An effective technique for constructing wood composite with superior dimensional stability

Materials

Poplar wood (Populus euramericana cv. “I-214”) with dimensions of 20×20×20 mm³ (longitudinal×radial×tangential) was prepared from sapwood sections. IA, acetone, ethanol, styrene and 2,2′-azobis (2-methylpropionitrile) (AIBN) were purchased from Sigma Aldrich (Beijing, China). Deionized water (DW) was used throughout the study.

Preparation of the esterification catalyst La³⁺~SO4²⁻/TiO₂-SiO₂

The esterification catalyst was prepared according to a previous paper (Li et al. 2013): 10 g mixture of Ti(SO₄)₂ and Na₂SiO₃ were dissolved in 90 g DW under vigorous stirring, and the molar ratio of Ti and Si was 25:1. Aqueous ammonia (28 wt%) was added drop-wise to adjust the pH value of the aforementioned solution to the range of 9–10. The addition continued until slowly reaching solidification. Subsequently, the obtained TiO₂-SiO₂ solid was separated and washed with DW, and was dried at 100°C for 24 h. An appropriate amount of La₂O₃ was dissolved in 2 mol l⁻¹ H₂SO₄ to prepare a mixture solution, and the concentration of the La³⁺ ion was 0.07 mol l⁻¹. The TiO₂-SiO₂ was immersed into the aforementioned mixture solution for 2 h and its ratio was 15 g TiO₂-SiO₂ per 100 ml solution. Finally, the precipitate was evaporated and calcined at 500°C to obtain La³⁺~SO4²⁻/TiO₂-SiO₂.

Activation of wood using water-soluble IA

Before activation, all wood samples were oven-dried at 103±2°C for 6 h, and the oven-dried weight and dimensions of the samples were measured. The following abbreviations will be used: natural wood (W), IA-activated wood (W_IA) and in situ IA-activated wood (W_IA/PS). The wood samples were activated as follows: dried wood cubes were evacuated in a glass container under high vacuum for 30 min. In another beaker, 20 g IA and 0.2 g La³⁺~SO4²⁻/TiO₂-SiO₂ were dissolved in 1000 ml DW at room conditions under stirring. Subsequently, the aforementioned solution was squashed into the glass container having the wood samples under atmospheric pressure. The glass container was vacuumed, sealed and maintained at 55°C for 2.5 h and then cooled to room temperature. After removing residual chemicals, the treated wood was dried at 103±2°C until weight constancy. Thus, activated wood samples were obtained, and the weight percent gain (WPG) and dimensions were determined.

Graft copolymerization of styrene monomers

The styrene oligomer solution was prepared by mixing styrene monomer and ethanol at a volume ratio of 1:1, containing 1 wt% AIBN as the initiator. The activated wood samples were placed at the bottom of a glass container and immersed in the aforementioned styrene oligomer solution. Then the glass container was vacuumed and heated at 70°C in water bath for 1.5 h; thus, the activated wood samples were subjected to vacuum-heat immersion treatment. After reaction completion, the wood cubes were washed with acetone, then with water and finally dried in the oven at 103±2°C for 24 h. The WPG and dimensions were measured.

Characterization

The chemical compositions of both natural wood and modified wood were qualitatively analyzed by the KBr pellet technique using a Fourier transform infrared (FTIR) spectrometer (Tensor 27, Bruker, Ettlingen, Germany). Mid-IR spectra were obtained by averaging 32 scans from 2000 to 400 cm⁻¹ at a spectral resolution of 4 cm⁻¹. For thermogravimetric analysis (TGA), roughly 5 mg of powdered wood samples were analyzed on a thermogravimetric analyzer (DTG-60, Shimadzu, Kyoto, Japan) (10°C min⁻¹ in N₂ atmosphere, 50–600°C). The microstructure morphologies of the modified and reference wood samples were visualized using a field emission scanning electron microscope (FE-SEM, SU 8010, Hitachi, Tokyo, Japan) operating at 3.0 kV. X-ray diffraction (XRD) analysis was carried out using an XRD Bruker D8 ADVANCE (Germany) (Cu Kα radiation with graphite monochromator, 40 kV and 40 mA). The patterns were obtained between 5° and 40° with 0.05° steps and a scan speed of 2° min⁻¹. According to the peak height method developed by Segal et al. (Segal et al. 1959) for native cellulose, the XRD crystallinity index (CrI) was calculated from the following height ratio:

