Abstract

Materials

The materials used in this study were WSF and two types of RPP packaging bags, i.e., RPP-DB and RPP-WB. The RPP packaging bags were obtained from Jiulong Regenerated Resources Recycling Co. Ltd, Jiangsu, China. WS came from Youfang Town, Jiangsu Province, China, and was prepared by grinding and screening, using a LH-08B Speed Grinder (Xinchang Hongli, CNC Instruments Inc., Jiangsu, China) with a 60-mesh screen. Silane coupling agent (KH550) from the Yaohua Co. (Shanghai, China) was used to treat WSF. The formulation of the WS/RPP composite was 50 percent WSF and 50 percent RPP. A 50 percent WSF composite formulation was chosen because it was representative of commercially available lumber products.

Sample preparation

WSFs were first ovendried at 105°C for 6 hours to reduce moisture content level to less than 2 percent. A speed mixer (K600-3205, Braun Electric, Germany) was used to mix WSF thoroughly, while stirring an aqueous solution (ethanol:water = 9:1 by weight) including a silane coupling agent (2%, by weight, of WSF) that was sprayed into them at 25°C for 30 minutes. After that, the mixture of WSF was dried at 105°C for 12 hours. The WS/RPP composites were prepared by blending the WSF and PP at 175°C for 5 to 7 minutes on a two-roll mill mixer (X-160 Banbury, Chuangcheng Rubber and Plastic Machinery Co., Ltd., Wuxi, China), after which the melting mixture was laid on a compression molding machine (XLB-0 Vulcanizing Mechine, Shunli Rubber Machinery Co., Ltd., Huzhou, China) at 180°C and 12.5 MPa for 12 minutes. The prepared board dimension was 120 by 100 by 5 mm, and then cut into the required size. All specimens were then conditioned for 72 hours at a temperature of 23°C ± 2°C and a relative humidity of 50 ± 5 percent for later characterization.

Accelerated weathering testing

The samples were subjected to an artificial weathering by exposing to 340-nm fluorescent UV lamps in a QUV accelerated weathering tester (ZN-P Aging Chamber, Nuoweier Electronic Technology Co., Ltd, Suzhou, China) for different intervals of 240, 480, 720, 960, and 1,200 hours. The weathering cycle involved a continuous UV light irradiation of 8 hours (60°C) followed by a water condensation for 4 hours (50°C). The specimens were exposed to the artificial weathering only on one side for all times.

Characterization

Colorimetric analysis

The surface color of composites was measured by a chroma meter (Hanpu HP-200, China) according to the CIE L*a*b* color system. Four specimens were taken for each formulation and measured. For each specimens five different points were measured, and the average values were calculated automatically by the chroma meter. The color difference (ΔE) was determined with the procedure outlined in ASTM D2244‐15a (ASTM International 2015):

where ΔL*, Δa*, and Δb* are the total changes on L*, a*, and b* values during weathering, respectively. An increase in L* means the sample is lightening (i.e., a positive ΔL* for lightening and a negative ΔL* for darkening). A positive Δa* signifies a color shift toward red, and a negative Δa* signifies a color shift toward green. A positive Δb* signifies a shift toward yellow, and a negative Δb* signifies a shift toward blue.

FTIR spectra measurement

The surface chemical changes in composites during accelerated weathering were monitored by a Fourier transform infrared (FTIR) spectrometer (Nicolet iS 10 FT-IR, Thermo Fisher Scientific, USA). The FTIR spectra were obtained in the range 4,000 to 400 cm⁻¹ at a resolution of 4 cm⁻¹ and a minimum of 60 scans. A typical approach consists of the monitoring of one or more bands of interest in relation to a band that did not change during the weathering process. Because the peak at 2,920 cm⁻¹, which corresponds to asymmetric stretching vibrations of methylene (–CH2–) groups of PP, showed the least change during degradation, it was used as a reference (Zhang et al. 1997, Achilias et al. 2008). The carbonyl index was calculated using the following equation:

where I denotes the intensity. The peak intensity in the carbonyl region normalized using the intensity of peak at 2,920 cm⁻¹, which corresponds to asymmetric stretching vibrations of methylene groups. The lignin index was calculated using the following equation:

The peak intensity of aromatic skeletal vibration originating from lignin (1,508 cm⁻¹) normalized using the intensity of peak at 2,920 cm⁻¹, which corresponds to asymmetric stretching vibrations of methylene groups (Achilias et al. 2008).

