moisture content and overall water concentration of the thermally modified wood specimens in a water-saturated state were lower than of the untreated wood. However, the overall moisture content of the specimens was similar, as the thermally modified wood specimens had a lower dry mass than the untreated specimens. The equilibrium moisture content of thermally modified wood is expected to be lower than untreated wood, as shown for water vapour sorption at relative humidities going from 0 to 95% (Hill et al. 2012; Altgen and Militz 2016). Few studies were found showing the moisture content at water saturation of thermally modified wood, though Guo et al. report an average moisture content at water saturation of about 160% for thermally modified and 170% moisture content for untreated radiata pine (Guo et al. 2018b). Implications and recommendations for future research The cooling step at lowered pressure allowed volatile degradation products to evaporate from thermally modified beech. The T 2 value for the lumen peak and cell wall peak as assessed with LFNMR, were higher for TT than for TT evap , indicating that retaining volatile degradation products during the thermal treatment process resulted in a more hydrophobic end-product. For Scots pine, the impact of allowing volatile degradation products to evaporate is less straightforward to interpret. No clear differences were found between TT and TT evap based on pH and \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) . The T 2 value of the lumen peak did have a higher increase for TT evap than TT and the 1056/1107 (hemicellulose/holocellulose) peak ratio was lower for TT evap than TT. However, the peak ratio of 1739/1502 (hemicellulose/lignin) gave the opposite result, being significantly lower for TT than for TT evap . For a follow-up study, it would be interesting to assess how both processes affect water vapour sorption at increasing relative humidity as well as the susceptibility to fungal decay of both wood species. Conclusion Retaining volatile degradation products during the thermal treatment process resulted in a more hydrophobic end-product in case of beech. For Scots pine, the type of thermal modification process did not have an effect on the pH and \(\:{c}_{\text{e}\text{x}\text{t}\text{r}}\) and as the impact on wood-water interactions was not in line with the findings on wood chemistry, the results are inconclusive. The methodological approach, including light microscopy, extraction and LFNMR assessment of volatile compounds, offered important insights into the LFNMR results, ensuring that conclusions related to increases in LFNMR signal were indeed linked to increased hydrophobicity and not related to other possible side-effects of the thermal modification process that might affect the LFMNR signal, such as in-or decreases in lumen size and the presence of volatile components. Declarations Competing Interests The authors have no relevant financial or non-financial interests to disclose. Ethics Approval Ethics approval was not required for this study. Funding The authors gratefully acknowledge financial support from Interreg Öresund-Kattegat-Skagerrak grant number 20201851, Aase & Ejnar Danielsens Fond, STARK fonden and Åforsk foundation. Author Contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by L.D.L., M.F. and E.E.T. The first draft of the manuscript was written by L.D.L. and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. References Altgen M, Militz H (2016) Influence of process conditions on hygroscopicity and mechanical properties of European beech thermally modified in a high-pressure reactor system. Holzforschung 70:971–979. https://doi.org/10.1515/hf-2015-0235 Altgen M, Willems W, Militz H (2016) Wood degradation affected by process conditions during thermal modification of European beech in a high-pressure reactor system. Eur J Wood Prod 74:653–662. https://doi.org/10.1007/s00107-016-1045-y Benjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B: Statistical Methodology 57:289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x Boonstra MJ, van Acker J, Kegel E, Stevens M (2007) Optimisation of a two-stage heat treatment process: durability aspects. Wood Sci Technol 41:31–57. https://doi.org/10.1007/s00226-006-0087-4 Cai C, Antikainen J, Luostarinen K, et al (2018) Wetting-induced changes on the surface of thermally modified Scots pine and Norway spruce wood. Wood Sci Technol 52:1181–1193. https://doi.org/10.1007/s00226-018-1030-1 Cai C, Javed MA, Komulainen S, et al (2020) Effect of natural weathering on water absorption and pore size distribution in thermally modified wood determined by nuclear magnetic resonance. Cellulose 27:4235–4247. https://doi.org/10.1007/s10570-020-03093-x Candelier K, Chaouch M, Dumarçay S, et al (2011) Utilization of thermodesorption coupled to GC–MS to study stability of different wood species to thermodegradation. Journal of Analytical and Applied Pyrolysis 92:376–383. https://doi.org/10.1016/j.jaap.2011.07.010 CEN (1997) EN 84 Wood preservatives - Accelerated ageing of treated wood prior to biological testing - Leaching procedure Cermak P, Rautkari L, Horacek P, et al (2015) Analysis of Dimensional Stability of Thermally Modified Wood Affected by Re-Wetting Cycles. BioResources 10:. https://doi.org/10.15376/biores.10.2.3242-3253 De Micco V, Balzano A, Wheeler EA, Baas P (2016) TYLOSES AND GUMS: A REVIEW OF STRUCTURE, FUNCTION AND OCCURRENCE OF VESSEL OCCLUSIONS. IAWA J 37:186–205. https://doi.org/10.1163/22941932-20160130 Dunn OJ (1961) Multiple Comparisons among Means. Journal of the American Statistical Association 56:52–64. https://doi.org/10.1080/01621459.1961.10482090 Dzurenda L, Dudiak M (2021) Cross-correlation of color and acidity of wet beech wood in the process of thermal treatment with saturated steam. WR 66:105–116. https://doi.org/10.37763/wr.1336-4561/66.1.105116 Esteves B, Velez Marques A, Domingos I, Pereira H (2013) Chemical changes of heat treated pine and eucalypt wood monitored by FTIR. Maderas Ciencia y tecnología 15:245–258. https://doi.org/10.4067/S0718-221X2013005000020 Esteves B, Videira R, Pereira H (2011) Chemistry and ecotoxicity of heat-treated pine wood extractives. Wood Sci Technol 45:661–676. https://doi.org/10.1007/s00226-010-0356-0 Fredriksson M, Digaitis R, Engqvist J, Thybring EE (2023) Effect of targeted acetylation on wood–water interactions at high moisture states. Cellulose. https://doi.org/10.1007/s10570-023-05678-8 Fredriksson M, Thybring EE (2019) On sorption hysteresis in wood: Separating hysteresis in cell wall water and capillary water in the full moisture range. PLOS ONE 14:e0225111. https://doi.org/10.1371/journal.pone.0225111 Fredriksson M, Thygesen LG (2017) The states of water in Norway spruce (Picea abies (L.) Karst.) studied by low-field nuclear magnetic resonance (LFNMR) relaxometry: assignment of free-water populations based on quantitative wood anatomy. Holzforschung 71:77–90. https://doi.org/10.1515/hf-2016-0044 Guo X, Wu Y, Yan N (2018a) In situ micro-FTIR observation of molecular association of adsorbed water with heat-treated wood. Wood Sci Technol 52:971–985. https://doi.org/10.1007/s00226-018-1020-3 Guo Y, Zhang M, Xie Y, et al (2018b) Effect of Thermal Treatment on the Heat of Vaporization of Bound Water by NMR and DSC Analysis. BioResources 13:5534–5542 Guo Y, Zhang M, Xie Y, et al (2018c) Effect of Thermal Treatment on the Heat of Vaporization of Bound Water by NMR and DSC Analysis. BioResources 13:5534–5542 Hill C, Altgen M, Rautkari L (2021) Thermal modification of wood—a review: chemical changes and hygroscopicity. J Mater Sci 56:6581–6614. https://doi.org/10.1007/s10853-020-05722-z Hill CAS, Ramsay J, Keating B, et al (2012) The water vapour sorption properties of thermally modified and densified wood. J Mater Sci 47:3191–3197. https://doi.org/10.1007/s10853-011-6154-8 Hofmann T, Rétfalvi T, Albert L, Niemz P (2008) Investigation of the chemical changes in the structure of wood thermally modified within a nitrogen atmosphere autoclave. WOOD RESEARCH 53:1–14 Holmbom B (1999) Extractives. In: Sjöström E, Alén R (eds) Analytical Methods in Wood Chemistry, Pulping, and Papermaking. Springer, Berlin, Heidelberg, pp 125–148 Hsieh CC, Cannella D, Jørgensen H, et al (2014) Cellulase inhibition by high concentrations of monosaccharides. J Agric Food Chem 62:3800–3805. https://doi.org/10.1021/jf5012962 Huang W, Pedersen NB, Fredriksson M, Thygesen LG (2019) Annual variations in Norway spruce xylem studied using infrared micro-spectroscopy. Forests 10:164. https://doi.org/10.3390/f10020164 Istratov AA, Vyvenko OF (1999) Exponential analysis in physical phenomena. Review of Scientific Instruments 70:1233–1257. https://doi.org/10.1063/1.1149581 Källbom S, Sedighi Moghaddam M, Wålinder MEP (2018) Liquid sorption, swelling and surface energy properties of unmodified and thermally modified Scots pine heartwood after extraction. Holzforschung 72:251–258. https://doi.org/10.1515/hf-2017-0064 Kekkonen PM, Telkki V-V, Jokisaari J (2010) Effect of Thermal Modification on Wood Cell Structures Observed by Pulsed-Field-Gradient Stimulated-Echo NMR. J Phys Chem C 114:18693–18697. https://doi.org/10.1021/jp1060304 Kocaefe D, Poncsak S, Boluk Y (2008) Effect of thermal treatment on the chemical composition and mechanical properties of birch and aspen. BioResources 3:. https://doi.org/10.15376/biores.3.2.517-537 Lupoi JS, Gjersing E, Davis MF (2015) Evaluating Lignocellulosic Biomass, Its Derivatives, and Downstream Products with Raman Spectroscopy. Front Bioeng Biotechnol 3:. https://doi.org/10.3389/fbioe.2015.00050 Mecca M, Todaro L, Giudice VL, et al (2021) GC-MS and SPME Techniques Highlighted Contrasting Chemical Behaviour in the Water Extractives of Modified Castanea sativa Mill. and Fagus sylvatica L. Wood. Forests. https://doi.org/10.3390/f12080986 Nhien LC, Long NVD, Kim S, Lee M (2017) Techno-economic assessment of hybrid extraction and distillation processes for furfural production from lignocellulosic biomass. Biotechnology for Biofuels 10:81. https://doi.org/10.1186/s13068-017-0767-3 Özgenç Ö, Durmaz S, Boyaci IH, Eksi-Kocak H (2017) Determination of chemical changes in heat-treated wood using ATR-FTIR and FT Raman spectrometry. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 171:395–400. https://doi.org/10.1016/j.saa.2016.08.026 Pedersen NB (2015) Microscopic and Spectroscopic Characterisation of Waterlogged Archaeological Softwood from Anoxic Environments Sluiter A, Ruiz R, Scarlata C, et al (2005) Determination of extractives in biomass. Laboratory analytical procedure (LAP) 1617:1–16 Sundqvist B, Karlsson O, Westermark U (2006) Determination of formic-acid and acetic acid concentrations formed during hydrothermal treatment of birch wood and its relation to colour, strength and hardness. Wood Sci Technol 40:549–561. https://doi.org/10.1007/s00226-006-0071-z Thybring EE, Digaitis R, Nord-Larsen T, et al (2020) How much water can wood cell walls hold? A triangulation approach to determine the maximum cell wall moisture content. PLOS ONE 15:e0238319. https://doi.org/10.1371/journal.pone.0238319 Thybring EE, Fredriksson M (2021) Wood Modification as a Tool to Understand Moisture in Wood. Forests 12:372. https://doi.org/10.3390/f12030372 Tjeerdsma B, Militz H (2005a) Chemical changes in hydrothermal treated wood: FTIR analysis of combined hydrothermal and dry heat-treated wood. Holz als Roh- und Werkstoff 63:102–111. https://doi.org/10.1007/s00107-004-0532-8 Tjeerdsma B, Militz H (2005b) Chemical changes in hydrothermal treated wood: FTIR analysis of combined hydrothermal and dry heat-treated wood. Holz als Roh- und Werkstoff 63:102–111. https://doi.org/10.1007/s00107-004-0532-8 Vek V, Oven P, Poljanšek I (2016) Review on Lipophilic and Hydrophilic Extractives in Tissues of Common Beech. Drvna Industrija 67:85–96. https://doi.org/10.5552/DRIND.2016.1511 Verkasalo E, Roitto M, Möttönen V, et al (2022) Extractives of Tree Biomass of Scots Pine (Pinus sylvestris L.) for Biorefining in Four Climatic Regions in Finland—Lipophilic Compounds, Stilbenes, and Lignans. Forests 13:779. https://doi.org/10.3390/f13050779 Windeisen E, Strobel C, Wegener G (2007) Chemical changes