Effect of thermal treatments on technological properties of wood from two Eucalyptus species

The effect of thermal treatments on physical and mechanical properties of rose gum and Sydney blue gum wood was evaluated. Wood samples were thermally modified in a combination: pre-treatment in an autoclave (127°C 1h) and treatment in an oven (180-240°C – 4h); and only treatment in an oven at 180-240°C for 4h. Chemical changes in the structure of woods were evaluated through infrared spectroscopy. Evaluation of physical properties was performed through mass loss, specific gravity, equilibrium moisture content and dimensional stability tests. Surface changes were analyzed through apparent contact angle technique and static bending tests were carried out to evaluate the mechanical behavior. Use of pre-treatment in autoclave affected the properties analyzed, however oven, resulted in the highest changes on wood from both species. Chemical changes were related to the degradation of hemicelluloses. Moreover, a significant decrease of hygroscopicity and mechanical strength of thermally modified woods was observed, while specific gravity did not significantly change for either of the species studied. The best results of decrease of wettability were found in low temperatures, while dimensional stability increased as a function of temperature of exposure in oven. The highest loss of mechanical strength was observed at 240°C for both species.


INTRODUCTION
Over the years, many processes have been developed in order to protect wood against xylophages agents, mainly treatments of impregnation with chemical products.Environmentally friendly and efficient alternatives towards changes of technological properties of wood were also developed, such as thermal treatments.According to Syrjanen (2001), thermal treatment is an ecological alternative to modify wood and can be used to produce many products like doors, windows, kitchen furniture, garden furniture and floors.
Thermal treatment occurs between 180 and 260°C due to temperatures lower than 140°C do not significantly affect the structure of material, and temperatures higher than 260°C result in PEDRO HENRIQUE G. DE CADEMARTORI et al. undesirable degradation (Hill 2006).According to Mitchell (1988), andKorkut andGuller (2008), the temperature is the parameter of process responsible for the highest effect of modification in the properties of thermally treated wood.
Many methods of thermal treatments were developed in different countries, mainly on the European continent, such as Finnish Thermo Wood (Finland), French Rectification and Bois Perdure (France), Oil Heat Treatment (Germany) and Dutch Plato Wood (Holland) (Militz 2002).
On the one hand, thermal treatment promotes changes in the physical properties of wood, especially as swelling and shrinkage, equilibrium moisture content, weathering resistance, color and mass loss.In this context, physical changes of thermally modified woods are highly dependent on the conditions of treatments (Hill 2006, Korkut et al. 2008).At the same time, thermal treatments modifying the wood's wettability, which is directly related to the chemical changes on wood structure, especially to the degradation of the hemicelluloses.
On the other hand, thermal treatments can change mechanical properties and, consequently, according to Borrega and Kärenlampi (2008), can limit the structural use of wood pieces jointly with mass loss.Moreover, breaking of hemicelluloses and cellulose chains can significantly reduce wood strength (Poncsák et al. 2006).According to Fengel and Wegener (2003), the components of thermally modified wood are stable up to 100°C.However, polysaccharides content decreases when exposed to high temperatures (above 100°C) due to its more sensitive reactivity in comparison to cellulose.
Wood from rose gum is rated as strength class S3 and SD4 and wood from Sydney blue gum is rated as class S3 and SD4 (AUSTRALIAN/NEW ZEALAND STANDARD 2000).Therefore, thermal treatments could be a great alternative to improve physical and aesthetic properties, increasing market value and, consequently, the number of applications of wood from rose gum or Sydney blue gum.
Considering these facts, the present study aimed to evaluate physical, surface and mechanical properties of rose gum (Eucalyptus grandis Hill ex.Maiden) and Sydney blue gum (Eucalyptus saligna Sm) woods thermally modified under different conditions.

