Material and heat treatment

Teak (Tectona grandis L. f.) wood samples with nominal dimensions of 150 x 75 x 20 mm³ were obtained from six trees of 12 years old. The material was separated in heartwood and sapwood, air dried and conditioned at 20ºC and 65% RH in a climate chamber until constant mass.

The heat treatment was carried out in an electric laboratorial muffle furnace from Linn Elektro Therm, with dimensions of 600 x 600 x 700 mm³ equipped with a system of temperature and time control. The treatment was conducted in four steps at two different temperatures: 180 and 200ºC. The four steps were: (1) heating up to 100ºC for 2 h; (2) increase of temperature from 100ºC to final temperature (180 or 200ºC) for 30 min; (3) treatment time in the selected temperature for 2 h and 30 min; and (4) cooling for approximately 1 h. The initial moisture content of the samples was 8-10% (based on the oven-dry of the wood). Heat-treated samples were conditioned at 20ºC and 65% RH until constant mass. The moisture content of the heat-treated heartwood and sapwood samples was 4% to 4,8%.

Contact angle analysis

The contact angle analysis was assessed by the DSA100 drop shape analyzer, version 1.92 from KRÜSS GmbH (Hamburg, Germany). The liquid probe was distilled water at 20ºC which was applied with a dosing syringe of 100 μl, needle of 0,5 mm in diameter at 3 mm height from sample surface. A 5 μl drop volume was standardized for all samples. The drop shape was analyzed every 5 s for 120 s, resulting in 24 readings at each point. Contact angle measurements were made on five points per sample with five replicates per treatment. Three variables were evaluated: initial contact angle (CA_initial) = 5 s after water droplet deposition on the sample surface; final contact angle (CA_final) = 120 s after water droplet deposition on the sample surface; and mean contact angle (CA_mean) = average of 24 readings taken for 120 s.

Experimental design and statistical analysis

The experimental design was composed of two factors (independent variables): wood type (heartwood and sapwood) and temperature (control, 180ºC and 200ºC), resulting in six treatment combinations. The dependent variables were the initial contact angle (CA_initial), final contact angle (CA_final) and mean contact angle (CA_mean).

Statistical analyzes were performed at Statistic 10.0. CA_initial and CA_final data were subjected to analysis of variance (Anova) and Tukey’s range test at 95% probability level. The normality and homoscedasticity were verified by Shapiro-Wilk and Levene’s tests, respectively, at 95% probability level. A logarithmic transformation (log10) was used to CA_final (heartwood + sapwood). CA_mean data did not meet the Anova assumptions when both independent variables (wood type and temperature) were included together in the analysis therefore they were analyzed separately. An Anova was performed to evaluate the effect of the temperature on CA_mean in each wood group and the Wilcoxon-Mann-Whitney nonparametric test was used to compare heartwood and sapwood.

Surface wettability

Behavior of the contact angle as a function of time

Figure 1 shows the behavior of the contact angle in heartwood and sapwood as a function of time for different heat treatments. Contact angle values of heartwood are higher than sapwood for a same treatment condition. The lower the contact angle the greater is the wettability of the surface of the material. Surfaces with contact angles of 0º have a complete wettability; surfaces with finite angle values are partially wettable (^{Daltin 2011}); and a fully hydrophobic surface has an angle of 180º. Therefore, the teak heartwood presents lower wettability than the sapwood (Figure 2). ^{Wang et al. (2015}) observed a similar behavior for poplar heartwood and sapwood (Populus cathayana), with smaller contact angle values for the sapwood during the surface wetting. The teak heartwood has a large amount of hydrophobic extractives (^{Yamamoto et al. 1998}) and this may explain its lower surface wettability.

Figure 1 Contact angle behavior of the teak heartwood and sapwood as a function of time at different heat treatments. 

Approximately up to 40 s, the curves showed a greater slope than the starting point for all treatments indicating that the greatest contact angle variations occurred at the beginning (Figure 1). After 40 s, the curves continued to decrease, but in a linear and constant way until reaching an equilibrium point close to 90 s (Figure 1). The wettability on the heterogeneous surface of the wood is complex since penetration of the liquid by capillarity occurs. According to the Forest Products Laboratory (2010), a surface with good wettability and good penetration absorbs the drop in up to 20 s; if the drop spreads but the water remains on the surface after 40 s, it has good wettability and low penetration, the adhesion may be difficult; and if after 40 s, the drop remains with its original shape, with little spreading, it can present problems of adhesion due to the surface inactivation in which there is low wettability and low penetration. For example, on the heartwood surface treated at 200ºC, the drop remains practically with the same shape until 120 s, indicating the presence of an inactive surface (Figure 2). On the other hand, on the untreated sapwood surface, a good drop spreading occurs indicating a greater wettability (Figure 2).

Figure 2: Drop shape changing on the teak heartwood and sapwood surfaces as a function of time to different heat treatments. 

As the heat treatments became more severe, the wood surface presented lower wettability, resulting in higher contact angle values. Figure 2 shows the reduction in drop volume as a function of time due to the penetration of the liquid into the porous structure of the wood. ^{Wang et al. (2015}) observed a similar behavior for two wood species (poplar (Populus cathayana Rehd.) and Scots pine (Pinus sylvestris L.)) treated under different temperatures (160, 180, 200, 220 or 240ºC) for 4 h. The heat treatment decreased the sapwood wettability of both wood species and the contact angle increased with the temperature increase (Wang et al. 2015). The contact angle in the sapwood treated at 200°C was similar to the untreated heartwood (Figure 1).

Initial, final and average contact angle

Table 1 shows the Anova results of the CA_initial and CA_final. Significative differences were found between heartwood/sapwood and temperature levels for CA_initial and CA_final. However, no significant interaction was observed between the independent variables (Table 1).

