Assessing the fire and thermal behavior of thermally modified, acetylated, and one-sided surface charred beech and birch wood

Abstract

The fire and thermal behavior of European beech (Fagus sylvatica L.) and silver birch (Betula pendula L.) woods were comparatively evaluated after being modified by thermal treatment (200 °C), acetylation, and one-sided surface charring. A small-flame ignition test (EN ISO 11925-2), gross heat of combustion via bomb calorimetry (EN ISO 1716), and thermo-gravimetric analysis (TG/DTG) were used for assessment. Thermal modification resulted in moderate mass loss (~3% to 4%), while acetylation achieved high weight percentage gain (>20%) and the strongest reduction in equilibrium moisture content; surface charring had only a negligible effect on bulk moisture properties. While none of the applied modifications altered the Euroclass-related reaction-to-fire classification, statistically significant differences in early-stage flame spread were observed. Acetylated wood exhibited increased flame spread, whereas surface-charred specimens showed a pronounced reduction, particularly for beech. All modified materials displayed higher gross heat of combustion compared with reference wood, reflecting increased carbon-rich constituents; however, this increase did not directly correspond to improved fire performance. TGA revealed similar degradation behavior for reference and thermally modified wood, whereas acetylated and surface-charred materials exhibited fundamentally different thermal responses because of chemical substitution and polysaccharide degradation with carbon-rich char formation, respectively. The study highlighted surface charring as an effective modification for reducing flame spread and demonstrates the necessity of combining complementary fire-testing methods to capture modification-specific fire behavior.

Full Article

Assessing the Fire and Thermal Behavior of Thermally Modified, Acetylated, and One-sided Surface Charred Beech and Birch Wood

Petr Čermák  , ^a,* Ondřej Prokop  ,^a Richard Slávik  ,^a Petr Sláčik,^b

Miroslav Zapletal,^b Jozef Ráhel  ,^a and Jakub Dömény  ,^a

The fire and thermal behavior of European beech (Fagus sylvatica L.) and silver birch (Betula pendula L.) woods were comparatively evaluated after being modified by thermal treatment (200 °C), acetylation, and one-sided surface charring. A small-flame ignition test (EN ISO 11925-2), gross heat of combustion via bomb calorimetry (EN ISO 1716), and thermo-gravimetric analysis (TG/DTG) were used for assessment. Thermal modification resulted in moderate mass loss (~3% to 4%), while acetylation achieved high weight percentage gain (>20%) and the strongest reduction in equilibrium moisture content; surface charring had only a negligible effect on bulk moisture properties. While none of the applied modifications altered the Euroclass-related reaction-to-fire classification, statistically significant differences in early-stage flame spread were observed. Acetylated wood exhibited increased flame spread, whereas surface-charred specimens showed a pronounced reduction, particularly for beech. All modified materials displayed higher gross heat of combustion compared with reference wood, reflecting increased carbon-rich constituents; however, this increase did not directly correspond to improved fire performance. TGA revealed similar degradation behavior for reference and thermally modified wood, whereas acetylated and surface-charred materials exhibited fundamentally different thermal responses because of chemical substitution and polysaccharide degradation with carbon-rich char formation, respectively. The study highlighted surface charring as an effective modification for reducing flame spread and demonstrates the necessity of combining complementary fire-testing methods to capture modification-specific fire behavior.

DOI: 10.15376/biores.21.3.5749-5767

Keywords: Wood modification; Thermal modification; Acetylation; Surface charring; Fire resistance; Combustion heat; Wood properties

Contact information: a: Department of Wood Science and Technology, Faculty of Forestry and Wood Technology, Mendel University in Brno, Zemědělská 3, 613 00 Brno, Czech Republic; b: Joinery Products Testing Institute, Faculty of Forestry and Wood Technology, Mendel University in Brno, Zemědělská 3, 613 00 Brno, Czech Republic; *Corresponding author: xcerma24@mendelu.cz

INTRODUCTION

Wood is widely valued in construction for its renewability, low embodied energy, and carbon storage potential, which aligns well with sustainable development goals (Schlamadinger and Marland 1996; Gustavsson et al. 2006). Its high strength-to-weight ratio and aesthetic appeal have also contributed to its resurgence in modern architecture. However, despite these unquestionably positive benefits, its vulnerability to moisture, fungal decay, and fire limits its broader application in safety-critical or long-life structures (White and Dietenberger 2010; Rowell 2013). Traditional preservative treatments can improve performance, but often they involve toxic compounds (Hill 2006). To address these limitations in an environmentally responsible way, wood modification technologies have been developed to improve dimensional stability, durability and, potentially, fire resistance (Esteves and Pereira 2009; Mantanis 2017; Sandberg et al. 2017; Jones and Sandberg 2020; Hill et al. 2021). According to Sandberg et al. (2017), wood modification includes chemical, thermo-hydro-mechanical, physical, and biological processes that alter the structure or composition of wood cell walls. Common techniques include acetylation, furfurylation, silane treatment, polymer impregnation, thermal modification, surface charring, etc. (Hill 2006; Sandberg et al. 2017). These modifications are typically non-toxic and result in wood products with improved service life and reduced maintenance needs. Among these, thermal modification, acetylation, and one-sided surface charring have attracted particular attention not only for enhancing dimensional stability, surface wettability, and decay resistance, but also for their potential impact on fire behavior (Rabe et al. 2020; Zelinka et al. 2022; Marino et al. 2024).

Thermal modification involves heating wood at elevated temperatures (typically 160 to 240 °C) in low-oxygen environments, leading to degradation of hemicelluloses and partial reorganization of lignin. These changes reduce the wood’s hygroscopicity and improve its biological durability (Esteves and Pereira 2009; Čermák et al. 2022). However, their effect on fire behavior is more nuanced, with several studies reporting reduced burning intensity despite changes in ignition behaviour. For instance, Martinka et al. (2016) showed that Norway spruce (Picea abies L.) and English oak (Quercus robur) thermally modified at 200 °C exhibited a marked reduction in peak heat release rate (pHRR), decreasing from 268 and 243 to 218 kW·m^-2 at an external heat flux of 50 kW·m^-2, respectively. In open-flame fire resistance tests, Čekovská et al. (2017a) reported lower weight loss for thermally modified spruce, decreasing from 4.72% in untreated wood to the range 3.98% to 4.48% after treatment at 160 to 210 °C. Cone calorimeter measurements by Xing and Li (2014) demonstrated that thermally modified larch (Larix spp., 180 to 210 °C, nitrogen atmosphere) showed reduced burning intensity but a shorter ignition time (22 s to 13 s) and increased total smoke production (+4.8% to +43.3%). In contrast, for teak (Tectona grandis Linn), Čekovská et al. (2017b) observed an increase in weight loss after 10-min direct flame exposure, rising from 4.8% in untreated wood to the range 11.2% to 13.5% in thermally modified wood, highlighting that fire-related benefits of thermal modification are not universal.