RAW MATERIAL
Six rose gum (Eucalyptus grandis Hill ex.Maiden) and six Sydney blue gum (Eucalyptus saligna Sm) trees were randomly selected from experimental fast-growing population located in the north coast of the state of Rio Grande do Sul, southern Brazil (30°14'09.73''S,50°19'55.07"W).Rose gum trees were 17 years old, average diameter at breast height (DBH, 1.30 m) of 37.1 cm and total height of 35.7 m; Sydney blue gum trees were 17 years old, average diameter at breast height (DBH, 1.30 m) of 29.5 cm and total height of 30 m.
Two hundred and seventy samples (15 samples per treatment for each specie) measuring 16 x 16 x 250 mm (radial x tangential x longitudinal) were cut from the first log (3.15 m length) of each tree.The samples were prepared only with straight grain, heartwood and absence of warps.All the samples were conditioned in a climatic chamber (20°C and 65% of relative humidity) to stabilize the equilibrium moisture at 12%, which was the initial point for the thermal treatments.

THERMAL TREATMENTS
Table I shows the eight distinct thermal treatments applied in rose gum and Sydney blue gum woods.Treatment 1 (control) was kept in a climatic chamber (20°C and 65% of relative humidity) the whole time.
One hundred and twenty wood samples were modified through combined treatments in an autoclave sterilizer and in an oven.To achieve this, a pre-treatment (wet conditions and indirect steam) in an autoclave at 127°C and 1.5 kgf.cm -2 for 1 hour was carried out.Pre-treated samples were conditioned in a climatic chamber to stabilize the equilibrium moisture.Then, these samples PROPERTIES OF TWO THERMALLY TREATED EUCALYPTS were treated in an oven (heating rate: 2°C.min -1 ) without forced air circulation at 180, 200, 220 and 240°C for 4 hours (dry conditions) after reaching the temperature of treatment.
Another one hundred and twenty wood samples were treated only in an oven with heating rate of 2°C.min -1 and without forced air circulation at 180, 200, 220 and 240°C for 4 hours after reaching the temperature of treatment.All the treated samples were kept in a climatic chamber (20°C and 65% of relative humidity) until they reached the equilibrium moisture content to perform the tests.

INFRARED SPECTROSCOPY
The chemical modifications were measured through attenuated total reflectance infrared spectroscopy (ATR-IR) using an equipment by direct transmittance at a resolution of 4 cm -1 for 32 scans in the range from 700 to 4000 cm -1 .The main evaluations occurs in the range 1800-700 cm -1 , which is known as fingerprint of wood (Pandey 1999).All the spectra were generated through an average of six spectra for each treatment.

MASS LOSS, EQUILIBRIUM MOISTURE CONTENT AND SPECIFIC GRAVITY
Mass loss (WL) was measured through the difference of weight of the samples before and after the thermal treatments (Cademartori et al. 2012).
Equilibrium moisture content (EMC) was measured using American Society for Testing and Materials D143-94 standard (ASTM 2000).Likewise, specific gravity (ρ b ) was measured by ratio of dry weight and wet volume.
DIMENSIONAL STABILITY AND WETTABILITY Dimensional stability was evaluated through volumetric (ΔV) and linear (Δl in radial and tangential direction) swelling.To achieve this, the samples (in equilibrium) measuring 16mm x 16mm x 50mm (radial x tangential x longitudinal) were immersed in distilled water until the fibers' saturation.After the immersion, weight and size (length, width and thickness) of the samples were measured.
Then, the samples were dried in an electrical oven at 103±2°C until they reached constant weight.The physical properties of volumetric (ΔV) and linear (Δl) swelling were measured as described by Cademartori et al. (2012).
Wettability was studied through the apparent contact angle technique using a goniometer (sessile drop method) as described by Cademartori et al. (2013).Measurement of the apparent contact angle were performed through the deposition of a distilled water droplet (5µl) in three distinct surface points of both radial and tangential direction of each of five samples (16mm x 16mm x 250mm) per treatment.The apparent contact angle was measured in seven distinct times with 15 seconds intervals up to a total of 90 seconds.The first measurements was after 5 seconds of the droplet contact on the wood surface.