Table 2 shows the means values and standard deviation of the CA_initial and CA_final of the untreated and heat-treated heartwood and sapwood. Heartwood had higher CA_initial and CA_final values than sapwood when compared at a same temperature (Table 2). For example, untreated heartwood had an average CA_initial of 68,42° while that of untreated sapwood was 57,87°. The average CA_final was of 49,28º for untreated heartwood and 37,32º for untreated sapwood (Table 2) showing that untreated heartwood has lower surface wettability than untreated sapwood.

Table 2: Descriptive statistical analysis (means and standard deviation) at initial (CA_inicial) and final (CA_final) contact angle of the teak wood. 

Means with the same letter are not significantly different by the Tukey test. Values between brackets = standard deviation.

Table 3 presents the means values and standard deviation for CA_mean of the untreated and heat-treated heartwood and sapwood. The Wilcoxon-Mann-Whitney test (Z value) showed significant differences between CA_mean of the heartwood and sapwood where the sum of ranks of the heartwood was higher than sapwood with values of 7186 and 4139, respectively (Z= 5,72; p <0,05) (non-tabulated data). The CA_mean of the untreated heartwood was 60,90º while that of the untreated sapwood was 45,72º (Table 3). The lower wettability of the untreated heartwood may be explained by the high extractive content. According to ^{Gardner et al. (1995}), extractives have a complex composition and the extractive group has a relevant effect on the wettability of the wood surface because some of them, when present at the interface (extractive-liquid), can orientate their hydrophilic functional group to the polar group of the liquid establishing hydrogen bonds, and thus provide an apparent change in the surface tension of the liquid, or even in the interfacial tension between wood-extractive-liquid decreasing or increasing the surface energy of the wood and consequently causing a higher or lower wettability.

Table 3: Descriptive statistical analysis (means and standard deviation) and F values to average contact angle (CA_mean) of the teak wood. 

Means with the same letter are not significantly different by the Tukey test. * Significant at 0,05 probability level.

The effect of temperature increase is remarkable in CA_initial and the CA_final. The temperature increase caused a gradual increase in the contact angle independent of wood type (Table 2). For the CA_initial, only wood treated at 200ºC was significantly different from untreated wood for both heartwood and sapwood, while treatments at 180ºC and 200ºC did not have difference between them. The treatment at 200ºC caused an average increase of 12% and 18% in the CA_initial of the heartwood and sapwood, respectively.

For CA_final, heartwood treated at 200ºC presented an increase of 30,5% when compared to untreated heartwood. The sapwood presented an average increase in the CA_final of 17% and 33% after treatment at 180°C and 200°C, respectively (Table 2). These results show that the treatment at 180°C does not significantly affect the wettability of the teak wood surface but on the other hand the surface wettability is decreased following the treatment at 200°C.

The CA_mean of the heartwood had an average increase of 27% following treatment at 200°C while for sapwood, it average increase about 18% at the same temperature (Table 3). However, the treatment at 180ºC did not affect the CA_mean of the heartwood or sapwood.

Heartwood treated at 200ºC showed the lowest wettability while untreated sapwood showed the highest wettability (Table 2). The low wettability following the treatment at 200ºC is due to the surface inactivation. Our results corroborate with those obtained by other authors (^{Sernek et al. 2004}, ^{Kocaefe et al. 2008}, ^{Cademartori et al. 2013}) to different wood species. A lower surface wettability was found to the Fraxinus americana and Acer rubrum woods treated at 205-215ºC (Kocaefe et al. 2008) and Eucalyptus cloeziana wood treated at 180-240°C (Cademartori et al. 2013). The surface inactivation is caused by the migration of hydrophobic extractives in a natural process of the wood or by physic-chemical modifications (Forest Products Laboratory 2010). Sernek et al. (2004) study the chemical modification of the heat-treated wood assessed by X-ray photoelectron spectroscopy (XPS) which provides the elemental composition (carbon and oxygen) of the wood surface. These authors evaluated the surface of two species (Liriodendron tulipifera and Pinus taeda) treated at 156ºC, 172ºC and 187ºC and observed that the wood exposed to the most severe temperature had the highest C1/C2 ratio in comparison to the untreated wood - C1 component corresponds to extractives and lignin and C2 component may arise from all constituents of the wood, but predominantly from cellulose. Also, ^{Wang et al. (2015}) evaluated the relation between the dynamic wetting behavior and chemical components of the heat-treated woods using three parameters (mass loss, C1/C2 ratio and O/C ratio) and they observed an increase in the C1/C2 ratio and mass loss and a decrease in the O/C ratio. The dynamic wetting showed a better correlation with O/C ratio than C1/C2 ratio, indicating that the amount of OH^- groups on wood surface played a fundamental role in the dynamic behavior of wood surface wetting. The dehydration reactions of the hemicelluloses were able to degrade OH^- groups during the heat treatment, which caused a decrease in water penetration on the wood surface.

Pétrissans et al. (2003) studied the chemical modifications of four wood species treated at 240°C assessed by ¹³C nuclear magnetic resonance (NMR) analysis and observed important differences between untreated and heat-treated woods at the level of crystallinity of carbohydrates. The C4 and C6 carbons of the cellulose resulted in two typical peaks of 89 and 85 ppm for the crystalline cellulose region and 66 and 65 ppm for the amorphous region in the treated and untreated woods, respectively. The considerable increase in peaks at 89 and 66 ppm in the heat-treated wood indicates greater crystallinity which leads to a more hydrophobic material. Several studies have reported that the sorption of water by amorphous cellulose is higher than for crystalline celluloses, and therefore the higher hydrophobicity of heat-treated wood can be explained also by the increase cellulose crystallinity (^{Pizzi and Stephanou 1993}).