WATER ABSORPTION, THICKNESS SWELLING AND MECHANICAL PROPERTIES OF CEMENT BONDED WOOD COMPOSITE TREATED WITH WATER REPELLENT

In this study, the purpose was to improve outdoor performance of cement bonded wood composite due to their biodegradation and sensitivity to moisture especially in warm and humid climates. Cement bonded wood composites were treated with different concentrations (10 %, 25 %, 50 %, 75 % and 100 %) of water repe - llent. Water repellent used was an organo-silicon based, nano-sized, eco-friendly, water-based agents. Dipping and pressure systems were applied for composite treatment. Water absorption, thickness swelling, accelerated weathering, color changes and mechanical properties after accelerated weathering were determined for treated and untreated cement-bonded composites. Results showed that treatment of composites with water repellent provided a transparent layer on composite surface. Thus, lower water absorption and thickness swelling results in the beginning of immersion in water. Treated and untreated composites were exposed to an accelerated weathering test for 350 h. Their mechanical strength including modulus of rupture, modulus of elasticity and internal bonding properties were decreased after 350 h of weathering. However, overall results after weathe-ring test showed that all panels’ mechanical properties provided minimum modulus of rupture, modulus of elasticity and internal bonding requirements of the EN standards.


INTRODUCTION
Cement-bonded wood composite materials have been studied for more than one hundred years but they were commercially utilized since 1930's (Frybort et al. 2008). Cement-bonded wood composites (CBWCs) are made of wood particles mixed with Portland cement (Okino et al. 2004, Marzuki et al. 2011. Utilization of different fiber sources in making cement-bonded composite material became an alternative material due to concerns related to policies of limiting asbestos (Moslemi 1999, Karade 2010. Large amount of different lignocellulosic materials can be used for composite production. Some of them are generated around the world during various human activities including agro-forestry wastes like rice husk and straw, sunflower stems, wheat-straw, bagasse, oil palm strands, hazel nuts and saw dust (Widyorini et al. 2005, Thygesen et al. 2005, Yel 2022). Compatibility of wood and cement is the main problem in cement-bonded wood composites due to inhibition effect of wood on cement settings. The inhibitory substances in wood such as hemicelluloses, starches, sugars, phenols and hydroxylated carboxylic acids cause hydration effects when cement is mixed with wood (Papadopoulos 2008, Frybort et al. 2008, Tittelein et al. 2012. Researchers reported that different wood treatments such as extraction, alkaline hydrolysis and retention of polysaccharides can be applied to minimize inhibition problems. Inhibitory water-soluble compounds in wood can be removed by extraction method while hemicelluloses and sugars can be degraded into non-inhibitory substances by alkaline hydrolysis. In the retention of polysaccarides, inhibitory substances are not released from the wood surface due to formation of thin coating layer on the wood surface (Quiroga et al. 2016).
CBWCs are widely preferred in building constructions for both interior and exterior applications due to their good properties on sound attenuation, fire resistance, thermal insulation, and structural performance (Huang andCooper 2000, Na et al. 2014). In addition, cement-bonded wood composites possesses both workability with conventional wood-working tools like sawn, drilled, nailed, sanded, glued, and screwed and durability like masonry construction material (Marzuki et al. 2011). However, using CBWC in exterior condition especially in warm and humid climates is limited due to biodegradation and sensitivity to moisture of wood in composites (Wei andTomita 2001, Kirkpatrick andBarnes 2006). Moreover, the wood composites are susceptible to UV radiation corruption. Wood components such as hemicellulose, lignin and cellulose are the main wood chemical elements affected by the photodegradation; lignin is the most susceptible to weathering. Photodegradation of lignin leads to the staining of wood. The UV radiation causes the formation of free radicals. As a result of this, begin by photochemical reactions in wood. Photochemical reactions induce the degradation of lignin and photooxidation of cellulose and hemicelluloses. This also leads to discoloration of wood and influence negatively on the wood's physical and mechanical properties (Temiz et al. 2007, Matuana et al. 2011, Durmaz et al. 2022).
Treatment of wood-based composites is not easy due to many types of wood-based composites such as particleboard, cement-bonded wood composite, medium density fiber board, wood-plastic composites etc. and manufacturing processes. Wood-based composites can be protected by various methods including pre-treatment of particles, in-process treatment during manufacturing process, post-treatment after manufacturing of composites (Kirkpatrick andBarnes 2006, Taşçıoğlu 2013). The most practical way to protect composites could be post-treatment after manufacturing the composites by dipping, spraying, brushing or vacuum-pressure treatment (Taşçıoğlu 2013). Depending on the desired properties of cement-bonded wood composites, physical and mechanical properties can be improved with several factors such as cement-wood ratio, particle size and geometry, orientation of particles, treatment of particles or treatment composites with water repellent among other. Moisture absorption and dimensional stability could be improved by impregnating water soluble polymers into composite structure after manufacturing the composites. Therefore, improving outdoor performance of CBWC with different treatment methods provides wider potential application of CB-WCs for replacing traditional building materials for interior and exterior use.
The objective of the research is to improve the outdoor performance of cement-bonded wood composites treated with water repellent. CBWCs were treated with different concentrations of organo-silicon based, nano-sized chemical and exposed accelerated weathering test. The effects of water repellent on water absorption, dimensional stability and mechanical properties of CBWCs after QUV test were determined.