Acetylation chemically modifies wood polymers by replacing the hydroxyl groups (–OH) with acetyl groups (–OCOCH₃), typically resulting in 18% to –25% weight percentage gain and substantially improved dimensional stability and durability (Popescu et al. 2014; Čermák et al. 2022). The effect of acetylation on flammability has been investigated in several studies with inconsistent outcomes. Cone calorimeter measurements on acetylated Monterey pine (Pinus radiata, WPG 20%) showed a shorter ignition time compared with unmodified wood (22 vs. 27 s) and a substantially higher first peak heat release rate (274 vs. 187 kW·m⁻²), indicating enhanced initial flammability (Rabe et al. 2020). In radiant panel tests on acetylated ash (Fraxinus excelsior L.), Morozovs and Bukšāns (2009) reported increased flame phase duration (1185 vs. 1675 s) and flame spread (437 vs. 600 mm) but decreased critical heat flux at extinguishment (4.8 vs. 2.6 kW·m^-2). Small-flame fire tests conducted by Papadopoulos et al. (2010) showed that acetylated OSB (WPG 11.2%) and particleboard (WPG 12.2%) exhibited longer ignition and flaming times (up to 245 vs. 93 s), but substantially reduced glowing durations, reflecting altered smoldering behaviour. These observations are consistent with findings by Hasburgh and Zelinka (2023), who attributed increased early heat release in acetylated wood to the emission of acetic acid and other volatile degradation products, which may act as additional combustible fuel under well-ventilated burning conditions.

Furthermore, surface charring is an age-old technique, traditionally known from the Japanese Yakisugi (Shou Sugi Ban) method, in which wood surfaces are intentionally burned to create a protective char layer (Kymäläinen et al. 2017; Ebner et al. 2024). The char layer acts as a carbonized barrier that is hydrophobic, fungal-resistant, and dimensionally stable, as most polysaccharides are decomposed during pyrolysis. Traditional flame charring exposes the wood surface to temperatures exceeding 800 °C, producing a thick (2 to 5 mm) but often cracked and brittle char layer (Ebner et al. 2022). In contrast, recent studies have explored low-temperature contact charring using hot plates at temperatures ranging between 250 and 400 °C (Kymäläinen et al. 2017; Šeda et al. 2021). Lin et al. (2023) demonstrated that pre-charred wood surfaces substantially improve fire performance when a continuous char layer is present. Increasing char thickness to 13.8 mm reduced the first peak heat release rate from 155 to 59 kW·m^-2, strongly suppressing early fire growth. Pre-charring also delayed ignition and increased ignition temperature to >650 °C. A minimum effective char thickness of ±6 mm was identified, below which fire performance improvements were negligible. Ebner et al. (2024) showed that one-sided surface charring improved fire resistance of silver fir (Abies alba Mill.) and European ash (Fraxinus excelsior), increasing burn-through time from 17.0 to 23.0 min and 22.0 to 25.0 min, with 6 and 3 mm char layer thickness, respectively. Contact charring using heated plate further confirmed the protective role of engineered char layers, as reported by Machová et al. (2021), who showed that one-sided charred beech wood (Fagus sylvatica L.) reduced weight loss during fire exposure from 16.6 % (untreated) to 6.8 % after surface charring at 250 °C for 6 min. It is evident that fire performance improvements are highly dependent on char uniformity and thickness and are not universally achieved across wood species (Hasburgh et al. 2021).

In addition to standardized fire tests, the fire response of wood is closely related to its thermal degradation behavior, which controls the onset of pyrolysis, volatile release, and char formation during heating. Thermogravimetric and calorimetric techniques are therefore widely used to describe the thermal stability and decomposition pathways of lignocellulosic materials, providing fundamental background information that complements flame-based fire testing (Grønli et al. 2002; Yang et al. 2007).

Although these modifications have been studied individually, comparative evaluations, especially on hardwoods like beech (Fagus sylvatica L.) and birch (Betula pendula L.), remain limited. Standardized testing using small-flame ignition (EN ISO 11925-2, 2020), gross heat of combustion (EN ISO 1716, 2018), and thermogravimetric analysis (TGA) can provide a clearer comprehensive picture of how these treatments influence flammability and combustion behavior. It is hypothesized that contact surface-charring will yield improved fire and thermal behavior, followed by thermal modification, while acetylation will provide either minimal or inconsistent benefits when compared to reference material.

EXPERIMENTAL

Material and Methods

European beech (Fagus sylvatica L.) and silver birch (Betula pendula L.) wood harvested in Training Forest Enterprise, Masaryk Forest Křtiny of Mendel University in Brno, Czech Republic, was studied. The tested materials were prepared from defect-free raw timber in accordance with the requirements of the applied test methods, with additional dimensional allowances to enable subsequent machining to the final specimen dimensions prior to testing.

Wood Modification

Thermal modification

Thermal modification at 200 °C (TM200) was carried out in a laboratory chamber (Katres spol. s r.o., CZ, volume 0.7 m³) in a superheated steam environment at atmospheric pressure. The treatment was performed in five stages: an initial heating ramp 10 °C/h to drying temperature (103 °C), 70 h at this temperature; a second ramp 4 °C/h to reach 130 °C; third ramp 30 °C/h until the maximum temperature (200 °C) was reached, maintained for 3 h; and natural cooling. The process intensity and degree of thermal modification was determined by mass loss (ML%), based on oven-dry (103 ±2 °C) mass before and after the thermal modification process. The ML was calculated according to Eq. 1:

ML (%) = (Dry mass_unmod – Dry mass_mod / Dry mass_unmod (1)

Acetic anhydride modification – Acetylation

The oven-dried specimens were vacuum-impregnated at 10 kPa (absolute pressure) for 0.5 h with acetic anhydride (Acet) using a laboratory vacuum-pressure impregnation plant (J. Hradecký spol. s r.o., CZ, vessel volume 50 L). After vacuum impregnation, the specimens were kept in the chemical at ambient pressure for 0.5 h, then placed in stainless-steel containers and sealed during the acetylation. The acetylation reaction took place at 120 °C in a hot-air oven (Sanyo MOV-212P) for 24 h. On completion, the specimens were dried to 0% MC (12 h at 30 °C, 12 h at 60° C and 48 h at 120 ±2 °C). The WPG values were calculated according to Eq. 2.