MECHANICAL PROPERTIES
Mechanical properties (modulus of elasticity and modulus of rupture) of thermally modified wood samples (16mm x 16mm x 250mm) were measured by static bending tests in a universal machine with capacity of 300kN.The tests were performed using a three-point bending apparatus according to American Society for Testing and Materials D143-94 standard (ASTM 2000).

STATISTICAL ANALYSIS
The collected data were analyzed by descriptive statistic and factorial analysis of variance with additional treatment in a factorial arrangement 2 x 4 + 1, i.e., use of steam pre-treatment in autoclave, temperature of treatment and control treatment, respectively.Due to the absence of a control treatment, mass loss (WL) was analyzed by analysis of variance in a factorial arrangement 2 x 4. For the analysis of variance, an induced p-value by F statistics at 5% and 1% of probability of error was used.When the null hypothesis (p<0.05) was rejected, the mean values were compared with Tukey Test at 5% of probability of error.

INFRARED SPECTROSCOPY (ATR-IR)
As observed in Fig. 1, the changes in peaks were only similar for the treatment in oven and in a combination of oven and autoclave.When compared to treatment 1 (control), thermal treatments significantly changed the chemical structure of rose gum and Sydney blue gum wood.
The high changes as a function of temperature of treatment were found in the peaks related to the hemicelluloses, mainly at 1731 cm -1 , which is typical in unconjugated ketone and carboxyl acids present in lignin and hemicelluloses (Colom et al. 2003).After treatment 8 and 9, the intensity of this peak was very small due to the action of the thermal treatment on the chemical structure of wood.According to Fengel and Wegener (2003), when wood is exposed to high temperatures, the polysaccharides content significantly decrease because have more sensible reactivity than the cellulose.The peak at 1105 cm -1 (OH association in hemicelluloses and cellulose) increases with increasing temperature of treatment for both species, which was also found by Ding et al. (2011) in a study with thermally modified Mongolian pine wood.
At 1592, 1504 and 1460cm -1 , peaks representative of lignin showed a stabilization or slight increase as a function of the temperature of treatment.This is related to the degradation of the polysaccharides and, consequently, the increase of proportion of lignin in the structure of thermally modified woods.

PHYSICAL PROPERTIES
Statistical analysis of physical properties (Table II) showed that most of the parameters evaluated were significantly influenced by the thermal treatments for both species.Nevertheless, specific gravity (ρ b ) did not significantly vary as a function of pre-treatment in the autoclave and temperature of heat treatment in oven factors.Pre-treatment in the autoclave was significant only for mass loss (WL) of both species studied and the equilibrium moisture content (EMC) of Sydney blue gum wood.The contrast Fact.vs Control was significant for all the properties evaluated, except for ρ b , which proves the influence of thermal treatments applied on hygroscopicity of rose gum and Sydney blue gum woods.Mass loss (WL) of thermally modified woods from two Eucalyptus species showed directly proportional behavior as a function of temperature of exposure (Table III), i.e, mass loss increased with increasing temperature of treatment in oven.
Mean values of WL significantly ranged from 12 to 25% for all the species, in which rose gum wood showed the lowest (12.63%), while Sydney blue gum wood presented the highest (24.47%).