during the production of thermo-treated beech wood. Wood Sci Technol 41:523–536. https://doi.org/10.1007/s00226-007-0146-5 Zelinka SL, Altgen M, Emmerich L, et al (2022) Review of Wood Modification and Wood Functionalization Technologies. Forests 13:1004. https://doi.org/10.3390/f13071004 Statements & Declarations Additional Declarations The authors declare no competing interests. Supplementary Files DeLigneetalSupplementaryinformation.pdf De Ligne et al - Supplementary information Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5191132","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":361387455,"identity":"db646ed4-8bc9-4ccc-88f8-5ebc25b2eabe","order_by":0,"name":"Liselotte De Ligne","email":"data:image/png;base64,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","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Liselotte","middleName":"","lastName":"De Ligne","suffix":""},{"id":361387456,"identity":"8e1af0fe-c2a7-422a-8682-1e61238a4a50","order_by":1,"name":"Maria Fredriksson","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"","lastName":"Fredriksson","suffix":""},{"id":361387457,"identity":"1a224ac9-4c5b-4fc3-bdac-c7c9b9b361c2","order_by":2,"name":"Lisbeth G. Thygesen","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Lisbeth","middleName":"G.","lastName":"Thygesen","suffix":""},{"id":361387458,"identity":"a770ca1d-d8f0-4c10-be58-047abab2ceb3","order_by":3,"name":"Emil E. Thybring","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Emil","middleName":"E.","lastName":"Thybring","suffix":""}],"badges":[],"createdAt":"2024-10-02 06:54:13","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-5191132/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5191132/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":66067861,"identity":"3cff9721-782e-455c-bb4a-ae56c11f71a1","added_by":"auto","created_at":"2024-10-07 11:35:44","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":87732,"visible":true,"origin":"","legend":"\u003cp\u003ea) wt% of extractives removed during extraction, b) pH measurements of wood powder at a powder:water ratio of 1:6 and c) pH measurements of water in which a wood specimen had been immersed, thus containing volatile components that “leached” from the wood\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/31f35021aca029ce4f4a4902.png"},{"id":66067864,"identity":"f6c3f9f3-9d5e-43f7-bb40-74d426bd2144","added_by":"auto","created_at":"2024-10-07 11:35:44","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":286869,"visible":true,"origin":"","legend":"\u003cp\u003eMean ATR-FTIR spectra for non-extracted beech and Scots pine, with the bands that mostly contributed to the difference between thermally modified and untreated wood highlighted. (1) 3600-3300 cm\u003csup\u003e-1\u003c/sup\u003e OH stretching from alcohols, with 3350-3000 cm\u003csup\u003e-1 \u003c/sup\u003eOH stretching of carboxylic acids, (2) 2925 cm\u003csup\u003e-1\u003c/sup\u003e and 2852 cm\u003csup\u003e-1\u003c/sup\u003e, (3) 1739 cm\u003csup\u003e-1\u003c/sup\u003e carbonyl peak, (4) aromatic peak shift: 1502 cm\u003csup\u003e-1\u003c/sup\u003e (untreated) and 1515 cm\u003csup\u003e-1\u003c/sup\u003e (thermally modified, (5) 1228 cm\u003csup\u003e-1 \u003c/sup\u003etentatively assigned to C-O absorption, (6,7,8) bands related to cellulose and (9) 977 cm\u003csup\u003e-1\u003c/sup\u003e C-O vibrations cellulose. Assignments based on (Esteves et al. 2013; Pedersen 2015; Lupoi et al. 2015; Özgenç et al. 2017; Huang et al. 2019). TT = thermally modified wood\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/632a1a41ed3cc2493c783411.png"},{"id":66069271,"identity":"e82146c7-a208-410c-a577-307d97d1387f","added_by":"auto","created_at":"2024-10-07 11:43:44","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":173598,"visible":true,"origin":"","legend":"\u003cp\u003eATR-FTIR peak ratios: 1056/1107 ratio represents hemicellulose relative to holocellulose, 1502/1107 ratio represents lignin relative to holocellulose and 1739/1502 represent the carbonyl peak of hemicelluloses relative to lignin. Error bars represent standard deviation of the mean for four replicate spectra of the same specimen. a) Beech, b) Scots pine. Peak ratios that were statistically different are indicated with a different letter (a, b)\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/9d6a1d0023df174c3f4212ec.png"},{"id":66067872,"identity":"ba9de093-9fa0-44c8-8441-f9b73cabbde1","added_by":"auto","created_at":"2024-10-07 11:35:44","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":130040,"visible":true,"origin":"","legend":"\u003cp\u003ePCA of the ATR-FTIR spectra: score plots for beech and Scots pine\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/5a09f864fedd7a60620577f4.png"},{"id":66069270,"identity":"49a5f57a-584f-4a0a-960b-838085b04273","added_by":"auto","created_at":"2024-10-07 11:43:44","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":709332,"visible":true,"origin":"","legend":"\u003cp\u003eLoading profiles of PC1 for beech, with the bands that mostly contributed to the difference between thermally modified and untreated wood highlighted. (1a) 3600-3300 cm\u003csup\u003e-1\u003c/sup\u003e OH stretching from alcohols, (1b) 3350-3000 cm\u003csup\u003e-1 \u003c/sup\u003eOH stretching of carboxylic acids, (2) 2925 cm\u003csup\u003e-1\u003c/sup\u003e and 2852 cm\u003csup\u003e-1\u003c/sup\u003e, (3) 1739 cm\u003csup\u003e-1\u003c/sup\u003e carbonyl peak, (4) aromatic peak shift: 1502 cm\u003csup\u003e-1\u003c/sup\u003e (untreated) and 1515 cm\u003csup\u003e-1\u003c/sup\u003e (thermally modified, (5) 1228 cm\u003csup\u003e-1 \u003c/sup\u003etentatively assigned to C-O absorption, (6,7,8) bands related to cellulose and (9) 977 cm\u003csup\u003e-1\u003c/sup\u003e C-O vibrations cellulose. Assignments based on (Esteves et al. 2013; Pedersen 2015; Lupoi et al. 2015; Özgenç et al. 2017; Huang et al. 2019). TT = thermally modified wood\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/cddd8d97f65718bd8e354dfd.png"},{"id":66069265,"identity":"bdad61c6-9a25-4ecc-8515-3659ea58bdd8","added_by":"auto","created_at":"2024-10-07 11:43:44","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":108544,"visible":true,"origin":"","legend":"\u003cp\u003eLumen areas (µ²) of TT, TT\u003csub\u003eevap\u003c/sub\u003e and untreated control of beech and Scots pine earlywood and latewood assessed with light microscopy and image analysis.\u0026nbsp; Data for TT\u003csub\u003eevap\u003c/sub\u003e beech earlywood are missing, as we did not succeed in extracting microtome slices due to the frailty of the TT\u003csub\u003eevap\u003c/sub\u003e earlywood. Lumen areas that were statistically different are indicated with a different letter (a, b, c).\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/3425cf1bd9bdfaaaa8131363.png"},{"id":66067869,"identity":"ab91fc80-66a9-4ae5-96dd-9f92692f2634","added_by":"auto","created_at":"2024-10-07 11:35:44","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":284102,"visible":true,"origin":"","legend":"\u003cp\u003eLight microscopy images at 40x magnification\u003cem\u003e \u003c/em\u003eof TT, TT\u003csub\u003eevap\u003c/sub\u003e and non-modified (untr) beech latewood (non-extracted) and earlywood (non-extracted), as well as two replicates of TT\u003csub\u003eevap\u003c/sub\u003e beech latewood (water-extracted), showing the large lumen sizes for both replicates. Note that only the vessels that are unruptured and completely in the field of view are included in the analysis (indicated with an x)\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/22057303fb2109faeeb82f44.png"},{"id":66069268,"identity":"9de253b3-e7de-4513-b3f2-84b834de4b97","added_by":"auto","created_at":"2024-10-07 11:43:44","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":86509,"visible":true,"origin":"","legend":"\u003cp\u003eContinuous \u003cem\u003eT\u003c/em\u003e\u003csub\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sub\u003e distributions for non-extracted beech and Scots pine\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/d6102a37da73068f0e0f3105.png"},{"id":66067866,"identity":"1141760d-3f91-4563-8aa8-af859a980cd5","added_by":"auto","created_at":"2024-10-07 11:35:44","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":91005,"visible":true,"origin":"","legend":"\u003cp\u003eMean values of \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e (ms) determined by continuous curve fitting for a) cell wall peak (5 replicates), b) lumen peak (5 replicates) and c) peak of liquid water specimen in which a wood specimen had been immersed: volatiles peak (3 replicates)\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/67ef8b7140a7d8c6d9305273.png"},{"id":66069267,"identity":"a2a72cc0-b547-4a8f-86e6-b6795522d09b","added_by":"auto","created_at":"2024-10-07 11:43:44","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":56474,"visible":true,"origin":"","legend":"\u003cp\u003eMean values of cell wall moisture content (%), in the water-saturated state assessed with DSC and LFNMR\u003c/p\u003e","description":"","filename":"image10.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/ce1bec958384c147d5c46e9e.png"},{"id":66069551,"identity":"7ce65c95-bc7d-47bf-b360-9349ad3ff86a","added_by":"auto","created_at":"2024-10-07 11:51:44","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":78400,"visible":true,"origin":"","legend":"\u003cp\u003eMoisture concentration (kg/m³) and moisture content (%) of TT, TT\u003csub\u003eevap\u003c/sub\u003e and non-modified (untr) beech and Scots pine in water-saturated state\u003c/p\u003e","description":"","filename":"image11.png","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/5deada0b741817ef46b140a8.png"},{"id":66070449,"identity":"73246b83-8349-4472-b79b-d3ba29a403a4","added_by":"auto","created_at":"2024-10-07 11:59:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2802641,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/a5e5d305-4693-422b-a0e2-8e56ff74a526.pdf"},{"id":66069266,"identity":"9ba138c5-e106-4850-937b-7a8d42546561","added_by":"auto","created_at":"2024-10-07 11:43:44","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":657500,"visible":true,"origin":"","legend":"\u003cp\u003eDe Ligne et al - Supplementary information\u003c/p\u003e","description":"","filename":"DeLigneetalSupplementaryinformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5191132/v1/a06df6df595237a942c9cd80.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eInfluence of volatile degradation products from thermal wood modification on wood-water interactions\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWood protection used to focus mainly on selecting naturally durable wood species, often from tropical regions, or treating non-durable wood species with fungicidal wood preservatives. General awareness of the negative impact of biocidal products on the environment initiated a new way of thinking about wood protection. Modern methods to protect wood from degradation aim at increasing the material's service life by removing or reducing a prerequisite for fungal degradation: water within the material. This can, for instance, be done by thermal modification. During thermal modification, wood is heated at high temperatures, usually between 160 and 240\u0026deg;C, to improve its dimensional stability and durability (Hill et al. 2021; Zelinka et al. 2022). Several process parameters can be varies: the duration and temperature applied during treatment steps, the treatment atmosphere (inert gas, air, vacuum), the use of \u0026lsquo;wet\u0026rsquo; or \u0026lsquo;dry\u0026rsquo; systems and open or closed systems (Hill et al. 2021). Thermal degradation reactions in wood are complex and depend on the wood species and the process conditions applied (Hill et al. 2021; Zelinka et al. 2022).