WPG (%) = (Dry mass_mod – Dry mass_unmod / Dry mass_unmod (2)

One-sided surface charring

Specimens were one-sided surface charred by contact heating using a laboratory hot plate (CERAN 33SR) with a built-in Ni-Cr-Ni sensor and electronic temperature controller. The specimens were first dried to 0% MC (103 ±2 °C), then immediately charred to avoid any shape deformation during contact heating. Surface charring was carried out at 300 °C for 60 s at atmospheric pressure. No additional pressure was applied on specimens during the charring process. The char layer thickness was approximately 3 mm, following the methodology of Šeda et al. (2021)

For all studied wood modifications, the moisture exclusion efficiency (MEE) at 50% RH was determined according to Eq. 3.

MEE (%) = (EMC_unmod – EMC_mod) / EMC_unmod (3)

Fire-Resistance Properties

Single-flame source test

Fire performance related to ignitability and surface flame spread was evaluated using the single-flame source test in accordance with EN ISO 11925-2 (2020). Rectangular specimens with nominal dimensions of 250 × 90 × 20 mm (length × width × thickness) were prepared from reference and modified wood materials. Prior to testing, all specimens were conditioned at 23 ± 2 °C and 50 ± 5% RH until constant mass was achieved. During the test, the specimens were mounted vertically in the test frame and exposed to a propane flame applied to the specimen surface at an angle of 45°. The distance between the burner nozzle and the specimen surface was maintained at approximately 5 mm. The flame application time was 30 s, in accordance with the applied test procedure. For each material variant, ten repetitions were performed, exceeding the minimum requirement of six specimens specified by the standard. Before the fire tests, airflow conditions in the test environment were verified using a Trotec TA300 anemometer, and the air velocity was measured at approximately 0.7 m·s^-1 to ensure stable and reproducible testing conditions. During the test, ignition behavior, flame development on the specimen surface, and persistence of flaming after flame removal were visually observed and recorded in accordance with the standard. In addition to the standard observations, the height of the charred surface resulting from flame exposure and the weight loss of the specimens after the test were measured as supplementary parameters to further characterize fire-induced material degradation. Each material group was measured in ten replicates.

Gross heat of combustion

The gross heat of combustion (PCS) was determined in accordance with EN ISO 1716 (2018) using an oxygen bomb calorimeter to quantify the total potential heat release of the tested materials under complete combustion conditions. For this purpose, the wood specimens were mechanically milled to obtain a fine and homogeneous wood powder, from which small pellets were subsequently formed to ensure uniform combustion behavior. For the surface-charred specimens, the PCS was determined separately for two successive material layers: (i) an isolated outer charred surface layer obtained by milling the uppermost 0.5 mm of the charred surface, and (ii) a subsurface layer obtained by milling the subsequent 0.5 mm (i.e., material taken from a depth of 0.5 to 1.0 mm). Approximately 1 g of wood powder was used for each test to prepare the pellets. All samples were oven-dried at 103 ± 2 °C to constant mass prior to testing. Each material group was analyzed in five independent repetitions. A known mass of the prepared pellet was combusted in a high-pressure oxygen atmosphere inside the calorimetric bomb, and the resulting temperature increase of the calorimetric system was continuously recorded. The gross heat of combustion was calculated according to Eq. 4,

PCS (MJ/kg) = (E*(T_m–T_i+c)–b)/m              (4)

where E is water equivalent (MJ/K), T_m is the maximum measured temperature after combustion (K), T_i is initial the temperature of the calorimetric system (K), c is temperature correction accounting for heat exchange with the surrounding (K), b is the correction for the heat contribution of auxiliary combustion material used during the test, such as cotton thread (MJ), and m is the dry mass of the test specimen (kg). Each material group was measured in five replicates.

Thermogravimetric analysis

Thermogravimetric measurements were carried out using a simultaneous thermal analysis instrument (Hitachi NEXTA STA200RV), combining thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) in a single unit. The instrument is equipped with RealView technology enabling optical monitoring of the sample during measurement. The balance weight resolution is 0.2 μg, with stability/drift/repeatability error < 10 μg. Wood dust (2 to 5 mg) was weighed into aluminium opened crucibles and measured against a blank reference. Experiments were performed under an inert nitrogen atmosphere (100 mL·min⁻¹). The temperature program started at 30 °C (2 min stabilization) and increased to 500 °C at a constant heating rate of 20 K·min⁻¹. Data were recorded every 0.2 s, and sample images were acquired every 15 s. Each material group was measured in five replicates.

Data Processing and Statistical Analysis

Data were processed, evaluated, and graphed using OriginPro 2022 (OriginLab Corporation). Statistical differences among the studied groups were assessed using the nonparametric Kruskal–Wallis test. When significant differences were detected, pairwise comparisons were performed using Dunn’s multiple comparison test with Holm adjustment. Statistical significance (α) was set to 0.05.

RESULTS AND DISCUSSION

Material Characteristics

Table 1 presents key indicators of modification intensity and moisture-related characteristics, such as mass loss (ML), weight percentage gain (WPG), equilibrium moisture content (EMC), and moisture exclusion efficiency (MEE), which together characterize the degree of chemical and structural alterations induced by the applied wood modifications and provide essential context for interpreting their performance.