where I₀₀₂ was the maximum intensity of the 002 diffraction peak at 2θ=22.5° and I_am was the height of the minimum at 2θ=18°.

Mechanical properties

Twelve specimens of natural and modified wood were analyzed using a universal mechanical testing machine (AG-100KNIMO, Tokyo, Japan). Tests were carried out according to the following Chinese standards: GB/T 1936.1–2009, GB/T 1935–2009 and GB/T 1933–2009. The surface hardness was tested using a TH210 durometer (Beijing TIME High Technology Ltd., Beijing, China) and the data were expressed using shore D hardness according to ASTM D2240. The mean result of 20 specimens was considered the final hardness.

Hygroscopicity (H), anti-swelling efficiency (ASE) and leachability

The H of both natural and modified wood was evaluated according to the Chinese standard method for determination of the water absorption of wood (GB/T 1934.1–2009). Specifically, the specimens were oven-dried and then were immersed in water for 10 days at room condition. After immersion, the excess water on the surface was removed using a soft cloth, and specimens were weighed immediately. The water uptake by the specimens increases with the immersion time. The increase in weight was calculated according to the following formula:

where W_t is the weight after immersion in water for a specific time period, and W_o is the weight of the original oven-dried sample.

The ASE of natural and modified samples was measured according to ASTM-1037 (1999) (Adebawo et al. 2016). Ten oven-dried specimens for each treatment with a dimension of 20×20×10 mm (radial×tangential×longitudinal) were immersed in a water bath at a temperature of 20±1°C for 240 h; weight and dimension of specimens were determined before and after soaking.

ASE was calculated according to equations (3) and (4) as follows:

where S_n=volumetric swelling coefficient of natural wood samples; S_m=volumetric swelling coefficient of modified wood samples;

The volumetric swelling coefficient (S) was calculated as follows:

where V₁=volume of wood before soaking; V₂=volume of water after soaking.

After 10 days of immersion, all specimens were oven-dried at 103±2°C to a constant weight before measurement. Leachability (L) due to water immersion was calculated as follows:

where M₀ and M₁ are the dry weights of modified wood specimens before and after immersion, respectively.

Physical and mechanical properties

The physical and mechanical properties of the W, W_IA and W_IA/PS are shown in Table 1. Both W_IA and W_IA/PS have a higher density than W. The air-dried density was increased by 11.63% and 48.84% from 0.43 g·cm⁻¹ to 0.48 g·cm⁻¹ and 0.64 g·cm⁻¹, respectively. In addition, the oven-dried density also was increased by 12.20% and 53.66% from 0.41 g·cm⁻¹ to 0.46 g·cm⁻¹ and 0.63 g·cm⁻¹, respectively. The aforementioned results are in accordance with the increasing tendency of WPG. Treatment with IA induced a good bending strength (88.01 MPa, 17.69% increment) and compressive strength parallel to the grain (62.31 MPa, 40.18%). These results demonstrated that IA had a limited promotion for physical and mechanical properties, and it mainly played a grafting anchor role in the construction process. After introducing PS within the W_IA, the W_IA/PS exhibited significantly enhanced bending strength (143.56 MPa) and compressive strength parallel to the grain (96.88 MPa), which increased by 91.98% and 117.95%, respectively, compared with that of the W. Furthermore, all three cut sections of W_IA/PS hardness also increased (Figure 2 ). These results demonstrated that in situ polymerization of styrene monomer effectively contributed to the mechanical performance of the material.