TABLE III Mean values of mass loss (WL), specific gravity (ρ b ) and equilibrium moisture content (EMC) of rose gum and Sydney blue gum thermally modified woods.
Mean values in the same column followed by the same lowercase letter or in the same line followed by the same uppercase letter for each parameter evaluated are not statistically different at level of 5% by the Tukey test.Values between parentheses corresponding to the standard deviation.AC = pre-treatment in autoclave; S/ AC = absence of pre-treatment in autoclave.PEDRO HENRIQUE G. DE CADEMARTORI et al.Mass loss occurs in high levels when wood is thermally modified in closed systems.However, this depends of temperature and atmosphere of treatments.Moreover, high levels of mass loss could be observed in treatments with air, wet conditions and hardwoods.On the other hand, low levels of mass loss occurs in treatments with inert atmospheres, open systems, dry conditions and softwoods (Hill 2006).Obtainment of high mean values of mass loss in this study was possibly due to the initial point of thermal treatments, which was the equilibrium moisture content (~12%).Therefore, part of WL is attributed to loss of water in the anatomical structure of wood.According to Surini et al. (2012), mass loss has high correlation to the moisture content of thermally treated wood.Bal and Bektas (2012) stated that the higher the mass loss, the higher the physical and mechanical modifications of the material will be.
According to Table III, ρ b did not show significant differences, which was also found by Calonego et al. (2012) in a study with rose gum.The mean values were 0.38-0.42g.cm -3 (rose gum) and 0.46-0.52g.cm -3 (Sydney blue gum).
Treatment at 180°C was sufficient to modify significantly the hygroscopicity of wood from the two species.This result was also found in other studies (Calonego et al. 2012, Esteves et al. 2007, Garcia et al. 2012).Moreover, from treatment at 220°C, values of EMC (Table III) significantly ranged between treatments with or without use of autoclave for Sydney blue gum wood, while this behavior was observed only in treatment at 240°C for rose gum wood.
The changes in EMC could be explained due to the decrease of OH groups, cleavage of the chains and loss of substances after exposure to high temperatures (Akyildiz and Ates 2008).The reduction of EMC in the present study was inversely proportional to the WL, which was also found by Esteves et al. (2007).
Volumetric swelling (ΔV) decreased significantly as a function of temperature of treatment in oven (Table IV) for both species studied, which confirms the modifications in the wood hygroscopicity.Pretreatment in autoclave was only influenced in the treatment at 200°C for Sydney blue gum wood.The highest reduction of ΔV in both species was observed in more drastic temperatures applied, which was ~ 64% and ~ 65% in rose gum and Sydney blue gum woods, respectively.
Linear swelling (Δl) in both radial and tangential direction for both rose gum and Sydney blue gum woods showed significant reduction.Pre-treatment in autoclave was significant only for Sydney blue gum wood, while temperature of exposure in oven was significant in both radial and tangential directions of both species studied.The highest reduction of Δl in radial section was ~60% and in tangential section was ~70%, both measured in rose gum wood treated at 240°C.Bak and Németh (2012) reported that changes in sorption of thermally treated wood are highly influenced by species.However, rose gum and Sydney blue gum showed similar behavior of dimensional stability in the present study, possibly due to homogeneous degradation of amorphous region in hemicelluloses, which was confirmed in the infrared spectroscopy analysis previously explained in this study.According to Ding et al. (2011), dimensional stability is determined mainly due to number of OH free groups on the amorphous polysaccharides present in the wood structure.Other factors can influence the level of swelling on wood such as permeability.Taghiary et al. ( 2011) observed that permeability of beech wood drastically decreased after exposure to hydrothermal and hygrothermal conditions due to the deposition of extractives on vessel perforations and blocking of fluid transfer through the porous of wood by cell walls.WETTABILITY Fig. 2 shows the kinetic of wetting for untreated and thermal treated samples of rose gum and Sydney blue gum woods.The thermal treatments significantly PROPERTIES OF TWO THERMALLY TREATED EUCALYPTS influenced the apparent contact angle of treated samples when compared to the untreated samples (treatment 1) in both radial and tangential sections.The hydrophobicity of thermally treated wood samples increased without significant differences between radial and tangential sections as described by Kocaefe et al. (2008).
The untreated samples showed a significant and constant decrease of the apparent contact angle as a function of the time, while the thermally treated wood samples presented low decrease rates, mainly in the treatments 2, 3, 4, 5, 6 and 7.An apparent contact angle lower than 90° in porous materials (wood for example) implies in a spontaneous penetration of the liquid (Wålinder and Gardnerb 1999), which was observed for the treatment 1 (control) in this study.According to Homan et al. (2000), the limited interaction between thermally treated wood and water is related to the reduction of the OH groups in the structure of hemicelluloses.
The highest increase of the apparent contact angle was observed in treatments 2, 3, 4 and 5. Treatments 8 and 9 (more aggressive) were higher than treatment 1 (control) but were lower than the treatments aforementioned (2, 3, 4 and 5).Therefore, it is possible to affirm that high temperatures applied in the thermal treatments are not always the best condition to obtain great changes in the material surface.Oliveira et al. ( 2010) described the same wettability behavior found in this study.Pétrissans et al. (2005) suggested that wettability decreases due to increase of cellulose crystallinity, whereas Hakkou et al. (2005) reported that changes in the conformational arrangement of wood is related to the lignin plasticization or the residual water.
Quality of wetting is influenced by many macroscopic characteristics, such as porosity, surface roughness, moisture content and fibers orientation (Kocaefe et al. 2008).Therefore, the most severe treatments applied in this study generated small cracks in the samples surface, which could have had an influence in the wood characteristics such as roughness and porosity and, consequently, in wettability.
The wettability of thermally treated wood is widely related to the material surface inactivation, which is associated to the reduction of attractive  Mean values in the same column followed by the same lowercase letter or in the same line followed by the same uppercase letter for each parameter evaluated are not statistically different at level of 5% by the Tukey test.Values between parentheses corresponding to the standard deviation.AC= pre-treatment in autoclave; S/ AC= absence of pre-treatment in autoclave.
forces.According to Christiansen (1991), mechanisms such as molecular surface reorientation, extractives migration for the surface, micropore closure, elimination of surface hydroxyl bonding sites and surface oxidation directly influence the reduction of attractive forces.