Table 1. Wood Modification Groups, Process, and Moisture-related Characteristics

Thermal modification at 200 °C resulted in a relatively low mass loss (ML) of 3.8% for beech and 3.2% for birch, indicating mild to moderate treatment severity, in which degradation is dominated by hemicellulose depolymerization and deacetylation, while cellulose degradation remains limited (Esteves and Pereira 2009). Such ML values are characteristic of thermally modified hardwoods processed under industrially relevant schedules aimed at improving hygroscopic behaviour while avoiding excessive degradation of structural biopolymers. Comparable ML values for beech treated at 200 °C were reported by Čermák et al. (2021, 2022), who observed ML in the range of approximately 2.7% to 5.7%, depending on treatment duration. Olarescu et al. (2014) reported ML values ranging between 3.2% and 5.1% for beech wood thermally modified for at 200 °C for 2 h and 3 h, respectively. Similarly, Borůvka et al. (2021) reported ML values of 5% for beech and 8% for birch wood modified at 210 °C. Mass loss of 6.4% for birch (Betula pendula) wood modified at 200 °C for 4 h was reported by Zeman et al. (2000) and Esteves et al. (2008). These findings confirm that a wide range of ML values can be observed across different processing conditions, while both diffuse-porous hardwood species exhibit comparable degradation kinetics. The ML values obtained in this study indicate a relatively mild treatment severity when compared with other studies employing similar process conditions.

The moderate ML observed in both species is directly reflected in their moisture behavior. Thermal modification significantly reduced EMC, corresponding to MEE values of 48.9% and 44.9%, respectively. These reductions are consistent with literature data reporting EMC decreases of approximately 40% to 55% for wood thermally modified at 200 °C (Altgen et al. 2016; Čermák et al. 2021; Engelund and Fredriksson 2021). Similar EMC reductions for thermally modified wood were also reported by Rautkari and Hill (2014), who attributed this behavior primarily to a reduced availability of accessible hydroxyl groups and partial collapse of the amorphous hemicellulose network. The close correspondence between ML and EMC reduction in the present study indicates that moderate thermal degradation is sufficient to substantially reduce hygroscopicity.

Acetylation resulted in a substantially higher degree of modification, with WPG values of 21.7% for beech and 24.6% for birch. These values are characteristic of strongly acetylated hardwoods and are consistent with results reported in laboratory- and pilot-scale acetylation studies. Čermák et al. (2022) reported a WPG of 24.9% for acetylated beech wood, accompanied by pronounced reductions in EMC. Similarly, Popescu et al. (2014) demonstrated that acetylated birch wood with WPG levels exceeding 16% exhibits a marked downward shift of the sorption isotherm, which can be attributed to chemical substitution of hydroxyl groups and permanent bulking of the cell wall. In the present study, acetylation led to a significant reduction in EMC, corresponding to MEE values of 63.0% and 65.2% for beech and birch, respectively. These reductions are in agreement with classical acetylation literature, where EMC decreases of approximately 60% to 70% are commonly reported for WPG levels in the range of 20% to 25% (Papadopoulos and Hill 2003; Rowell 2006; Čermák et al. 2022). Compared with thermal modification, acetylation therefore provides a more pronounced and stable reduction in wood hygroscopicity.

In contrast, one-sided surface charring induced only negligible changes in bulk moisture behaviour. The EMC values of charred wood remained close to those of the corresponding reference materials, resulting in low MEE values (2.2% to 4.5%). This response is characteristic of one-sided or superficial charring, where modification is confined to a thin surface layer and does not substantially affect the hygroscopic properties of the wood core. Similar findings were reported for one-sided charred beech by Čermák et al. (2019) and Šeda et al. (2021), who showed that although surface wettability and density profiles are markedly altered, the EMC of the bulk material remains largely unchanged. Comparable conclusions were also drawn by Kymäläinen et al. (2022), demonstrating that surface carbonization primarily affects surface moisture interactions rather than the equilibrium sorption of the entire specimen.

Fire Resistance Properties

The reaction-to-fire behavior of reference beech and birch wood, as well as modified wood was assessed using the EN ISO 11925-2 (2020) small flame test. All tested materials exhibited flame spread behavior comparable to their respective reference specimens, with no specimen reaching the critical flame spread height of 150 mm within the specified observation time and no visually significant occurrence of flaming droplets or sustained post-flame combustion. According to the classification framework of EN 13501-1, solid wood products of this type fall within Euroclass D when assessed by the small flame test alone, and from that perspective the applied modification treatments did not result in any observable change in classification (Östman et al. 2010).

Although the EN ISO 11925-2 (2020) assessment did not indicate any change in reaction-to-fire classification according to EN 13501-1 (2018), Fig. 1 clearly demonstrates statistically significant differences in flame spread height among the studied modifications. These results suggest that while the applied treatments do not affect compliance-based criteria, they do influence the early-stage flame propagation behaviour under small-flame exposure. For both beech and birch, reference wood and thermally modified wood at 200 °C (TM200) indicating comparable flame spread behaviour. This seems to be consistent with previous findings showing that moderate thermal modification temperatures around 160 to 200 °C result in no statistically significant changes to ignitability and flame spread, as hemicellulose degradation and char yield enhancement remain insufficient to alter flame propagation under small-flame conditions (Martinka et al. 2016; Čekovská et al. 2017a; Luptáková et al. 2019).

Fig. 1. Flame spread height (mm) measured during the small flame (n=10) test for beech (left) and birch (right) wood specimens

In contrast, acetylated wood exhibited significantly higher flame spread heights in both studied species. This observation is consistent with some studies showing that acetylation does not improve, and may locally increase, flammability indicators in small-scale and bench-scale fire tests (Mohebby et al. 2007; Rabe et al. 2020). This behaviour can be explained by a combination of chemical and physical changes induced by acetylation. During thermal decomposition, acetylated wood releases increased amounts of acetyl-related volatile products, particularly acetic acid, originating from cleavage of ester bonds. This has been detected in pyrolysis and combustion gases and is known to participate in flame chemistry under well-ventilated conditions (Schwarzinger and List 2010; Hasburgh and Zelinka 2023). In addition, acetylation substantially reduces EMC, thereby limiting the endothermic heat sink associated with water evaporation during heating. Reduced moisture buffering shortens the preheating phase prior to pyrolysis and ignition, facilitating earlier volatile release and flame propagation, a relationship that is well established for wood materials in ignition and flame-spread studies (Spearpoint and Quintiere 2000; Cuevas and Maluk 2023). Together, these effects explain why acetylated wood may display increased flame spread under small-flame exposure despite its improved material properties.