Composite treatment process
CBWCs used in this study were commercially obtained from TEPE Betopan Co. Inc. in Ankara, Turkey. CBWCs were treated with different concentrations (10 %, 25 %, 50 %, 75 % and 100 %) of water repellent to improve outdoor performance of composites. Water repellent used was an organo-silicon based, nano-sized, eco-friendly, water-based agents. It was obtained from ARDChem in Türkiye. Since the size of water repellent less than 3 nm -6 nm, it penetrated deep into the material pores 3 nm -5 mm. Therefore, the building material becomes stronger and highly water repellent (Köse et al. 2014). Two different treatment processes, dipping and vacuum-pressure systems were applied to treat with water repellent. The first group samples were dipped into solutions for 10 min. Pressure system was conducted in a small-scale impregnation container using a vacuum of 0,09 MPa for 30 min followed by 0,6 MPa pressure for 30 min. The retentions for each treatment solution were calculated according to ASTM D 1413ASTM D (2007 standard by using the Equation 1: (1) Where; G Weight of treating solution absorbed by composite (kg), C is the concentration (%), V is the sample volume (m 3 ).

Water absorption and thickness swelling tests
Water absorption (WA) and thickness swelling (TS) tests were performed according to CEN EN 317 (1993). Fifteen replicates of each group with the dimensions of 50 mm (length) x 50 mm (width) x 10 mm (thickness) were conditioned in climate room (20 ºC ± 2 ℃ temperature and 65 % ± 5 % RH) until they reach constant weigh. Treated and untreated (control) samples were placed into distilled water test tanks and measured water absorption and thickness swelling after 30 min, 2 h, and 24 h. The WA and TS values were calculated according to Equation 2 and Equation 3 after each period.
Where; W 1 and W 2 are the weight of the boards before and after test, T 2 is the thickness of boards at any given time during water soaked condition and T 1 is the initial thickness of the boards.

Accelerated weathering test
The weathering test was performed in a QUV accelerated weathering tester. UVB-313 type lamp was used in QUV. The reason for choosing UV-B light is that the energy level in UV-B zone was enough to break the chemical bonds of carbon-nitrogen in polymer carbon-carbon, nitrogen-hydrogen, carbon-oxygen, carbon-hydrogen compounds. Composites, pressure treated with 25 % and 50 % of concentration water repellent were exposed to QUV due to showed better dimensional stability. Weathering test cycle comprised exposure to 340 nm UVB-313 light irradiation for 4 h followed by condensation for 4 h. The average irradiation was set about 0,71 W/m 2 at 340 nm with temperature of 60 ℃ while condensation temperature was set to 50 °C. Seven replicates of each treatment were exposed for 50 h, 100 h, 200 h and 350 h to QUV and determined color change according to ISO 7724-1 (1984).