\u003c/p\u003e \u003cp\u003eDuring thermal treatment, degradation products ensue from the breakdown of cell wall polymers and wood extractives. For instance, the amount of hemicellulose degradation products, such as acetic acid and furfurals, increases with increased treatment temperature (Candelier et al. 2011) or with increased vapour pressure in closed systems (Altgen et al. 2016). Within a closed system, degradation products accumulate within the treatment vessel, producing acidic conditions that catalyse the depolymerisation of polysaccharides (Sundqvist et al. 2006; Altgen et al. 2016). In an open system, volatile by-products of degradation and exhaust gases are expected to be expelled into the atmosphere. This is likely also the case in a closed system process if the vessel pressure is released before the cooling step. Since the presence of volatile degradation products can affect the chemical reactions taking place, the venting off of volatile degradation products in the cooling step might affect the wood chemistry and wood-water interactions of thermally modified wood.\u003c/p\u003e \u003cp\u003eIn this paper, we investigate the impact of retaining or evaporating volatile degradation products in a closed thermal treatment process on the wood chemistry and wood-water interactions of thermally modified beech and Scots pine.\u003c/p\u003e"},{"header":"Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eWood specimens\u003c/h2\u003e \u003cp\u003eBoards (25 x 100 x 1200 mm\u0026sup3;) of Scots pine (\u003cem\u003ePinus sylvestris\u003c/em\u003e L.) and beech (\u003cem\u003eFagus sylvatica\u003c/em\u003e L.) were exposed to two thermal modification treatments based on the ThermoTreat 2.0 process. This closed thermal modification process modifies the wood by hydrolysis at high pressures but lower temperatures than typical open processes, whereby pressure acts as a substitute for temperature. In the high-pressure thermal treatment (TT), the wood was heated to 170\u0026deg;C over 4 hours and kept at this temperature for 2 hours at 15\u0026ndash;18 bar pressure before cooling down over 8 hours while the pressure was kept high (\u0026gt;\u0026thinsp;12 bar). The procedure for the second thermal treatment follows the same steps, but the cooling steps occurred after the pressure had been lowered to 1\u0026ndash;3 bar, allowing volatile degradation products to evaporate (TT\u003csub\u003eevap\u003c/sub\u003e). In total, 60 untreated controls, 60 specimens of the higher and 60 of the lower pressure thermally modified per wood species were then cut to a size of 5 x 5 x 10 mm\u003csup\u003e3\u003c/sup\u003e, where 10 mm was in the longitudinal direction.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eExtraction\u003c/h3\u003e\n\u003cp\u003eFor each type of extraction protocol and wood species, 20 specimens per heat-treated material and 30 specimens per untreated control material (5 x 5 x 10 mm\u003csup\u003e3\u003c/sup\u003e) were oven-dried at 60\u0026deg;C in a BINDER VD23 vacuum oven (BINDER GmbH, Tuttlingen, Germany) for 24 hours, to determine the oven-dry mass of the specimens before extraction (\u003cem\u003em\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e). After each extraction protocol, the specimens were oven-dried once more at 60\u0026deg;C in the vacuum oven for 24 hours to determine the oven-dry weight of the specimens after extraction (\u003cem\u003em\u003c/em\u003e\u003csub\u003e1\u003c/sub\u003e) (0.1 mg resolution, QUINTIX224-1S, Sartorius Lab Instruments GmbH \u0026amp; Co. KG, Goettingen, Germany). Note that the 20 or 30 specimens of the same type were weighed together before and after extraction, as labelling was impossible. The average amount of extractives \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e removed by the extraction protocol was calculated as follows:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}=\\:\\frac{{m}_{0}-\\:{m}_{1}}{{m}_{0}}\\:.\\:100\\%$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\n\u003ch3\u003eMulti-solvent extraction\u003c/h3\u003e\n\u003cp\u003eThe specimens were placed in cellulose thimbles, with all specimens of one treatment in one thimble. First, the specimens were subjected to multi-solvent extraction with a 9:1 acetone-water mixture for 24 hours. Then, the specimens were subjected to multi-solvent extraction with a 1:2 ethanol-toluene mixture for 24 hours (Holmbom 1999; Sluiter et al. 2005). While setting up the second extraction, the specimens were left to evaporate in the fumehood.\u003c/p\u003e\n\u003ch3\u003eWater extraction\u003c/h3\u003e\n\u003cp\u003eThe specimens were leached/water-extracted in accordance with the procedure described in EN 84 (1997). The specimens were placed in two-neck round bottom flasks (volume of 100 ml), with the top neck connected to a vacuum pump RZ 2.5 (Vacuubrand, GMBH, Werthelm, Germany) and the side-neck closed off with a rubber stopper. A vacuum corresponding to 4 kPa was installed for 15 minutes, after which deionised water was injected into the flask through the rubber stopper with a syringe. After 5 minutes of immersion, the vacuum was lifted, and the immersed specimens were left for 2 hours, after which the water was removed, and the bottom flasks were refilled with fresh deionised water. The specimens remained submerged for 14 days in deionised water. The water was changed at the end of the first and second day of immersion and then seven more times during the remaining 12 days.\u003c/p\u003e\n\u003ch3\u003epH measurements\u003c/h3\u003e\n\u003cp\u003eFor each sample type (TT, TT\u003csub\u003eevap\u003c/sub\u003e, untreated) and extraction method (Non-, Multi-solvent- and Water extracted), a pH measurement was performed (1 replicate). Wood specimens were milled in a Retsch MM400 mill (Retsch, Germany). For each specimen, 0.7 g of wood powder was added to a 50 ml tube. Deionised water was added with a pipette at a weight ratio of 1:6 (wood:water). A magnet was placed inside each tube, and a rotation speed of 300 rpm was installed to ensure adequate mixing of the wood powder and water. The pH was measured with a pH sensor (827 pH lab, Metrohm, Switzerland).\u003c/p\u003e \u003cp\u003eIn addition, pH measurements were performed on the water in which the LFNMR wood specimens (see 2.6) had been immersed (1 replicate per thermal treatment and extraction method). It is assumed that the water contains leached volatile components from the wood. The water was transferred from the Eppendorf tubes containing the wood specimens into a broader glass vile and measured with the pH sensor.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eAttenuated total reflectance Fourier transform infrared spectroscopy\u003c/h2\u003e \u003cp\u003eAttenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) spectroscopy was applied to determine changes in the wood polymers due to the two thermal modification procedures. Two replicates of the non-extracted and multi-solvent-extracted beech and Scots pine specimens of both thermally modified wood and untreated controls were dried in a BINDER VD23 vacuum oven (BINDER GmbH, Tuttlingen, Germany) for 24 hours at 60\u0026deg;C. The specimens were placed on the ATR-FTIR platform and covered with a custom-made conditioning chamber with a continuous airflow of 0% RH. FTIR measurements were taken of each specimen using a Nicolet 6700 FT-IR, Pike Technologies GladiATR diamond spectrometer (Thermo Scientific, Waltham, MA, USA), with a working temperature of 25\u0026deg;C. The specimens were placed on the FTIR crystal using the openings on the side of the humidity chamber, to cause minimal disturbance to the targeted 0% RH conditions during measurement. The spectral range included was 3700\u0026ndash;500 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and spectra were obtained using 64 scans (128 for the background) with a resolution of 2 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Atmospheric correction was applied in OMNIC (version 9.8.372, Thermo Fisher Scientific Inc.). Junk data between 1800 and 2700 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e was excluded.\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eFor each replicate, the borders of 13 regions of interest were identified and linear baseline correction was applied to these individual peaks. A baseline-corrected spectrum was obtained for each replicate and normalised by dividing all absorbance values by the absorbance value of the highest peak (around 1026 cm\u003csup\u003e-1\u003c/sup\u003e). A smoothed spectrum was obtained for each specimen after smoothing with a Savitzky-Golay filter (polynomial\u0026thinsp;=\u0026thinsp;3, window\u0026thinsp;=\u0026thinsp;9). Baseline corrected areas were taken (before normalisation) of four reference peaks: holocellulose (1107 cm\u003csup\u003e-1\u003c/sup\u003e), hemicellulose (1056 cm\u003csup\u003e-1\u003c/sup\u003e), lignin (1502 cm\u003csup\u003e-1\u003c/sup\u003e) and the carbonyl peak in hemicellulose (1739 cm\u003csup\u003e-1\u003c/sup\u003e), and peak ratios were determined.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003ePrincipal component analysis (PCA) was applied to the smoothed FTIR spectra to determine the impact of thermal modification on the FTIR spectra of beech and Scots pine and to identify which wavenumbers in the FTIR spectra were related to thermal modification (Python, sklearn.decomposition.PCA package).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eLight microscopy\u003c/h3\u003e\n\u003cp\u003eLight microscopy images were taken from microtome slices to determine the cross-sectional tracheid cell lumen areas (Scots pine) and vessel areas (beech) per cell. Specimens of Scots pine and beech were cut to a size of 2 x 2 x 10 mm\u0026sup3; for each treatment and extraction method with two replicates per material type. The wood specimens were placed in a Leica RM2255 automated rotary microtome (Leica Biosystems, United States) and slices were cut at 30 and 40 \u0026micro;m thickness for thermally treated Scots pine and beech, respectively. The untreated wood specimens were cut at 20 \u0026micro;m thickness and stained with Toluidine Blue. The thermally modified wood slices were not died as the darkened wood cells due to modification allowed sufficient visualisation. The wood slices were placed in a Leica compound microscope, and images were taken at 40x magnification with a Leica ICC50 camera (Leica Microsystems). For each specimen, an image was taken from an earlywood (EW) and a latewood zone (LW), taken at the beginning and end of the growth ring, respectively. Since the thermally modified wood was brittle, cell tissue ruptured easily, especially near the sides of the slice, and the images were taken where the tissue was most intact. Images were pre-processed in Fiji (Schindelin et al. 2012) to obtain binary images. Incomplete cell areas were manually coloured to be excluded from the measurements, and tracheid/lumen cell areas were determined in WinCELL (Regent Instruments Inc., Canada).\u003c/p\u003e\n\u003ch3\u003eLow-Field Nuclear Magnetic Resonance measurements\u003c/h3\u003e\n\u003cp\u003eLow-Field Nuclear Magnetic Resonance (LFNMR) spectroscopy was applied to gain knowledge on the water populations in the wood specimens in water-saturated state. Five replicates (5 x 5 x 10 mm\u0026sup3;) for each treatment, extraction method and species were placed in two-neck round bottom flasks, with the top neck connected to a vacuum pump RZ 2.5 (Vacuubrand, GMBH, Werthelm, Germany) and the side-neck closed off with a rubber stopper. A vacuum corresponding to 4 kPa was installed for 15 minutes, after which deionised water was injected into the flask through the rubber stopper with a syringe. After 5 minutes of immersion, the vacuum was lifted, and the immersed specimens were transferred to water-filled specimen tubes and left overnight. Then, the specimens were put in a pre-weighed glass LFNMR tube, after the excess surface water was removed by dapping on a wet cloth (Fredriksson and Thygesen 2017). The wet mass of the specimen was measured with a resolution of 0.1 mg. A solid Teflon rod with an O-ring sealing was inserted into the glass LFNMR tube to fill the remaining space, thereby limiting water evaporation from the specimen during the experiment. The glass LFNMR tube containing the specimen was placed in a Bruker mq20 minispec with a 0.47 T permanent magnet (Bruker, Billerica, MA, USA) to perform LFNMR measurements. The temperature in the LFNMR around the specimen was kept constant at 22\u0026deg;C by water circulation in the probe using a refrigerated and heating circulator (Julabo GmbH, Seelbach, Germany). The specimen was allowed to thermally equilibrate in the instrument for two minutes before starting each measurement. The Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence was used to measure the spin-spin relaxation time (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e) of the specimens with a pulse separation (\u003cem\u003eτ\u003c/em\u003e) of 0.1 ms, 8000 echoes, 32 scans and a recycle delay of 30 s. Exponential decay analysis (Istratov and Vyvenko 1999) was applied to the recorded LFNMR decay curves to give smooth, continuous distributions of \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e relaxation times. 