The one-sided surface-charred specimens exhibited a clear reduction in flame spread height, with the effect being most pronounced for beech. This behaviour can be attributed primarily to the formation of a continuous surface char layer, which acts as a physical and thermal barrier, reducing heat transfer into the underlying wood and thereby limiting the generation and release of flammable pyrolysis gases during small-flame exposure (Ebner et al. 2024). Although surface charring reduces the equilibrium moisture content, the protective effect of the char layer dominates, suppressing ignition and flame spread. In this sense, surface charring alters the initial fire–surface interaction, weakening flame attachment and slowing flame propagation along the specimen surface. The stronger response observed for beech compared with birch is likely related to species-specific differences in density, anatomical structure, and char formation behaviour. Beech has a higher average density (680 kg/m³) and a more uniform diffuse-porous structure, which may promote the development of a denser and more cohesive char layer with better barrier properties. Birch, although also diffuse-porous, typically exhibits lower density (610 kg/m³) and higher porosity, which likely result in a more fragmented or mechanically weaker char layer under identical charring conditions. Such species-dependent differences in char cohesion, density, porosity, and thermal response have been shown to influence flame spread and surface burning behaviour in wood (Lowden and Hull 2013; Bartlett et al. 2019; Mensah et al. 2023).

Fig. 2. Weight loss measured during the small flame test (n=10) for beech (left) and birch (right) wood specimens

Figure 2 shows the weight loss measured after small-flame exposure for beech and birch specimens. Overall, no statistically significant differences were observed among the studied modification types within each species, indicating that studied modifications did not substantially affect weight loss under the applied small-flame conditions. When comparing species, birch exhibited a slightly higher weight loss than beech, although the differences were small. This tendency can be attributed to species-related factors such as the lower density and higher porosity of birch, which may facilitate faster heat penetration and a more pronounced release of volatile decomposition products during short-term flame exposure.

Figure 3 presents the gross heat of combustion (PCS) determined for beech and birch after small-flame exposure. PCS represents the total amount of energy released by complete combustion of a unit mass of dry material, as measured by bomb calorimetry, and reflects the overall energy content of the solid material. For both wood species, all modified groups exhibited higher PCS values than the respective reference materials, indicating an increased energy density. This increase is consistent with modification-induced changes in chemical composition, such as a relative enrichment in lignin and carbon-rich structures due to hemicellulose degradation (thermal modification) or surface charring, as well as the incorporation of acetyl groups in acetylated wood. Lignin and carbonized structures are known to possess higher calorific values than polysaccharides, which explains the systematic increase in PCS observed here. Comparable PCS ranges and similar trends—higher calorific values for modified or carbon-enriched wood materials—have been reported for hardwoods using the same bomb-calorimetric approach (Demirbaş 2001; Telmo and Lousada 2011; Poletto et al. 2012).

Importantly, higher PCS does not directly imply improved fire performance. While materials with higher energy content can release more heat once fully involved in combustion, early-stage fire behaviour (such as flame spread height in EN ISO 11925-2 (2020)) is primarily governed by surface processes, heat transfer, and volatile release kinetics. This helps to explain why acetylated and charred specimens may show higher PCS yet contrasting flame spread behaviour, with acetylation promoting flame spread and surface charring suppressing it. Thus, the PCS results provide complementary information on material energy potential, rather than contradicting the flame-spread observations.

Fig. 3. Gross heat of combustion (n=5) recorded for beech (left) and birch (right) wood specimens

The thermogravimetric (TG) and derivative thermogravimetric (DTG) curves obtained for beech and birch revealed systematic effects of wood species and modification type on thermal degradation behaviour (Fig. 4). Despite these differences, all samples followed the characteristic multi-stage thermal decomposition pattern of lignocellulosic materials. In all specimens, an initial minor mass loss (ML) was observed up to approximately 120 to 150 °C. This stage was associated with the release of physically bound moisture and low-molecular volatile compounds and was reflected by very low DTG intensities. The magnitude of this low-temperature ML differed slightly among the studied groups, with acetylated wood showing a noticeably reduced ML compared with reference samples. This behaviour is consistent with its substantially lower EMC (Table 1), indicating reduced moisture uptake and, consequently, limited moisture desorption during the initial heating stage.

Fig. 4. Thermogravimetric (TG) and derivative thermogravimetric (DTG) curves of reference, thermally modified, acetylated, and surface-charred beech and birch wood (n=5)

The main ML occurred between approximately at 200 ℃ and 380–400 °C and dominated the TG curves of all reference and modified samples. Within this region, a shoulder or secondary DTG feature was observed at lower temperatures, around 250 to 350 °C, followed by a pronounced DTG maximum at approximately 350 to 400 °C. The lower-temperature shoulder is commonly attributed to hemicellulose degradation, whereas the main DTG peak corresponds primarily to cellulose depolymerization (Grønli et al. 2002; Yang et al. 2007). This shoulder seems to be more pronounced in birch than in beech, which is consistent with the higher xylan content of birch and its tendency toward earlier and broader hemicellulose decomposition (Shen et al. 2010). The evolution of the solid residue during the heating process is illustrated in Figs. 5 and 6 for both beech and birch. The photographic documentation confirms the progressive formation of carbonaceous material and the gradual structural collapse of the samples with increasing temperature. These images provide visual support for the TG/DTG results and illustrate the physical changes associated with the thermal degradation process.

Fig. 5. Visual appearances of solid residues of reference, thermally modified, acetylated, and surface-charred beech wood samples at selected temperatures during TGA

Thermally modified samples exhibited TG and DTG profiles that were broadly similar to those of the respective reference wood, with only small shifts in the onset and shape of the main degradation region. These changes indicate partial degradation of thermally labile hemicelluloses during the modification process, while the cellulose-dominated decomposition peak were largely unchanged in position. Such behaviour is consistent with thermal modification at moderate temperatures (~180 to 220 °C), where hemicelluloses undergo deacetylation and depolymerisation, whereas cellulose retains its structural integrity and decomposes within a similar temperature range as in untreated wood (Grønli et al. 2002; Yang et al. 2007; Esteves and Pereira 2009). In this context, TGA has been shown to be a sensitive tool for capturing these modification-induced changes in wood composition and for assessing the degree of thermal modification, particularly through ML in the hemicellulose-dominated temperature range (Korošec et al. 2017).