Figure 2:

Hardness of natural wood (W), IA-activated wood (W_IA) and in situ IA-activated wood (W_IA/PS).

Polymer distribution studied with FE-SEM

FE-SEM was conducted to study the polymer distribution at the wood cell level after the in situ polymerization of the W_IA sample (Figure 3). The SEM image of W_IA confirmed that the wood morphology remains almost unchanged after the modification. This result was consistent with a similar WPG result (Table 1). The only visible change of W and W_IA was the more compact W_IA cell walls. This was attributed to the IA linkage effect pulling neighboring wood fibers together (Figure 3b). After the in situ polymerization of styrene, almost all of the vessels and tracheids were completely filled with PS (Figure 3c), indicating this polymerization process occurred substantially. The magnification image of W_IA/PS (Figure 3d) showed that the cell wall became very dense and some structures (such as ray) were covered by PS. There is a seamless connectivity at the interface between lumens and cell walls, which is crucial to prevent water invasion. From the aforementioned results, the hydrophobic PS was compactly distributed in cell walls and lumens, which can comprehensively enhance water repellence, leading to a durable dimensional stability.

Figure 3:

FE-SEM images of transverse section.

(a) Natural wood (W), (b) IA-activated wood (W_IA) and (c, d) in situ IA-activated wood (W_IA/PS).

FTIR analysis

In order to confirm the esterification and in situ polymerization, the W, W_IA and W_IA/PS were characterized by FTIR spectroscopy (spectra are shown in Figure 4). The IR spectrum for the W reveals the characteristic peaks from holocellulose and lignin. After modification with IA, the presence of carboxylic acid groups can be confirmed by the clear increased peak intensities corresponding to the carbonyl C=O vibration at 1724 cm⁻¹, the C–O stretching at 1156 cm⁻¹ and the alkene stretching at 1635 cm⁻¹ (Tammer 2004). For the W_IA/PS, the typical peaks of PS at 698 and 756, and 1601 cm⁻¹ appeared, which can be attributed to the out-of-plane C–H bending of the aromatic ring, and an aromatic ring breathing mode, respectively (Socrates 2004). Moreover, the intensity of the double bond has a clear decrease, indicating the styrene was successfully grafted to the W_IA (Table 2).

Figure 4:

FTIR spectra curves of natural wood (W), IA-activated wood (W_IA) and in situ IA-activated wood (W_IA/PS), and absorption band assignment (1724 cm⁻¹, C=O stretching vibrations; 1635 cm⁻¹, C=O stretching vibrations; 1601 cm⁻¹, C=O stretching vibrations; 1156 cm⁻¹, C=O stretching vibrations; 756 cm⁻¹, C=O stretching vibrations; 698 cm⁻¹, C=O stretching vibrations).

XRD analysis

Figure 5a presents the XRD patterns of the natural and modified wood samples. All of the wood samples showed diffraction peaks around 15.6, 22.5 and 35° 2θ indicating the (101), (002) and (040) crystal planes of the cellulose I structure (Cave 1997). Comparing the three curves, the position of the peaks did not change, indicating modification treatments do not destroy the crystalline structure of cellulose, which is important for the physical and mechanical properties of W_IA and W_IA/PS. The only difference among samples was the intensity change in the peaks, denoting changes in the wood crystallinity (Figure 5b). Compared with W (27.8% crystallinity), the W_IA (26.0% crystallinity) had a slight decrease, showing IA just plays a grafting anchor role. As expected, the W_IA/PS (15.05% crystallinity) had a significant decrease compared to W_IA, which is attributed to the in situ polymerization of styrene monomer in wood hierarchical structure, increasing the proportion of the amorphous region.

Figure 5:

XRD patterns of natural wood (W), IA-activated wood (W_IA) and in situ IA-activated wood (W_IA/PS) (a) and the corresponding crystallinity (b).