200 exponentials were fitted to the data with fixed, logarithmically spaced T2 relaxation times from the first to the last time point (0.2 to 2500 ms). For each peak, the relative sum of amplitudes of the exponential components and their \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values corresponding to maximum peak intensity were determined. After the experiment, the LFNMR specimens were weighed (\u003cem\u003em\u003c/em\u003e\u003csub\u003e\u003cem\u003ew\u003c/em\u003e\u003c/sub\u003e), oven-dried at 60\u0026deg;C in a BINDER VD23 vacuum oven (BINDER GmbH, Tuttlingen, Germany) for 24 hours and weighed again (\u003cem\u003em\u003c/em\u003e\u003csub\u003e\u003cem\u003ed\u003c/em\u003e\u003c/sub\u003e).\u003c/p\u003e \u003cp\u003eWhen researching the impact of thermal modification, it is interesting to assess both total moisture concentration and moisture content of the wood specimens, as the oven-dry weight of thermally modified wood is lower than untreated wood. The moisture concentration (\u003cem\u003ec\u003c/em\u003e\u003csub\u003esat\u003c/sub\u003e) of the LFNMR specimens after water saturation was calculated as follows, with \u003cem\u003eV\u003c/em\u003e the volume of the wood specimens:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{c}_{sat}=\\:\\frac{{m}_{w}-\\:{m}_{d}}{V}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe moisture content (ω) of the LFNMR specimens after water saturation was calculated as follows:\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:\\omega\\:=\\:\\frac{{m}_{w}-\\:{m}_{d}}{{m}_{d}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eTo determine the influence of water-soluble components on LFNMR, three specimens per material and extraction type were water-saturated as described above and stored in Eppendorf tubes in a refrigerator. After 18 days of immersion, a 1 ml syringe was used to suck up water from the Eppendorf tube in which the specimen had been immersed. The liquid was passed through a Nylon syringe filter (MicroLab Scientific, diameter 13 mm, pore size 0.45 \u0026micro;m) while transferred to a new Eppendorf. 350 \u0026micro;l of liquid was transferred into a glass LFNMR tube. The Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence was used to measure the spin-spin relaxation time (\u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e) of the specimens with a pulse separation (\u003cem\u003eτ\u003c/em\u003e) of 0.1 ms, 10000 echoes, 32 scans and a recycle delay of 30 s. Additionally, MilliQ water was assessed with LFNMR. For the MilliQ water and water from the multi-solvent extracted specimens, 32000 echoes were necessary.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eCharacterisation of maximum cell wall moisture content with DSC\u003c/h2\u003e \u003cp\u003eThe cell wall moisture content in water-saturated state was determined with Differential Scanning Calorimetry (DSC) as described by Fredriksson et al. (Fredriksson et al. 2023). Here, the maximum cell wall moisture content can be determined by measuring the heat required to thaw a frozen water-saturated specimen. Since the water outside the cell wall freezes, but the cell wall water does not, the total moisture content can be separated into capillary water and cell wall water, respectively (Fredriksson and Thygesen 2017; Thybring et al. 2020). Specimens (5 x 5 x 10 mm\u003csup\u003e3\u003c/sup\u003e) from the LFNMR experiments of all material types were vacuum-saturated with water. From these, circular discs with a diameter of approximately 5 mm and a height of 1.4 mm were cut to fit DSC-pans of 40 \u0026micro;l. The specimens were cut while keeping them in water to avoid drying. Excess water was then removed by wiping the surfaces of the specimen with a moist cloth (Wettex, Freudenberg Home and Cleaning Solutions GmbH, Malm\u0026ouml;, Sweden). Five replicates were cut for each material type, so 90 specimens were measured.\u003c/p\u003e \u003cp\u003eAfter cutting, each specimen was placed in a pre-weighed DSC pan, which was then hermetically sealed. The water-saturated mass, \u003cem\u003em\u003c/em\u003e\u003csub\u003esat\u003c/sub\u003e (g), was determined. Then, the specimens were placed in the sample robot of the DSC (DSC 3+, Mettler Toledo AG, Schwerzenbach, Switzerland) where the following method was applied: \u0026minus; 20\u0026deg;C for 5 min, a ramp from \u0026minus;\u0026thinsp;20 to 20\u0026deg;C with 2\u0026deg;C min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and finally a temperature stabilisation period of 5 min at 20\u0026deg;C. After the measurements, the lids of the DSC pans were pierced with a needle and placed in a vacuum oven at 60\u0026deg;C for 24 h. Each specimen's dry mass, \u003cem\u003em\u003c/em\u003e\u003csub\u003ed\u003c/sub\u003e, was determined after temperature stabilisation in a desiccator with molecular sieves. The cell wall moisture content, \u003cem\u003eω\u003c/em\u003e\u003csub\u003ecw,\u003c/sub\u003e was then evaluated (Fredriksson and Thybring 2019; Fredriksson et al. 2023) by:\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\:{\\omega\\:}_{\\text{c}\\text{w}}=\\frac{{m}_{\\text{s}\\text{a}\\text{t}}-{m}_{\\text{d}}-\\left(\\frac{\\text{Q}\\bullet\\:{m}_{\\text{s}\\text{a}\\text{t}}}{{H}_{\\text{f}}}\\right)}{{m}_{\\text{d}}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003eH\u003c/em\u003e\u003csub\u003ef\u003c/sub\u003e (J g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) is the enthalpy of fusion of water of 333.7 J g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and \u003cem\u003eQ\u003c/em\u003e (J) is the heat of melting of the capillary water, which was determined by integration of the melting peak. This integration was performed in STARe Evaluation Software (version 17.00, Mettler Toledo AG, Schwerzenbach, Switzerland). Calibration of the DSC for enthalpy of fusion was done with deionised water (melting point 0\u0026deg;C, enthalpy of fusion 333.7 J g\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eStatistical data analysis\u003c/h2\u003e \u003cp\u003eThe Kruskal\u0026ndash;Wallis H-test, a nonparametric statistical test, was applied to test whether the median values of untreated, TT and TT\u003csub\u003eevap\u003c/sub\u003e datasets differed. When the difference between medians was significant, Dunn\u0026rsquo;s multiple comparison test (Dunn 1961) was applied to pinpoint for which materials the medians differed. Benjamini-Hochberg correction (Benjamini and Hochberg 1995) was performed to control the false discovery rate (Type I error).\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eExtractive content and pH\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe thermally modified wood was clearly more acidic for both species than the non-modified wood (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb,c). The cooling down step at atmospheric pressure allowed volatile degradation products to evaporate in the case of beech (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). For Scots pine, the results are inconclusive, as no significant difference was observed in amount of extractives between TT and TT\u003csub\u003eevap\u003c/sub\u003e.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eThermally modified beech contained higher amounts of extractives than the untreated beech, both for multi-solvent and water-extracted specimens (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). For beech, more extractives were removed from TT as compared to TT\u003csub\u003eevap\u003c/sub\u003e, indicating that less volatile components were present in TT\u003csub\u003eevap\u003c/sub\u003e as compared to TT (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) and that the cooling step at lowered pressure indeed allowed volatile degradation products to evaporate. Likewise, the pH of TT was slightly more acidic than TT\u003csub\u003eevap\u003c/sub\u003e, confirming the lower presence of acidic, volatile degradation products in TT\u003csub\u003eevap\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb and c). However, the magnitude of the pH difference was different for the two methods used. For pH measurements in water in which the wood had been immersed, the difference in pH between TT and TT\u003csub\u003eevap\u003c/sub\u003e was higher than for the measurements made on wood powder. This is likely because of the high acidity of the wood compounds that remained in the wood cell wall (volatile and non-volatile) and came free due to the milling. The multi-solvent-extracted specimens all had a high pH, but there was still a difference between TT and TT\u003csub\u003eevap\u003c/sub\u003e for the pH measurements on wood powder. This high pH is not unexpected, as less volatile components remain in the wood after multi-solvent extraction, and the relative amount of acid components relative to water (pH 7) is less. Note that the pH of the thermally modified wood powder was more acidic than the non-modified wood powder both for beech and Scots pine, which is expected for thermally modified wood (Boonstra et al. 2007).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFor Scots pine, only 2\u0026ndash;3 wt% was removed after multi-solvent and water extraction and only small (\u0026lt;\u0026thinsp;1%) and no consistent differences were observed in\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e between TT, TT\u003csub\u003eevap\u003c/sub\u003e and the untreated wood. The extractive content in untreated Scots pine wood is generally around 3\u0026ndash;5% (Verkasalo et al. 2022). Likewise, the pH of the water in which the wood specimens had been immersed was similar between thermally modified and untreated Scots pine, with only minor differences of 0.2 (water extracted, non-extracted) in pH value (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). In contrast, the pH of the thermally modified, non-extracted wood powder was significantly lower than the that of the untreated wood, indicating that the wood components themselves had become more acidic after the thermal modification (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). The pH of the multi-solvent-extracted specimens was very high, both for the wood powder and the leachate.