Fig. 6. Visual appearance of solid residues of reference, thermally modified, acetylated, and surface-charred birch wood samples at selected temperatures during TGA

Acetylated wood exhibited a more pronounced modification of the thermal degradation profile compared with reference and thermally modified samples. In both beech and birch, the onset of rapid mass loss was shifted towards slightly higher temperatures, indicating delayed initiation of thermal decomposition. However, the temperature of the main DTG maximum remained essentially unchanged and occurred at approximately the same temperature as in the reference and thermally modified wood (±380 °C). The primary difference was in the shape of the DTG curves: the peak intensity was reduced and the characteristic shoulder around ~300 °C, visible in reference and TM samples, was strongly diminished or absent in acetylated wood. This behaviour suggests a reduced contribution of early-degrading fractions, which are typically associated with hemicelluloses, and a more gradual overall degradation process.

These effects are consistent with the chemical substitution of hydroxyl groups by acetyl groups, which reduces hygroscopicity, limits the accessibility of reactive sites, and alters the thermal response of polysaccharide components without fundamentally changing the cellulose-dominated decomposition stage (Hill 2006; Rowell 2013). Notably, acetylated wood exhibits a higher DTG peak intensity than the reference and thermally modified groups, indicating a more concentrated ML event during the main decomposition stage despite a similar peak temperature. The stronger response observed for acetylated birch can plausibly be linked to its slightly higher degree of acetylation (Table 1), which would further reduce the contribution of early-degrading hemicellulose-related reactions and shift a larger fraction of the devolatilization into the main cellulose-dominated peak (Hill 2006; Yang et al. 2007; Rowell 2013).

The most pronounced differences were observed for the charred samples. Both charred layers show strongly suppressed ML throughout the temperature range up to approximately 350 to 400 °C, reflecting the extensive thermal degradation induced during the surface charring process. The characteristic hemicellulose-related shoulder and the main cellulose DTG peak were largely absent, particularly in the material taken from the outermost 0.5 mm of the surface, where the DTG curves approached a nearly flat profile. This behaviour indicates that the majority of thermally labile polysaccharides in this surface layer had already been decomposed during direct exposure to the heating source. Sample taken from beneath the surface layer showed only minor differences compared with the outermost charred layer, which can be likely attributed to its slightly greater distance from the heating source and the consequently lower degree of thermal degradation during charring process.

CONCLUSIONS

1. The applied wood modification techniques differed markedly in modification intensity and resulting fire and thermal behaviors. Thermal modification resulted in moderate mass loss (ML) and a reduction in equilibrium moisture content (EMC). Acetylation achieved the highest degree of chemical modification with the strongest reduction in hygroscopicity, whereas one-sided surface charring caused only negligible changes in bulk moisture characteristics, confirming that the modification was confined to a thin surface layer.
2. Despite these differences in modification intensity, none of the treatments altered the Euroclass-related reaction-to-fire classification under the EN ISO 11925-2 (2020) small-flame test; however, clear differences in early-stage fire behavior were observed. Acetylated wood exhibited increased flame spread due to reduced moisture buffering and the release of additional combustible volatiles, whereas surface-charred specimens showed a pronounced reduction in flame propagation, particularly for beech, due to the presence of a continuous, thermally insulating char layer.
3. All modified materials exhibited increased gross heat of combustion compared with reference wood, indicating higher energy content associated with carbon enrichment or chemical modification. However, higher energy density did not translate into improved fire resistance, as total combustion energy reflects potential heat release rather than the kinetics of ignition and flame spread.
4. Thermogravimetric analysis showed that thermal modification retained comparable degradation pathways to reference wood, whereas acetylated wood exhibited altered devolatilization behavior and charred layers displayed strongly suppressed mass loss due to prior thermal degradation.
5. Overall, the results confirm surface charring as the most effective modification for reducing flame spread and highlight the necessity of combining flame-based, calorimetric, and thermal analysis methods for a reliable assessment of fire behavior in modified wood.

ACKNOWLEDGMENTS

This study was funded by the Czech Science Foundation (GAČR), project no. GA24-10430S “Microwave thermal modifications with enhanced surface selectivity for wood-based materials”.

REFERENCES CITED

Altgen, M., Hofmann, T., and Militz, H. (2016). “Wood moisture content during the thermal modification process affects the improvement in hygroscopicity of Scots pine sapwood,” Wood Science and Technology 50, 1181-1195. https://doi.org/10.1007/s00226-016-0845-x

Bartlett, A. I., Hadden, R. M., and Bisby, L. A. (2019). “A review of factors affecting the burning behaviour of wood for application to tall timber construction,” Fire Technology 55, 1-49. https://doi.org/10.1007/s10694-018-0787-y

Borůvka, V., Šedivka, P., Novák, D., Holeček, T., and Turek, J. (2021). “Haptic and aesthetic properties of heat-treated modified birch wood,” Forests 12(8), article 1081. https://doi.org/10.3390/f12081081

Čekovská, H., Gaff, M., Osvald, A., Kačík, F., Kubš, J., and Kaplan, L. (2017a). “Fire resistance of thermally modified spruce wood,” BioResources 12(1), 947-959. https://doi.org/10.15376/biores.12.1.947-959

Čekovská, H., Gaff, M., Osvaldová, L. M., Kačík, F., Kaplan, L., and Kubš, J. (2017b). “Tectona grandis Linn. and its fire characteristics affected by thermal modification of wood,” BioResources 12(2), 2805-2817. https://doi.org/10.15376/biores.12.2.2805-2817

Čermák, P., Baar, J., Dömény, J., Výbohová, E., Rousek, R., Pařil, P., Oberle, A., Čabalová, I., Hess, D., Vodák, M., and Brabec, M. (2022). “Wood-water interactions of thermally modified, acetylated and melamine formaldehyde resin impregnated beech wood,” Holzforschung 76(5), 437-450. https://doi.org/10.1515/hf-2021-0164

Čermák, P., Dejmal, A., Paschová, Z., Kymäläinen, M., Dömény, J., and Rautkari, L. (2019). “One-sided surface charring of beech wood,” Journal of Materials Science 54, 9497-9506. https://doi.org/10.1007/s10853-019-03589-3

Čermák, P., Hess, D., and Suchomelová, P. (2021). “Mass loss kinetics of thermally modified wood species as a time-temperature function,” European Journal of Wood and Wood Products 79, 547-555. https://doi.org/10.1007/s00107-020-01634-6

Cuevas, J., and Maluk, C. (2023). “The role of moisture on the flame extinction potential of cross-laminated timber,” Fire Safety Journal 141, article 103937. https://doi.org/10.1016/j.firesaf.2023.103937