Color measurement
The surface color of composites was determined according to ISO 7724-1 (1984) standard using a Minolta model color measurement device. CIELab (Commission International de i'Eclairage) system is characterized by three parameters, L*, a* and b*. L* axis represents light stability, a* and b* are the chromatographic coordinates, (+a* for red, -a* for green, +b* for yellow, -b* for blue). .
Where; ∆L*, ∆a*, ∆b* are the changes between the initial (i) and in different periods (f). A low ∆E* corresponds to a low color change or a stable color.

Mechanical properties
Modulus of rupture (MOR), modulus of elasticity (MOE) and internal bonding (IB) of the CBWCs, conditioned at 20 °C and 65 % RH were performed according to CEN EN 310 (1993) and CEN EN 319 (1993) standards, respectively. Rectangular 250 mm (length) x 50 mm (width) x 10 mm (thickness) samples were used to determine three-point bending measurement of MOR and MOE. The flexural properties were detected using a Zwick/Roell Z010 Universal Testing system equipped with a load cell capacity of 10 kN. Specimens of 50 mm (length) x 50 mm (width) x 10 mm (thickness) were used for IB measurement. The samples were tested on Mohr+Federhaff+Losenhause universal testing machine equipped with a load cell capacity of 10 kN. Twenty replicates were tested for each composite formulation. The result of CBWC was analyzed with ANOVA test using SPSS 21.0 software (SPSS 2020). The significance (p<0,05) between the treatments was compared with DUNCAN homogeneity groups.

Water absorption and thickness swelling ratios
Water absorption and thickness swelling of composites are given in Table 1 with their retention values. According to results, water absorption values of control groups were determined to an increase from 6,22 % after 30 min. to 14,29 % after 24 h of exposure in water. All the samples treated with water repellent significantly reduced water uptake in the beginning of immersion time. Increasing water repellent concentration in pressure system generally decreases water absorption values. After 24 h exposure in water, there were differences between test and control (untreated) groups. Samples pressure treated with 75 % of water repellent showed the lowest water absorption value after 30 min. exposure in water while all water repellent treatment significantly reduced water absorption in the beginning of water exposure. Samples there were treated with the pressure system showed lower water absorption than dipping process due to higher retention values. Pressure treatment system showed lower water absorption than dipping process due to higher retention values. Several researchers have reported that organo-silicon based treatment improved water absorption of wood by means of water repellency system (Köse et al. 2014, Glohamiyan 2010. Regarding the thickness swelling, maximum swelling of control (untreated) composites after 24 h in water was 0,91 %. This value is lower than resin-bonded particleboards subjected to similar conditions due to "encased" nature of wood in the cement, restricting wood from expanding volumetrically (Jorge et al. 2004, Yel andUrun 2022). The moisture absorption in composites is mainly due to the presence of lumens, pores, and hydrogen bonding sites in the wood fiber along with voids (English and Falk 1996). Treatment with water repellent provided better dimensional stability of composites than that of untreated (control) groups. Some groups treated with water repellent showed minus values due to creating a transparent coating layer of test chemical on the composite surface. Depending on the exposure time in water, this water repellent layer on the composite surface was washed out with water and samples reached their initial thickness. The lowest thickness swelling was determined on the composites pressure treated with 75 % of WR (-1,14 %). Decreasing WR amount in composites resulted in lower thickness swelling. Table 1: Water absorption and thickness swelling of the CBWC treated with water repellent.