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eWood composition assessed with ATR-FTIR\u003c/h2\u003e \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e \u003ch2\u003eATR-FTIR spectra\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe ATR-FTIR spectra of non-extracted beech show clear differences among TT, TT\u003csub\u003eevap\u003c/sub\u003e and untreated (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). In the OH range, there was a clear decrease in the region 3600\u0026thinsp;\u0026minus;\u0026thinsp;3400 cm\u003csup\u003e-1\u003c/sup\u003e (peak 1 in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), assigned to OH-stretching in alcohols (Esteves et al. 2013) for thermally modified beech compared to untreated beech. For TT\u003csub\u003eevap\u003c/sub\u003e there was an increase in the region 3350\u0026thinsp;\u0026minus;\u0026thinsp;3000 cm\u003csup\u003e-1\u003c/sup\u003e (peak 2), assigned to OH-stretching of carboxyl acids (Esteves et al. 2013). The intensity of the carbonyl peak at 1739 cm\u003csup\u003e-1\u003c/sup\u003e (peak 3) had decreased for thermally modified beech as compared to untreated beech, which is an indication of the degradation of hemicelluloses, such as xylan and glucomannan (Tjeerdsma and Militz 2005a; Guo et al. 2018a). The aromatic rings in hardwood lignin (Guaiacyl - G and Syringyl - S) usually exhibit benzene ring stretching vibrations at approximately 1505 cm\u003csup\u003e-1\u003c/sup\u003e (Esteves et al. 2013), which is indeed the case for the untreated beech wood (1502 cm\u003csup\u003e-1\u003c/sup\u003e) (peak 4). For thermally modified beech, the peak maximum shifted to 1515 cm\u003csup\u003e-1\u003c/sup\u003e, which can be attributed to the breaking of aliphatic side chains in lignin and/or cross-linking by condensation reactions (Windeisen et al. 2007; Kocaefe et al. 2008; Esteves et al. 2013). There is a clear decrease at 1228 cm\u003csup\u003e-1\u003c/sup\u003e (peak 5), tentatively assigned to C-O absorption (Windeisen et al. 2007; Pedersen 2015; Huang et al. 2019), for thermally modified wood compared to untreated wood, indicating alteration and cleavage of carbonyl groups (Windeisen et al. 2007). The bands at 1161 cm\u003csup\u003e-1\u003c/sup\u003e, 1109 cm\u003csup\u003e-1\u003c/sup\u003e and 1057 cm\u003csup\u003e-1\u003c/sup\u003e (peaks 6,7 and 8), related to cellulose (Lupoi et al. 2015; \u0026Ouml;zgen\u0026ccedil; et al. 2017), increased after thermal modification. Similar observations could be made based on the ATR-FTIR spectra of the multi-solvent extracted beech samples (Suppl. Fig. S1).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eThe ATR-FTIR spectra of non-extracted Scots pine show a higher standard deviation than of beech, especially for untreated and TT\u003csub\u003eevap\u003c/sub\u003e. For TT, a clear decrease in the region 3600\u0026thinsp;\u0026minus;\u0026thinsp;3400 cm\u003csup\u003e-1\u003c/sup\u003e (peak 1), assigned to OH-stretching in alcohols (Esteves et al. 2013) could be observed, but not for TT\u003csub\u003eevap,\u003c/sub\u003e while the latter could be observed in the ATR-FTIR spectra of the multi-solvent extracted samples (Suppl. Fig. S1). Similarly, the intensity of the carbonyl peak at 1739 cm\u003csup\u003e-1\u003c/sup\u003e (peak 3) had decreased for thermally modified Scots pine as compared to untreated Scots pine, indicating the degradation of hemicelluloses, but not for TT\u003csub\u003eevap\u003c/sub\u003e. The bands at 1161 cm\u003csup\u003e-1\u003c/sup\u003e and 1109 cm\u003csup\u003e-1\u003c/sup\u003e (peaks 6 and 7), related to cellulose (Lupoi et al. 2015; \u0026Ouml;zgen\u0026ccedil; et al. 2017) increased for both thermal modifications (TT and TT\u003csub\u003eevap\u003c/sub\u003e).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003ePeak ratios\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003ePeak ratios were calculated for specific FTIR wavenumber regions representing holocellulose, hemicellulose and lignin, as well as the carbonyl peak in hemicellulose (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Overall, the standard deviation of the non-extracted peak ratios was higher than the multi-solvent-extracted peak ratios, especially for TT\u003csub\u003eevap\u003c/sub\u003e. For beech, no significant differences were found for the 1056/1107 peak ratio representing hemicellulose relative to holocellulose. Indeed, from the ATR-FTIR spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), it could be observed that both peaks increased for thermally modified beech. Similarly, no significant difference was found for the 1508/1107 peak ratio representing lignin relative to holocellulose. While there was a clear peak shift from 1502 to 1515 cm\u003csup\u003e-1\u003c/sup\u003e, the overall peak area taken for calculating the peak ratio did not change (area taken from wavenumber region 1480\u0026ndash;1530 cm\u003csup\u003e-1\u003c/sup\u003e). In contrast, the 1739/1502 peak ratio for the TT\u003csub\u003eevap\u003c/sub\u003e beech was significantly lower than the untreated wood for multi-solvent extracted specimens (Suppl. Table\u0026nbsp;1). Likewise, the non-extracted specimens followed the same pattern, but no significant difference was found with Kruskal-Wallice, due to the high standard deviation of TT\u003csub\u003eevap\u003c/sub\u003e.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFor Scots pine, similar results were found, except the non-extracted 1056/1107 peak ratio, where TT was significantly higher than TT\u003csub\u003eevap\u003c/sub\u003e and untreated wood (Suppl. Table\u0026nbsp;1). The 1739/1502 peak ratio for TT Scots pine was significantly lower than the untreated wood for non-extracted specimens (Suppl. Table\u0026nbsp;1). Interestingly, this was not the case for the multi-solvent extracted specimens, for which the median peak ratios were similar.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003ePCA of ATR-FTIR spectra\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eFor beech, untreated (untr) and thermally modified wood (TT and TT\u003csub\u003eevap\u003c/sub\u003e) were clearly distinguishable based on PC1 (56.1% of variance) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The spectral ranges that were most important for separating thermally modified beech wood from untreated beech wood based on the ATR-FTIR spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) were peak 1a, 3600\u0026thinsp;\u0026minus;\u0026thinsp;3300 cm\u003csup\u003e-1\u003c/sup\u003e OH assigned to stretching from alcohols (Esteves et al. 2013), peak 3 representing the 1739 cm\u003csup\u003e-1\u003c/sup\u003e carbonyl peak, peak 5 at 1228 cm\u003csup\u003e-1\u003c/sup\u003e tentatively assigned to C-O absorption (Windeisen et al. 2007; Pedersen 2015; Huang et al. 2019) and peak 7 and 8, related to cellulose (Lupoi et al. 2015; \u0026Ouml;zgen\u0026ccedil; et al. 2017). The increase in the region 3350\u0026thinsp;\u0026minus;\u0026thinsp;3000 cm\u003csup\u003e-1\u003c/sup\u003e (peak 2), assigned to OH-stretching of carboxyl acids (Esteves et al. 2013) is more clearly visible in the loading profiles (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) than in the average ATR-FTIR spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFor Scots pine, a distinction between thermally modified and untreated wood could also be made, based on a combination of PC3 (2.2%) and PC4 (4.6%). The PCA score plots also distinguished between multi-solvent-extracted and non-extracted beech samples, based on PC4, with mostly negative values for multi-solvent-extracted specimens.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eLight microscopy\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe light microscopy images showed significantly lower cross-sectional vessel areas per cell for thermally modified beech (TT) than untreated beech for earlywood and latewood in non-extracted samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, Suppl. Tables\u0026nbsp;2 and 3). Interestingly, larger cross-sectional vessel cell areas were found for TT earlywood and TT\u003csub\u003eevap\u003c/sub\u003e latewood for water-extracted beech specimens, but not for TT beech latewood with light microscopy (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). No significant difference in earlywood cross-sectional cell lumen areas was found between TT, TT\u003csub\u003eevap\u003c/sub\u003e and untreated Scots pine for non-extracted and water-extracted specimens (Suppl. Table\u0026nbsp;3). The latewood cell lumen areas of TT\u003csub\u003eevap\u003c/sub\u003e were significantly smaller than untreated wood (100\u0026ndash;130 \u0026micro;m\u0026sup2;) and TT (50\u0026ndash;150 \u0026micro;m\u0026sup2;), independent of extraction method (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Suppl. Table\u0026nbsp;3). The cell lumen areas clearly differed between water extracted specimens compared to the other extraction methods for TT and TT\u003csub\u003eevap\u003c/sub\u003e, except for TT beech latewood (Suppl. Tables\u0026nbsp;4 and 5). Interestingly, a significant difference was only found twice for the untreated specimens (Beech EW: Water x Non and Scots pine EW: Water x Multi-solvent).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eWood-water interactions in water-saturated state\u003c/h2\u003e \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e \u003ch2\u003eWater populations assessed with LFNMR\u003c/h2\u003e \u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eThe water populations of each material were assessed with LFNMR in water-saturated state. Continuous \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e distributions for non-extracted beech and Scots pine are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e. All continuous \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e distributions can be found in Supplementary materials, plotted per extraction method (Fig. S3) and per treatment method (Fig. S4). Thermal treatment decreased the mean \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the cell wall peak, both for beech and Scots pine \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003ea), with the decrease statistically different between TT\u003csub\u003eevap\u003c/sub\u003e and untr. (Suppl. Table\u0026nbsp;6\u0026ndash;7). TT\u003csub\u003eevap\u003c/sub\u003e had a lower \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the cell wall peak as compared to TT for beech (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003ea). While this difference was not significant, it did occur consistently for all extraction methods (Suppl. Table\u0026nbsp;7, Suppl. Figures\u0026nbsp;3 and 4). A similar effect could be observed to a lesser extent for Scots pine, but the standard deviation of TT\u003csub\u003eevap\u003c/sub\u003e was high.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eFor the non-extracted and multi-solvent-extracted specimens, the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the vessel peak of beech was slightly higher for TT than the untreated specimens, and slightly lower for TT\u003csub\u003eevap\u003c/sub\u003e, with the difference between TT and TT\u003csub\u003eevap\u003c/sub\u003e significant (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eb, Suppl. Figure\u0026nbsp;3, Suppl. Table\u0026nbsp;7). Interestingly, the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values for the water-extracted beech specimens were much higher than those of the non- and multi-solvent-extracted specimens, but only for the thermally modified wood. For Scots pine, the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the lumen peak increased after thermal modification, with a higher increase for TT\u003csub\u003eevap\u003c/sub\u003e than TT, for all extractions (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eb, Suppl. Figure\u0026nbsp;3, Suppl. Table\u0026nbsp;7).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eThe volatiles peak shows the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values for liquid water in which a wood specimen had been immersed for 18 days (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003ec). The presence of volatiles caused the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value to decrease, with non-extracted wood having lower \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values than extracted wood, thermally modified wood having lower \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values than untreated wood and deionized water having the highest \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eCell wall moisture content at saturation\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe maximum cell wall moisture content of the thermally modified beech and Scots pine was significantly lower than for untreated wood (Suppl. Tables\u0026nbsp;8\u0026ndash;9), for both LFNMR and DSC (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e). No significant difference was found between TT and TT\u003csub\u003eevap,\u003c/sub\u003e though for beech the cell wall moisture content for TT\u003csub\u003eevap\u003c/sub\u003e was consistently lower than TT. While both techniques gave similar results, the absolute values for moisture content were lower for LFNMR than for DSC measurements, which has been shown before (Thybring et al. 2020). Interestingly, the cell wall moisture content for the thermally modified samples was higher for multi-solvent extracted specimens, as shown with DSC (both beech and Scots pine) and LFNMR (only beech).