Demirbaş, A. (2001). “Relationships between lignin contents and heating values of biomass,” Energy Conversion and Management 42(2), 183-188. https://doi.org/10.1016/S0196-8904(00)00050-9

Ebner, D. H., Barbu, M.-C., Prokop, O., and Čermák, P. (2024). “Fire resistance of one-sided, surface-charred silver fir and European ash timber,” Forests 15(7), article 1109. https://doi.org/10.3390/f15071109

Ebner, D. H., Barbu, M.-C., Gryc, V., and Čermák, P. (2022). “Surface charring of silver fir wood cladding using an enhanced traditional Japanese Yakisugi method,” BioResources 17(2), 2031-2042. https://doi.org/10.15376/biores.17.2.2031-2042

Engelund, E., and Fredriksson, M. (2021). “Wood modification as a tool to understand moisture in wood,” Forests 12(3), article 372. https://doi.org/10.3390/f12030372

EN 13501-1 (2018). “Fire classification of construction products and building elements – Part 1: Classification using data from reaction to fire tests,” European Committee for Standardization, Brussels.

EN ISO 11925-2 (2020). “Reaction to fire tests – Ignitability of products subjected to direct impingement of flame – Part 2: Single-flame source test,” European Committee for Standardization, Brussels.

EN ISO 1716 (2018). “Determining of the heat of combustion/calorific value,” European Committee for Standardization, Brussels.

Esteves, B., Domingos, I. J., and Pereira, H. P. (2008). “Pine wood modification by heat treatment in air,” BioResources 3(1), 142-154. https://doi.org/10.15376/biores.3.1.142-154

Esteves, B., and Pereira, H. (2009). “Wood modification by heat treatment: A review,” BioResources 4(1), 370-404. https://doi.org/10.15376/biores.4.1.370-404

Grønli, M. G., Várhegyi, G., and Di Blasi, C. (2002). “Thermogravimetric analysis and devolatilization kinetics of wood,” Industrial & Engineering Chemistry Research 41(17), 4201-4208. https://doi.org/10.1021/ie0201157

Gustavsson, L., Pingoud, K., and Sathre, R. (2006). “Carbon dioxide balance of wood substitution: Comparing concrete- and wood-framed buildings,” Mitigation and Adaptation Strategies for Global Change 11(3), 667-691. https://doi.org/10.1007/s11027-006-7207-1

Hasburgh, L. E., and Zelinka, S. L. (2023). “Flammability and acetic acid emissions from acetylated wood under well-ventilated burning conditions,” Forests 14(6), article 1186. https://doi.org/10.3390/f14061186

Hasburgh, L. E., Zelinka, S. L., Bishell, A. B., and Kirker, G. T. (2021). “Durability and fire performance of charred wood siding (Shou Sugi Ban),” Forests 12(9), article 1262. https://doi.org/10.3390/f12091262

Hill, C. A. S. (2006). Wood Modification: Chemical, Thermal and Other Processes. Wiley, Chichester. https://doi.org/10.1002/0470021748

Hill, C., Altgen, M., and Rautkari, L. (2021). “Thermal modification of wood – A review: Chemical changes and hygroscopicity,” Journal of Materials Science 56, 6581-6614. https://doi.org/10.1007/s10853-020-05722-z

Jones, D., and Sandberg, D. (2020). “A review of wood modification globally – Updated findings from COST FP1407,” Interdisciplinary Perspectives on the Built Environment 1, 1-31. https://doi.org/10.37947/ipbe.2020.vol1.1

Korošec, R. C., Renko, S., Rep, G., and Bukovec, P. (2017). “Determination of the thermal modification degree of beech wood using thermogravimetry,” Journal of Thermal Analysis and Calorimetry 130, 1383-1390. https://doi.org/10.1007/s10973-017-6446-6

Kymäläinen, M., Dömény, J., and Rautkari, L. (2022). “Moisture sorption of wood surfaces modified by one-sided carbonization,” Coatings 12(9), article 1273. https://doi.org/10.3390/coatings12091273

Kymäläinen, M., Rautkari, L., and Hill, C. A. S. (2017). “Surface modification of solid wood using contact charring,” Journal of Materials Science 52, 6111-6123. https://doi.org/10.1007/s10853-017-0850-y

Lin, S., Qin, Y., Huang, X., and Gollner, M. (2023). “Use of pre-charred surfaces to improve fire performance of wood,” Fire Safety Journal 137, 103745. https://doi.org/10.1016/j.firesaf.2023.103745

Lowden, L. A., and Hull, T. R. (2013). “Flammability behaviour of wood and a review of char structure characteristics,” Fire Science Reviews 2(4), 1-19. https://doi.org/10.1186/2193-0414-2-4

Luptáková, J., Kačík, F., Mitterová, I., and Zachar, M. (2019). “Influence of temperature of thermal modification on the fire-technical characteristics of spruce wood,” BioResources 14(2), 3795-3807. https://doi.org/10.15376/biores.14.2.3795-3807

Machová, D., Oberle, A., Zárybnická, L., Dohnal, J., Šeda, V., Dömény, J., Vacenovská, V., Kloiber, M., Pěnčík, J., Tippner, J., and Čermák, P. (2021). “Surface characteristics of one-sided charred beech wood,” Polymers 13(10), article 1551. https://doi.org/10.3390/polym13101551

Mantanis, G. I. (2017). “Chemical modification of wood by acetylation or furfurylation: A review of the present scaled-up technologies,” BioResources 12(2), 4478-4489. https://doi.org/10.15376/biores.12.2.4478-4489

Marino, S., Gaff, M., Sethy, A. K., Kamboj, G., Kačík, F., and Hosseini, S. B. (2024). “Enhancing the fire resistance properties of thermally modified Robinia pseudoacacia wood with natural and synthetic flame retardants: Chemical characterization and fire behaviour,” European Journal of Wood and Wood Products 82, 1145-1157. https://doi.org/10.1007/s00107-024-02079-x

Martinka, J., Kačíková, D., Rantuch, P., and Balog, K. (2016). “Investigation of the influence of spruce and oak wood heat treatment upon heat release rate and propensity for fire propagation in the flashover phase,” Acta Facultatis Xylologiae Zvolen 58(1), 5-14. https://doi.org/10.17423/afx.2016.58.1.01