The effect of accelerated weathering test on the color change of CBWC
Color changes of the panel groups after 50 h, 100 h, 200 h and 350 h of accelerated weathering are given in Table 2. According to CIELab coordinates, positive Δa* values show red in UV process, on the other hand, negative Δa* values indicates green. Again in Δb* values indicate yellow for positive and blue for negative. Positive ΔL* values represent getting white, negative ΔL* values represent getting grey.
After the 350-h process, the highest color change was observed at the panel group impregnated with 25 % of WR (ΔE: 14,46) while the lowest color change was determined at the group treated with 50 % of WR among the treatment groups. The reason for higher color change for WR groups than control groups could be due to destroying and washing out the coating layer on composite surface, formed by WR treatment. Water absorption results was also showed that this surface layer protected composites against water absorption but with increasing immersion time, composites treated with WR absorbed water, indicates destroying this layer.

The ffects of accelerated weathering test on mechanic properties of CBWC
Treated and untreated wood-cement composites were exposed 350 h to an accelerated weathering test. Mechanical properties including modulus of rupture (MOR), modulus of elasticity (MOE) and internal bonding (IB) of unexposed and exposed to QUV were determined. The results of MOR, MOE, and IB of experimental boards including homogeneity group (Duncan test) values are given in Table 3.
According to results, all wood-cement composites tested in this study provide the minimum MOR and MOE requirement of the CEN EN 634-2 (2007) standard. In addition, the minimum requirements of IB strength in the CEN EN 634-2 (2007) standards are 0,5 MPa for general usage and 0,3 MPa after the cyclic test. Therefore, all types of board met the minimum IB requirement of the EN standards. Changes on the mechanical properties after QUV test are shown in Figure 1.  As expected that, mechanical properties of wood-cement composites decreased after 350 h weathering test. This can be attributed that the weathering test can lead to surface damage due to forming the micro-cracking, resulted in reduction on mechanical properties (Page andPage 2007, Hung at al. 2017). The lowest mechanical changes were obtained from the composites treated with 25 % of WR. Higher concentration of WR increased the mechanical properties losses. This may be related with some negative effect of cement in composite structure and WR.
Strength properties of wood are very dependent on the moisture content of the cell wall. Because of increasing the rate of moisture barrier coated with a population that formed between water and wood panels as a result of decreases. For this reason, wood materials become more fragile and brittle (Rowell 1984, Taşçıoğlu et al. 2016, Shang et al. 2012. Impregnation of wood decreased flexible properties and shock resistance (Shang et al. 2012. When the examination of effect the mechanical properties of accelerated weathering, it was determined that a reduced amount of MOR, MOE and IB resistance compared to Duncan test results. The natural or accelerated weathering of composite materials can change their mechanical and strength properties (Butylina et al. 2012). Güntekin and Şahin (2009) defined that significant reduction in mechanical properties with the accelerated weathering test in fiber cement composites. As stated above, photodegradation of lignin adversely affected wood physical and mechanical properties. In addition, the physical and mechanical properties of cement-based material affected the environmental conditions greatly (Kockal and Turker 2007). Wetting-drying cycles had negative effects on strength properties and causes durability problems at concrete.

CONCLUSIONS
Commercially obtained cement-bonded wood composites were treated with water repellent in order to improve outdoor performance. Two different treatment processes (dipping and pressure treatment for water repellent) were applied as post-treatment after manufacturing the composites. Main findings are summarized as follow: Water repellent treatment significantly reduced water absorption in the beginning of water absorption test due to forming a transparent coating layer on the composite surface. However, increasing exposure time decreased water absorption differences between treated and untreated (control) composites.
Treatment of composites with water repellent provided better dimensional stability of composites than that of untreated (control) groups.
Mechanical properties of wood-cement composites decreased after 350 h weathering test. The lowest mechanical changes were obtained from the composites treated with 25 % of WR. However, the highest color change was observed at the panel group impregnated with 25 % of WR after 350 h weathering test.