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eTotal moisture concentration and moisture content in water-saturated state\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eFor thermally modified wood, it is interesting to assess both total moisture concentration and moisture content of the specimens are provided in Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e. Both for beech and Scots pine, the moisture concentration of the thermally modified specimens in the water-saturated state was lower than that of the untreated samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e), with TT\u003csub\u003eevap\u003c/sub\u003e (but not TT) significantly lower than the untreated wood for beech and both thermal treatments significantly lower for Scots pine (Suppl. Table\u0026nbsp;9). The moisture concentration of TT\u003csub\u003eevap\u003c/sub\u003e was lower than TT for beech, but not significantly. While the total moisture concentration of the thermally modified beech specimens was lower, the total moisture content values were similar to untreated wood, as the oven-dry weight of the thermally modified wood specimens was lower (Suppl. Fig. S5). Interestingly, for Scots pine, the oven-dry mass of TT\u003csub\u003eevap\u003c/sub\u003e was lower than the untreated wood, but not that of TT, which explains the lower moisture content (Suppl. Fig. S5).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003eImpact of thermal treatment on wood chemistry\u003c/h2\u003e \u003cp\u003eThermal modification leads to changes in the wood cell wall polymers. The decreased pH of the thermally treated beech and Scots pine samples is a clear indication of this (Hofmann et al. 2008; Cai et al. 2018; Dzurenda and Dudiak 2021), as (acetic) degradation products are formed during the thermal degradation of hemicellulose and lignin (Hill et al. 2021). Thermally modified beech contained higher amounts of extractives than the untreated beech, which was not the case for Scots pine. A study on steam thermal treatment under N\u003csub\u003e2\u003c/sub\u003e gas also showed a clear increase in \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e after thermal treatment for beech (from 2.4\u0026ndash;3.6 to 11.9\u0026ndash;13.7%) for ethanol-toluene extracts (Hofmann et al. 2008). For Scots pine, only small differences (\u0026lt;\u0026thinsp;1%) were observed for multi-solvent extraction, in line with two other studies on multi-solvent extraction of thermally modified Scots pine (0.5-2%) (Hofmann et al. 2008; K\u0026auml;llbom et al. 2018). Hardwoods have been shown to be more sensitive to thermodegradation than softwoods, for which degradation products appear at slightly higher temperature (Candelier et al. 2011). For beech, the lesser amount of extractives and the less acidic pH value of TT\u003csub\u003eevap\u003c/sub\u003e confirm that cooling at lowered pressure allows volatile degradation products to evaporate. In contrast, for Scots pine, there was no discernible difference in pH and extractives among the two thermal treatments. Do note that there was no mass decrease after multi-solvent extraction in untreated beech, while 2\u0026ndash;3 wt% would be expected (Hofmann et al. 2008).\u003c/p\u003e \u003cp\u003eDuring thermal treatment, the amorphous polysaccharide content degrades, predominantly affecting hemicelluloses. Hardwoods have indeed been shown to experience a greater degree of hemicellulose degradation compared to softwoods when subjected to the same conditions (Hill et al. 2021). Hardwoods degrade at lower temperatures due to the higher acetyl group content in the hemicelluloses, with the xylans of hardwoods being acetylated, in contrast to those of softwoods (Hill et al. 2021). In dry thermal treatment conditions, degradation of carbohydrates is likely to lead to dehydration products, such as furfurals (Tjeerdsma and Militz 2005b), which have the potential to polymerise and form water-insoluble hydrophobic materials in the cell wall. While not removable with water extraction, furfurals are easily extractable with toluene (Nhien et al. 2017), possibly accounting for the high \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e for multi-solvent extracted thermally modified beech. Non-modified beech wood contains 1.0% lipophilic and 3.7% hydrophilic extractives, with saturated fatty acids, fatty alcohols, and free sterols as the main extractive components in the heartwood (Vek et al. 2016). Besides hemicellulose degradation during thermal modification, degradation of wood extractives occurs, as well as migration of wood extractives towards the wood surface, if these are mobile under the thermal treatment conditions (Hill et al. 2021). However, Mecca et al. (2021) reported that there were no substantial changes in the type of compounds in beech extractives after thermal modification. The non-volatile extractive compounds in Scots pine are mainly free fatty acids, resin acids and phenolic compounds. Estevez et al. show an increase in ethanol-soluble extractives between 1-2.5% (Esteves et al. 2011). This increase in extractives was mainly attributed to phenolic compounds and anhydrosugars, with the appearance of new compounds after thermal modification, such as vanillin, vanillic acid and 3-vanillyl propanol.\u003c/p\u003e \u003cp\u003eInterestingly, the pH of the multi-solvent-extracted specimens was very high, both for beech and Scots pine. This could simply indicate that the multi-solvent extraction removes acidic products better than water extraction. That would indicate that although the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e in Scots pine was low, a certain amount of (volatile) acidic compounds were still removed during multi-solvent extraction. Another hypothesis is that part of the solvents remained in the wood after multi-solvent extraction and contributed to the higher pH, with ethanol and acetone (pH 7) the most likely solvents as toluene is a non-polar molecule.\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe wood composition was assessed with ATR-FTIR to determine the modification of wood cell wall polymers. Several indicators of thermal modification were present when comparing the ATR-FTIR spectra of TT and TT\u003csub\u003eevap\u003c/sub\u003e to the untreated wood, such as a decrease in the region 3600\u0026thinsp;\u0026minus;\u0026thinsp;3400 cm\u003csup\u003e-1\u003c/sup\u003e, assigned to OH-stretching in alcohols (Esteves et al. 2013), a decrease in intensity of the carbonyl peak (Tjeerdsma and Militz 2005a; Guo et al. 2018a) at 1739 cm\u003csup\u003e-1\u003c/sup\u003e and an increase in intensity for bands related to cellulose (Lupoi et al. 2015; \u0026Ouml;zgen\u0026ccedil; et al. 2017). Both for beech and Scots pine a decrease in the hemicellulose/lignin ratio could be observed, which is expected after thermal modification (K\u0026auml;llbom et al. 2018), though this was not the case for the multi-solvent extracted specimens of the latter. Interestingly, for thermally modified beech, the peak maximum shifted to 1515 cm\u003csup\u003e-1\u003c/sup\u003e, which can be attributed to the breaking of aliphatic side chains in lignin and/or cross-linking by condensation reactions (Windeisen et al. 2007; Kocaefe et al. 2008; Esteves et al. 2013).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eThe chemical composition of the thermally modified beech wood differed slightly between TT and TT\u003csub\u003eevap\u003c/sub\u003e based on the PCA of the FTIR spectra, with TT having more negative values than TT\u003csub\u003eevap\u003c/sub\u003e for PC1. This difference in chemical composition was also confirmed by the difference in pH and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e. We would expect the hemicellulose/lignin ratio to be higher for TT\u003csub\u003eevap\u003c/sub\u003e than TT, which is indeed the case for the non-extracted wood, but not for the multi-solvent extracted wood. For Scots pine, the ATR-FTIR peak ratio of 1739/1502 was significantly lower for TT as TT\u003csub\u003eevap\u003c/sub\u003e, which might indicate a higher level of thermal modification, while the 1056/1107 (hemicellulose/holocellulose) peak ratio was lower for TT\u003csub\u003eevap\u003c/sub\u003e than TT. In addition, only a small part of the variance in ATR-FTIR spectra was assignable to a difference between thermally modified and untreated Scots pine wood in the PCA.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eWood-water interactions in water-saturated state\u003c/h2\u003e \u003cdiv id=\"Sec27\" class=\"Section4\"\u003e \u003ch2\u003eCapillary water interactions\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the lumen peak increased after the thermal modification of Scots pine, with a higher increase for TT\u003csub\u003eevap\u003c/sub\u003e than TT for all extractions. This increase in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values for the lumen peak typically occurs in thermally modified wood, and has been shown for Scots pine (Cai et al. 2020), Norway spruce (Cai et al. 2020) and radiata pine (Guo et al. 2018b). Our results show that this increase in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value was not caused by an increase in cell lumen diameter, as no increase in tracheid lumen area was observed for thermally modified Scots pine with light microscopy imaging, and even a decrease in cell lumen area occurred in the latewood. Such decreases in lumen dimensions have been shown for thermally modified Scots pine before, in a study applying pulsed-field-gradient stimulated-echo NMR and field emission scanning electron microscopy (Kekkonen et al. 2010). Wood volume typically decreases after thermal modification, mainly due to the loss of degraded cell wall material. Additionally, there is a strong possibility that molecular restructuring within the cell wall contributes to this volume decrease (Hill et al. 2021).\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFurthermore, the increase in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the lumen peak was not linked to the presence of volatile compounds. Volatiles caused the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value to decrease, with non-extracted wood having lower \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values than extracted wood, non-extracted thermally modified wood having lower \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values than non-extracted untreated wood and deionised water having the highest \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values. \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values have indeed been shown to decrease with increasing concentration of dissolved molecules, for sugars for example (Hsieh et al. 2014), because the average molecular distance between the solute becomes shorter. As the cell lumen areas had not increased and volatile compounds have a decreasing effect on \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values, the increase in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values for the lumen peak is a clear indication of the increased hydrophobicity of the modified Scots pine. The differences in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values and pH confirmed the presence of volatile components in the leached water, supporting the hypothesis that (part of) the volatiles remained in the water as assumed before the start of the experiments.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFor beech, the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the lumen peak increased for TT and decreased for TT\u003csub\u003eevap\u003c/sub\u003e. However, like for Scots pine, the light microscopy images showed significantly lower vessel cell areas for thermally modified beech (TT) as compared to untreated beech, both for earlywood and latewood in non-extracted samples. Likewise, less volatile compounds, which have a decreasing effect on \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values, are expected in TT\u003csub\u003eevap\u003c/sub\u003e. as compared to TT. The higher \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the lumen peak for TT as compared to TT\u003csub\u003eevap\u003c/sub\u003e is, therefore, likely an indication of an increased hydrophobicity. Indeed, more hydrophobic compounds were present in TT, as confirmed by the higher \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e after multi-solvent extraction. Note that in the case of beech, the light microscopy data might be slightly skewed towards smaller vessel areas, as cross-sections that had smaller vessels were less likely to rupture and provide an undamaged imaging region for data analysis.