MECHANICAL PROPERTIES
Statistical analysis of mechanical properties (Table V) showed that interaction between pre-treatment in the autoclave and temperature of heat treatment in oven factors was significant only for modulus of rupture (MOR) of rose gum.The contrast Fact.vs Control showed that both MOR of rose gum and Sydney blue gum woods significantly changed after thermal treatments, which implies that mechanical properties of thermally modified wood samples were distinct when compared to the untreated wood samples.
Modulus of elasticity (MOE) of thermally modified woods (Table VI) did not show a tendency as a function of thermal treatments and did not present significant difference in comparison to control treatment, which as previously observed in Table V.The absence of significant range in stiffness is related to the characteristics of the wood developed during the thermal treatment.Bhuiyan et al. (2000) reported Modificações na superfície do material foram analisadas pela técnica de ângulo de contato e o comportamento mecânico foi avaliado com testes de flexão estática.O uso de pré-tratamento em autoclave afetou as propriedades analisadas, no entanto a exposição em estufa denotou as maiores modificações na madeira de ambas a espécies.
As modificações químicas foram relacionadas com a degradação das hemiceluloses.Além disso, observou-se redução significativa da higroscopicidade e da resistência mecânica da madeira termorretificada, enquanto que a massa específica básica não possuiu modificações significativas para a madeira das duas espécies estudadas.Mean values in the same column followed by the same lowercase letter or in the same line followed by the same uppercase letter for each parameter evaluated are not statistically different at level of 5% by the Tukey test.Values between parentheses corresponding to the standard deviation.AC= pre-treatment in autoclave; S/ AC= absence of pre-treatment in autoclave.

TABLE VI Mean values of modulus of elasticity (MOE) and modulus of rupture (MOR) of rose gum and Sydney blue gum thermally modified wood.
Os melhores resultados de redução da molhabilidade foram encontrados para as temperaturas mais baixas, enquanto que a estabilidade dimensional elevou-se em função da temperatura de tratamento em estufa.A maior perda de resistência mecânica foi observada no tratamento a 240°C para ambas as espécies.

Figure 1 -
Figure 1 -ATR-IR spectra of untreated and thermally treated wood.

Figure 2 -
Figure 2 -Kinetic of wetting for untreated and thermally treated wood samples.

TABLE I Thermal treatments performed in rose gum and Sydney blue gum woods
.PEDRO HENRIQUE G. DE CADEMARTORI et al.

TABLE II Summary of the analysis of variance of the physical properties of thermally treated rose gum and Sydney blue gum woods.
Fact = Factorial; Control = Control treatment (untreated samples); ns = not significant; ** = Significant at 1% of probability of error; * = Significant at 5% of probability of error.