TGA analysis

The TGA and derivative thermogravimetric (DTG) curves for the W, W_IA and W_IA/PS are presented in Figure 6 . Wood is vulnerable to thermal decomposition and its three main components (cellulose, hemicellulose and lignin) have different thermal stabilities due to their different structures (Kim et al. 2004; Schirp and Su 2016; Qiao et al. 2019). Typically, hemicellulose is easy to be degraded, and its pyrolysis is focused at 220–315°C. The pyrolysis of cellulose mainly happened at 315–400°C, while that of lignin covered a whole temperature range (150–600°C) (Yang et al. 2007). It can be observed the thermal degradation of W_IA is faster than that of W between 200 and 330°C, which indicates IA has an acid hydrolysis for hemicellulose, making it easier to be pyrolyzed. For the W_IA/PS, the TG and DTG curves are shifted to higher temperature regions, indicating PS has a strong connection with wood components and it can effectively prevent wood from thermal degradation. PS has a good thermal stability and the hydrolysis section at 400–500°C is related to it (Liu et al. 2019). There is no doubt that modified wood using PS will obtain a good thermal stability. There is more residual weight (ash content) at 600°C for the W_IA/PS, also reflecting the full polymerization of styrene monomer in wood tissue.

Figure 6:

TGA (a) and DTG (b) curves of natural wood (W), IA-activated wood (W_IA) and in situ IA-activated wood (W_IA/PS).

Water-repellent property and leachability

The H values of the samples were measured to evaluate the water repellency of the W, W_IA and W_IA/PS upon water soaking for 10 days (Figure 7a). Within this time period, the W and W_IA showed almost same hygroscopic tendencies and they gained about 133% and 126% water, respectively. When styrene monomer underwent in situ polymerization within the W_IA, the H value of the W_IA/PS was reduced by more than half of W. It can be attributed to the polymerization of the hydrophobic styrene with the cell wall components and cell lumens’ bulk filling of PS, which occupied the flow path (vessels and tracheids) originally available for water and obstructed water uptake. In addition, after 10 days of washing, the leaching tests showed that the W_IA/PS has a very low leaching rate (3.51%) compared to the W (3.22%) (Figure 7b). This behavior represents PS has been firmly connected to the wood structure and there is no residual styrene monomer in W_IA/PS. Consequently, such W_IA/PS could be potentially utilized as a reliable engineering material for outdoor construction due to its superior durability in water.

Figure 7:

(a) Water uptake of natural wood (W) (A), IA-activated wood (W_IA) (B) and in situ IA-activated wood (W_IA/PS) (C). (b) Leachability of natural and modified wood samples.

Dimensional stability

The ASE was determined to evaluate the dimensional stability of the treated wood upon water soaking cycles (Figure 8). The sample with a lower S and higher ASE value can possess better dimensional stability. When the W was activated using IA, the S of the modified wood dropped by 25.97%, from 14.26% to 11.32%. According to Figure 8, the effect of styrene polymerization on S and ASE was significant. The S of W_IA/PS was 101.70% and 60.11% smaller than the W and W_IA, respectively. The incorporation of PS into the wood leads to a further increase in the ASE when compared to the W_IA: after 10 days in water, the ASE of W_IA is about 22.91%, while that of W_IA/PS can be up to 50.46%. The obtained ASEs were also in good agreement with the results of the FE-SEM, XRD and H measurements. Lumen filling can decrease the space available for water, block cell cavities (such as pits) and consequently delays, and avoids water entrance into the cell walls (Hadi and Massijaya 2018). Wood cell wall modification is an efficient way for improving dimensional stability (Kong et al. 2018). So, excellent dimensional stability of the W_IA/PS can be achieved through the reinforcement of the cell walls by the modification of IA and styrene and their polymerization in the wood cell lumens.

Figure 8:

S and ASE values of natural wood (W), IA-activated wood (W_IA) and in situ IA-activated wood (W_IA/PS).