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eInterestingly, the \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e values for the water-extracted beech specimens were much higher than those of the non- and multi-solvent-extracted specimens, but only for the thermally modified wood. Similarly, larger vessel areas were found for TT earlywood and TT\u003csub\u003eevap\u003c/sub\u003e latewood, but not for TT latewood with light microscopy. This is unexpected as thermal modification is generally expected to decrease lumen size (Kekkonen et al. 2010). However, beech is known for the formation of tyloses (De Micco et al. 2016), which could indeed be observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, non-extracted beech earlywood TT\u003csub\u003eevap\u003c/sub\u003e). It could be that the water extraction procedure, in which the samples are water extracted under vacuum pressure and kept submerged in water for 14 days, caused the tyloses in the thermally modified wood to rupture, enlarging the space for capillary water.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eCell wall moisture content at saturation\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThermal treatment decreased the mean \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the cell wall peak, both for beech and Scots pine. However, the removal of hemicelluloses due to thermal degradation is expected to increase the available space for moisture in the cell wall (Thybring and Fredriksson 2021). Presumably, the decrease in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value is linked to a decrease in pore size in the wood cell wall in water-saturated state, due to increased crosslinking within the cell wall polymers, changes in the mobility of the polymer network and bulking of thermally degraded components that have remained in the cell wall (Thybring and Fredriksson 2021; Hill et al. 2021). Such a decrease in the cell wall peak has also been shown for thermally modified radiata pine (200\u0026deg;C, 24 h, N\u003csub\u003e2\u003c/sub\u003e) (Guo et al. 2018b). However, no decrease in \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the cell wall peak was observed for thermally modified Scots pine under steam (180\u0026deg;C, 2 h, steam-treatment) (Cai et al. 2020). For beech, the TT\u003csub\u003eevap\u003c/sub\u003e had a lower \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the cell wall peak as compared to TT. Since less extractives were present in TT\u003csub\u003eevap\u003c/sub\u003e as compared to TT beech, this is not likely linked to increased bulking of thermally degraded components. Like for the increased \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the lumen peak, this could be an indication of increased hydrophobicity, related to the higher content of hydrophobic compounds in TT, as compared to TT\u003csub\u003eevap\u003c/sub\u003e. The retaining of volatile degradation products during the thermal modification process could also have affected other parameters, such as crosslinking within the cell wall polymers and changes in the mobility of the polymer network.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003eWater holding capacity\u003c/h2\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe cell wall moisture content of Scots pine reduced from around 30% to around 20% after thermal modification, as assessed with LFNMR. Similar values were found for Radiata pine in a study by Guo et al. (2018b). The cell wall moisture content for the thermally modified samples was higher for multi-solvent extracted specimens, as shown with DSC (both beech and Scots pine) and LFNMR (only beech). Multi-solvent extraction removes hydrophobic compounds and is consequently expected to increase the hydrophilicity (K\u0026auml;llbom et al. 2018). In addition to that, mass loss from extraction typically results in an increase in cell wall moisture content, as soluble components that have leached out of the cell wall free up space for water molecules. Interestingly, the cell wall moisture content of water-extracted Scots pine specimens was similar to that of untreated wood. This phenomenon has also been shown in earlier studies demonstrating that although thermal modification in dry conditions results in a reduction of (equilibrium) moisture content, the effects can be partially reversible when exposed to high RH levels or liquid water. This reversible effect has been shown before for Scots pine (Hill et al. 2012), though also for beech (Cermak et al. 2015), where the water extraction procedure in our study did not have the same effect.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eThe maximum cell wall moisture content and overall water concentration of the thermally modified wood specimens in a water-saturated state were lower than of the untreated wood. However, the overall moisture content of the specimens was similar, as the thermally modified wood specimens had a lower dry mass than the untreated specimens. The equilibrium moisture content of thermally modified wood is expected to be lower than untreated wood, as shown for water vapour sorption at relative humidities going from 0 to 95% (Hill et al. 2012; Altgen and Militz 2016). Few studies were found showing the moisture content at water saturation of thermally modified wood, though Guo et al. report an average moisture content at water saturation of about 160% for thermally modified and 170% moisture content for untreated radiata pine (Guo et al. 2018b).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eImplications and recommendations for future research\u003c/h3\u003e\n\u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eThe cooling step at lowered pressure allowed volatile degradation products to evaporate from thermally modified beech. The \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value for the lumen peak and cell wall peak as assessed with LFNMR, were higher for TT than for TT\u003csub\u003eevap\u003c/sub\u003e, indicating that retaining volatile degradation products during the thermal treatment process resulted in a more hydrophobic end-product. For Scots pine, the impact of allowing volatile degradation products to evaporate is less straightforward to interpret. No clear differences were found between TT and TT\u003csub\u003eevap\u003c/sub\u003e based on pH and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e. The \u003cem\u003eT\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e value of the lumen peak did have a higher increase for TT\u003csub\u003eevap\u003c/sub\u003e than TT and the 1056/1107 (hemicellulose/holocellulose) peak ratio was lower for TT\u003csub\u003eevap\u003c/sub\u003e than TT. However, the peak ratio of 1739/1502 (hemicellulose/lignin) gave the opposite result, being significantly lower for TT than for TT\u003csub\u003eevap\u003c/sub\u003e.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eFor a follow-up study, it would be interesting to assess how both processes affect water vapour sorption at increasing relative humidity as well as the susceptibility to fungal decay of both wood species.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eRetaining volatile degradation products during the thermal treatment process resulted in a more hydrophobic end-product in case of beech. For Scots pine, the type of thermal modification process did not have an effect on the pH and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{c}_{\\text{e}\\text{x}\\text{t}\\text{r}}\\)\u003c/span\u003e\u003c/span\u003e and as the impact on wood-water interactions was not in line with the findings on wood chemistry, the results are inconclusive. The methodological approach, including light microscopy, extraction and LFNMR assessment of volatile compounds, offered important insights into the LFNMR results, ensuring that conclusions related to increases in LFNMR signal were indeed linked to increased hydrophobicity and not related to other possible side-effects of the thermal modification process that might affect the LFMNR signal, such as in-or decreases in lumen size and the presence of volatile components.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eCompeting Interests\u003c/h2\u003e \u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e \u003ch2\u003eEthics Approval\u003c/h2\u003e \u003cp\u003eEthics approval was not required for this study.\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThe authors gratefully acknowledge financial support from Interreg \u0026Ouml;resund-Kattegat-Skagerrak grant number 20201851, Aase \u0026amp; Ejnar Danielsens Fond, STARK fonden and \u0026Aring;forsk foundation.\u003c/p\u003e\u003ch2\u003eAuthor Contributions\u003c/h2\u003e \u003cp\u003eAll authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by L.D.L., M.F. and E.E.T. The first draft of the manuscript was written by L.D.L. and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAltgen M, Militz H (2016) Influence of process conditions on hygroscopicity and mechanical properties of European beech thermally modified in a high-pressure reactor system. Holzforschung 70:971\u0026ndash;979. https://doi.org/10.1515/hf-2015-0235\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAltgen M, Willems W, Militz H (2016) Wood degradation affected by process conditions during thermal modification of European beech in a high-pressure reactor system. Eur J Wood Prod 74:653\u0026ndash;662. https://doi.org/10.1007/s00107-016-1045-y\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenjamini Y, Hochberg Y (1995) Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B: Statistical Methodology 57:289\u0026ndash;300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoonstra MJ, van Acker J, Kegel E, Stevens M (2007) Optimisation of a two-stage heat treatment process: durability aspects. Wood Sci Technol 41:31\u0026ndash;57. https://doi.org/10.1007/s00226-006-0087-4\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai C, Antikainen J, Luostarinen K, et al (2018) Wetting-induced changes on the surface of thermally modified Scots pine and Norway spruce wood. Wood Sci Technol 52:1181\u0026ndash;1193. https://doi.org/10.1007/s00226-018-1030-1\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai C, Javed MA, Komulainen S, et al (2020) Effect of natural weathering on water absorption and pore size distribution in thermally modified wood determined by nuclear magnetic resonance. Cellulose 27:4235\u0026ndash;4247. https://doi.org/10.1007/s10570-020-03093-x\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCandelier K, Chaouch M, Dumar\u0026ccedil;ay S, et al (2011) Utilization of thermodesorption coupled to GC\u0026ndash;MS to study stability of different wood species to thermodegradation. Journal of Analytical and Applied Pyrolysis 92:376\u0026ndash;383. https://doi.org/10.1016/j.jaap.2011.07.010\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCEN (1997) EN 84 Wood preservatives - Accelerated ageing of treated wood prior to biological testing - Leaching procedure\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCermak P, Rautkari L, Horacek P, et al (2015) Analysis of Dimensional Stability of Thermally Modified Wood Affected by Re-Wetting Cycles. BioResources 10:. https://doi.org/10.15376/biores.10.2.3242-3253\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Micco V, Balzano A, Wheeler EA, Baas P (2016) TYLOSES AND GUMS: A REVIEW OF STRUCTURE, FUNCTION AND OCCURRENCE OF VESSEL OCCLUSIONS. IAWA J 37:186\u0026ndash;205. https://doi.org/10.1163/22941932-20160130\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunn OJ (1961) Multiple Comparisons among