Mensah, R. A., Jiang, L., Renner, J. S., and Xu, Q. (2023). “Characterisation of the fire behaviour of wood: From pyrolysis to fire retardant mechanisms,” Journal of Thermal Analysis and Calorimetry 148(4), 1407-1422. https://doi.org/10.1007/s10973-022-11442-0

Mohebby, B., Talaii, A., and Kazemi Najafi, S. (2007). “Influence of acetylation on fire resistance of beech plywood,” Materials Letters 61(2), 359-362. https://doi.org/10.1016/j.matlet.2006.04.117

Morozovs, A., and Bukšāns, E. (2009). “Fire performance characteristics of acetylated ash (Fraxinus excelsior L.) wood,” Wood Material Science & Engineering 4(1–2), 76-79. https://doi.org/10.1080/17480270903315580

Olarescu, C. M., Campean, M., and Olarescu, A. (2014). “Dimensional stabilization of wood originating from small-diameter trees through heat treatment,” BioResources 9(3), 4844-4861. https://doi.org/10.15376/biores.9.3.4844-4861

Papadopoulos, A. N., and Hill, C. A. S. (2003). “The sorption of water vapour by acetylated wood,” Wood Science and Technology 37, 221-231. https://doi.org/10.1007/s00226-003-0192-6

Papadopoulos, A. N., Tountziarakis, P., and Pougioula, G. (2010). “Fire resistance of two panel products made from chemically modified raw material,” Maderas. Ciencia y Tecnología 12(1), 53-55. https://doi.org/10.4067/S0718-221X2010000100006

Poletto, M., Zattera, A. J., Forte, M. M. C., and Santana, R. M. C. (2012). “Thermal decomposition of wood: Influence of wood components and cellulose crystallite size,” Bioresource Technology 109, 148-153. https://doi.org/10.1016/j.biortech.2011.11.122

Popescu, C. M., Hill, C. A. S., Curling, S. F., Ormondroyd, G. A., and Xie, Y. (2014). “The water vapor sorption behaviour of acetylated birch wood: How acetylation affects the sorption isotherm and accessible hydroxyl content,” Journal of Materials Science 49, 2362-2371. https://doi.org/10.1007/s10853-013-7937-x

Rabe, S., Klack, P., Bahr, H., and Schartel, B. (2020). “Assessing the fire behavior of woods modified by N-methylol crosslinking, thermal treatment, and acetylation,” Fire and Materials 44, 530-539. https://doi.org/10.1002/fam.2809

Rautkari, L., and Hill, C. (2014). “Effect of initial moisture content on the anti-swelling efficiency of thermally modified Scots pine sapwood treated in a high-pressure reactor under saturated steam,” Holzforschung 68(3), 323-326. https://doi.org/10.1515/hf-2013-0078

Rowell, R. M. (2006). “Chemical modification of wood: A short review,” Wood Material Science & Engineering 1(1), 29-33. https://doi.org/10.1080/17480270600670923

Rowell, R. M. (2013). Handbook of Wood Chemistry and Wood Composites, 2^nd Ed., CRC Press. https://doi.org/10.1201/b12487

Sandberg, D., Kutnar, A., and Mantanis, G. I. (2017). “Wood modification technologies – A review,” iForest 10, 895-908. https://doi.org/10.3832/ifor2380-010

Schlamadinger, B., and Marland, G. (1996). “The role of forest and bioenergy strategies in the global carbon cycle,” Biomass and Bioenergy 10(5–6), 275-300. https://doi.org/10.1016/0961-9534(95)00113-1

Schwarzinger, C., and List, M. (2010). “Identification of marker compounds in pyrolysis–GC/MS of various acetylated wood types,” Journal of Analytical and Applied Pyrolysis 87(1), 144-153. https://doi.org/10.1016/j.jaap.2009.11.001

Šeda, V., Machová, D., Dohnal, J., Dömény, J., Zárybnická, L., Oberle, A., Vacenovská, V., and Čermák, P. (2021). “Effect of one-sided surface charring of beechwood on density profile and surface wettability,” Applied Sciences 11(9), 1-10. https://doi.org/10.3390/app11094086

Shen, D. K., Gu, S., and Bridgwater, A. V. (2010). “The thermal performance of the polysaccharides extracted from hardwood: Cellulose and hemicellulose,” Carbohydrate Polymers 82(1), 39-45. https://doi.org/10.1016/j.carbpol.2010.04.018

Spearpoint, M. J., and Quintiere, J. G. (2000). “Predicting the burning of wood using an integral model,” Combustion and Flame 123(3), 308-325. https://doi.org/10.1016/S0010-2180(00)00162-0

Telmo, C., and Lousada, J. (2011). “Heating values of wood pellets from different species,” Biomass and Bioenergy 35(7), 2634-2639. https://doi.org/10.1016/j.biombioe.2011.02.043

White, R. H., and Dietenberger, M. A. (2010). “Fire safety of wood construction,” in: Wood Handbook, USDA Forest Service. https://doi.org/10.2737/FPL-GTR-190

Xing, D., and Li, J. (2014). “Effects of heat treatment on thermal decomposition and combustion performance of Larix spp. wood,” BioResources 9(3), 4274-4287. https://doi.org/10.15376/biores.9.3.4274-4287

Yang, H., Yan, R., Chen, H., Lee, D. H., and Zheng, C. (2007). “Characteristics of hemicellulose, cellulose and lignin pyrolysis,” Fuel 86(12–13), 1781-1788. https://doi.org/10.1016/j.fuel.2006.12.013

Zelinka, S. L., Altgen, M., Emmerich, L., Guigo, N., Keplinger, T., Kymäläinen, M., Thybring, E. E., and Thygesen, L. G. (2022). “Review of wood modification and wood functionalization technologies,” Forests 13(7), article 1004. https://doi.org/10.3390/f13071004

Zeman, A., Alén, R., and Kotilainen, R. (2000). “Thermal behaviour of Scots pine (Pinus sylvestris) and silver birch (Betula pendula) at 200–230 °C,” Wood and Fiber Science 32(2), 138-143.

Article submitted: March 25, 2026; Peer review completed: April 23, 2026; Revisions accepted: May 1, 2026; Published: May 5, 2026.

DOI: 10.15376/biores.21.3.5749-5767