Means. Journal of the American Statistical Association 56:52\u0026ndash;64. https://doi.org/10.1080/01621459.1961.10482090\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDzurenda L, Dudiak M (2021) Cross-correlation of color and acidity of wet beech wood in the process of thermal treatment with saturated steam. WR 66:105\u0026ndash;116. https://doi.org/10.37763/wr.1336-4561/66.1.105116\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEsteves B, Velez Marques A, Domingos I, Pereira H (2013) Chemical changes of heat treated pine and eucalypt wood monitored by FTIR. Maderas Ciencia y tecnolog\u0026iacute;a 15:245\u0026ndash;258. https://doi.org/10.4067/S0718-221X2013005000020\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEsteves B, Videira R, Pereira H (2011) Chemistry and ecotoxicity of heat-treated pine wood extractives. Wood Sci Technol 45:661\u0026ndash;676. https://doi.org/10.1007/s00226-010-0356-0\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFredriksson M, Digaitis R, Engqvist J, Thybring EE (2023) Effect of targeted acetylation on wood\u0026ndash;water interactions at high moisture states. Cellulose. https://doi.org/10.1007/s10570-023-05678-8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFredriksson M, Thybring EE (2019) On sorption hysteresis in wood: Separating hysteresis in cell wall water and capillary water in the full moisture range. PLOS ONE 14:e0225111. https://doi.org/10.1371/journal.pone.0225111\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFredriksson M, Thygesen LG (2017) The states of water in Norway spruce (Picea abies (L.) Karst.) studied by low-field nuclear magnetic resonance (LFNMR) relaxometry: assignment of free-water populations based on quantitative wood anatomy. Holzforschung 71:77\u0026ndash;90. https://doi.org/10.1515/hf-2016-0044\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo X, Wu Y, Yan N (2018a) In situ micro-FTIR observation of molecular association of adsorbed water with heat-treated wood. Wood Sci Technol 52:971\u0026ndash;985. https://doi.org/10.1007/s00226-018-1020-3\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo Y, Zhang M, Xie Y, et al (2018b) Effect of Thermal Treatment on the Heat of Vaporization of Bound Water by NMR and DSC Analysis. BioResources 13:5534\u0026ndash;5542\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo Y, Zhang M, Xie Y, et al (2018c) Effect of Thermal Treatment on the Heat of Vaporization of Bound Water by NMR and DSC Analysis. BioResources 13:5534\u0026ndash;5542\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHill C, Altgen M, Rautkari L (2021) Thermal modification of wood\u0026mdash;a review: chemical changes and hygroscopicity. J Mater Sci 56:6581\u0026ndash;6614. https://doi.org/10.1007/s10853-020-05722-z\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHill CAS, Ramsay J, Keating B, et al (2012) The water vapour sorption properties of thermally modified and densified wood. J Mater Sci 47:3191\u0026ndash;3197. https://doi.org/10.1007/s10853-011-6154-8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHofmann T, R\u0026eacute;tfalvi T, Albert L, Niemz P (2008) Investigation of the chemical changes in the structure of wood thermally modified within a nitrogen atmosphere autoclave. WOOD RESEARCH 53:1\u0026ndash;14\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolmbom B (1999) Extractives. In: Sj\u0026ouml;str\u0026ouml;m E, Al\u0026eacute;n R (eds) Analytical Methods in Wood Chemistry, Pulping, and Papermaking. Springer, Berlin, Heidelberg, pp 125\u0026ndash;148\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHsieh CC, Cannella D, J\u0026oslash;rgensen H, et al (2014) Cellulase inhibition by high concentrations of monosaccharides. J Agric Food Chem 62:3800\u0026ndash;3805. https://doi.org/10.1021/jf5012962\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang W, Pedersen NB, Fredriksson M, Thygesen LG (2019) Annual variations in Norway spruce xylem studied using infrared micro-spectroscopy. Forests 10:164. https://doi.org/10.3390/f10020164\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIstratov AA, Vyvenko OF (1999) Exponential analysis in physical phenomena. Review of Scientific Instruments 70:1233\u0026ndash;1257. https://doi.org/10.1063/1.1149581\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eK\u0026auml;llbom S, Sedighi Moghaddam M, W\u0026aring;linder MEP (2018) Liquid sorption, swelling and surface energy properties of unmodified and thermally modified Scots pine heartwood after extraction. Holzforschung 72:251\u0026ndash;258. https://doi.org/10.1515/hf-2017-0064\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKekkonen PM, Telkki V-V, Jokisaari J (2010) Effect of Thermal Modification on Wood Cell Structures Observed by Pulsed-Field-Gradient Stimulated-Echo NMR. J Phys Chem C 114:18693\u0026ndash;18697. https://doi.org/10.1021/jp1060304\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKocaefe D, Poncsak S, Boluk Y (2008) Effect of thermal treatment on the chemical composition and mechanical properties of birch and aspen. BioResources 3:. https://doi.org/10.15376/biores.3.2.517-537\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLupoi JS, Gjersing E, Davis MF (2015) Evaluating Lignocellulosic Biomass, Its Derivatives, and Downstream Products with Raman Spectroscopy. Front Bioeng Biotechnol 3:. https://doi.org/10.3389/fbioe.2015.00050\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMecca M, Todaro L, Giudice VL, et al (2021) GC-MS and SPME Techniques Highlighted Contrasting Chemical Behaviour in the Water Extractives of Modified Castanea sativa Mill. and Fagus sylvatica L. Wood. Forests. https://doi.org/10.3390/f12080986\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNhien LC, Long NVD, Kim S, Lee M (2017) Techno-economic assessment of hybrid extraction and distillation processes for furfural production from lignocellulosic biomass. Biotechnology for Biofuels 10:81. https://doi.org/10.1186/s13068-017-0767-3\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u0026Ouml;zgen\u0026ccedil; \u0026Ouml;, Durmaz S, Boyaci IH, Eksi-Kocak H (2017) Determination of chemical changes in heat-treated wood using ATR-FTIR and FT Raman spectrometry. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 171:395\u0026ndash;400. https://doi.org/10.1016/j.saa.2016.08.026\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePedersen NB (2015) Microscopic and Spectroscopic Characterisation of Waterlogged Archaeological Softwood from Anoxic Environments\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSluiter A, Ruiz R, Scarlata C, et al (2005) Determination of extractives in biomass. Laboratory analytical procedure (LAP) 1617:1\u0026ndash;16\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSundqvist B, Karlsson O, Westermark U (2006) Determination of formic-acid and acetic acid concentrations formed during hydrothermal treatment of birch wood and its relation to colour, strength and hardness. Wood Sci Technol 40:549\u0026ndash;561. https://doi.org/10.1007/s00226-006-0071-z\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThybring EE, Digaitis R, Nord-Larsen T, et al (2020) How much water can wood cell walls hold? A triangulation approach to determine the maximum cell wall moisture content. PLOS ONE 15:e0238319. https://doi.org/10.1371/journal.pone.0238319\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThybring EE, Fredriksson M (2021) Wood Modification as a Tool to Understand Moisture in Wood. Forests 12:372. https://doi.org/10.3390/f12030372\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTjeerdsma B, Militz H (2005a) Chemical changes in hydrothermal treated wood: FTIR analysis of combined hydrothermal and dry heat-treated wood. Holz als Roh- und Werkstoff 63:102\u0026ndash;111. https://doi.org/10.1007/s00107-004-0532-8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTjeerdsma B, Militz H (2005b) Chemical changes in hydrothermal treated wood: FTIR analysis of combined hydrothermal and dry heat-treated wood. Holz als Roh- und Werkstoff 63:102\u0026ndash;111. https://doi.org/10.1007/s00107-004-0532-8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVek V, Oven P, Poljanšek I (2016) Review on Lipophilic and Hydrophilic Extractives in Tissues of Common Beech. Drvna Industrija 67:85\u0026ndash;96. https://doi.org/10.5552/DRIND.2016.1511\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerkasalo E, Roitto M, M\u0026ouml;tt\u0026ouml;nen V, et al (2022) Extractives of Tree Biomass of Scots Pine (Pinus sylvestris L.) for Biorefining in Four Climatic Regions in Finland\u0026mdash;Lipophilic Compounds, Stilbenes, and Lignans. Forests 13:779. https://doi.org/10.3390/f13050779\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWindeisen E, Strobel C, Wegener G (2007) Chemical changes during the production of thermo-treated beech wood. Wood Sci Technol 41:523\u0026ndash;536. https://doi.org/10.1007/s00226-007-0146-5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZelinka SL, Altgen M, Emmerich L, et al (2022) Review of Wood Modification and Wood Functionalization Technologies. Forests 13:1004. https://doi.org/10.3390/f13071004\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStatements \u0026amp; Declarations\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"27fb55ca-fc3d-402a-9a36-b23fb3f9bbe9","identifier":"10.13039/100013276","name":"Interreg","awardNumber":"20201851","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University of Copenhagen","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"thermal modification, volatile degradation products, wood-water interactions, LFNMR, ATR-FTIR, beech, Scots pine","lastPublishedDoi":"10.21203/rs.3.rs-5191132/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5191132/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThermal modification is a widely used technology for enhancing the dimensional stability and durability of wood. However, thermal degradation reactions in wood are complex and depend on the process conditions applied. Volatile degradation products from thermal wood modification are expected to affect the wood chemistry and the wood-water interactions of the thermally modified wood. In this paper, we investigate the impact on wood chemistry and wood-water interactions of retaining or evaporating volatile degradation products while thermally modifying beech and Scots pine in a closed thermal treatment process. The presence of extractable degradation products was determined based on water and multi-solvent extraction. pH measurements and ATR-FTIR analyses were performed to determine differences in wood chemistry. Light microscopy images of xylem cross-sections were taken to determine the vessel and tracheid lumen areas, and LFNMR and DSC measurements were performed to determine wood-water interactions. We found that thermal treatment with a cooling step at atmospheric pressure indeed allowed volatile degradation products to evaporate, in the case of beech, resulting in a less hydrophobic end product. However, for Scots pine, evaporating volatile degradation products during the thermal modification process did not have an effect on the pH and the amount of extractives, and as the impact on wood-water interactions was not in line with the findings on wood chemistry, the results are inconclusive. Our results demonstrate that volatile degradation products can have an impact on the wood-water interactions of thermally modified wood in the cooling step and that the results are wood species dependent.\u003c/p\u003e","manuscriptTitle":"Influence of volatile degradation products from thermal wood modification on wood-water interactions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-07 11:35:39","doi":"10.21203/rs.3.rs-5191132/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"256d509f-7e22-42d7-8c6f-600adae77d7c","owner":[],"postedDate":"October 7th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-10-07T11:35:39+00:00","versionOfRecord":[],"versionCreatedAt":"2024-10-07 11:35:39","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5191132","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5191132","identity":"rs-5191132","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}