Surface morphology

Surface morphology of composites during weathering is shown in Figure 2. Before exposure, the WS/RPP-DB and WS/RPP-WB composites showed a smooth and flat surface, exhibiting that WSF could be completely wrapped by PP matrix (Figs. 2a and 2e). After 240 hours of exposure (Fig. 2b), the bond between WSF and PP was destroyed at first, with cracks around the WSF edge extending and expanding to all around, leaving distinguishable unprotected WSF on the exposed surface. These cracks were considered to be related to the photo-oxidation of PP, which caused chain scission and molecular weight reduction, and would initiate and propagate cracks (Fabiyi et al. 2009). The composites produced from WSF swelled greatly when exposed to water during the condensation process because of its higher hygroscopicity than PP (Reddy et al. 2010). During 240 to 720 hours of exposure, the bared WSF became the main target of UV radiation and water, which made WSF swell when exposed to alternate wetting and drying cycles, and combined with PP bearing alternate stress, resulted in cracks accumulating. Because the swelling WSF created microcracks in the matrix and degraded WSF particles were washed away by condensation water, many smaller cracks were observed (Fig. 2c). Though long-time exposure (1,200 h) caused visible changes on the surface layer, including protrusion and debonding of WSF, WS/RPP-DB still showed a relatively flat surface (Fig. 3a), indicating that the degradation mainly happened on the surface and the overall composites largely remained integral and dense. So for WS/RPP-DB, the weathering degradation process was gradual from surface to interior.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Figures 2e through 2h present micrographs of weathered surfaces of WS/RPP-WB. After 240 hours of weathering, some tiny cracks and microparticles were observed on the surface of WS/RPP-WB (Fig. 2f). With increasing exposure time, these cracks and microparticles also increased. After 720 hours of exposure, the degraded RPP-WB was unable to bond WSF and formed many deep holes on the surface (Fig. 2g), facilitating UV light and water penetration deeper into deeper composite and further degradation (Stark and Matuana 2006). When the weathering test was carried out to 1,200 hours (Fig. 2g), the degraded WSF and PP were washed away by condensation water, which made these holes deeper and the whole composite became loose and porous. In Figure 3b, the matrix surrounding WSF had been hollowed out and a very deep groove on the surface could be observed, indicating that the internal structure of WS/RPP-WB had been destroyed. Therefore, the stress cannot be transferred effectively from fiber to plastic matrix and it could be the reason that WS/PP-WB showed much lower impact strength than WS/RPP-DB.

Color change

As can be seen in Figure 4 the curves of ΔL* substantially kept in step with ΔE, indicating parameters of a* and b* values had a low effect on the total color change. When the weathering test began, WS/RPP-DB and WS/RPP-WB were photobleached rapidly, then slightly slowed up to 480 hours. This phenomenon could be attributed to the loss of methoxyl content of lignin in WSF, photodissociation of carbon–carbon bonds, and formation of carbonyl-based chromphoric groups (Michel and Billington 2012, Peng et al. 2014). Lignin accounted for 80 to 95 percent of light absorption, which was suggested to have made a significant effect on the discoloration (Homkhiew et al. 2014). It should be noted that the discoloration process of WS/RPP-WB was weaker than that of WS/RPP-DB, which might be because of the residual light stabilizer in PP woven bags. Usually, to prevent the performance decline of PP woven bags in the process of long-term outdoor use, some UV stabilizers are added by manufacturers (Karian 2003), which could enhance the light resistance of WS/RPP-WB to some extent. The changes in ΔE values were found to increase quickly again until 960 hours. The ΔL* of WS/RPP-DB and WS/RPP-WB increased up to 960 hours, which were, respectively, 1.7 and 2.2 times the 240 hours. Besides the effect of UV light, the presence of condensation water enhanced the rate of photodegradation of the composites. Absorption of condensation water initiated the cell wall swell of WSF to create microcracks in the matrix and facilitate UV light farther into WS/RPP (Athijayamani et al. 2009, Assarar et al. 2011). Also during the condensation process, the loose WSF particles, degraded lignin, and hemicellulose at the surface were washed away by water, which was in line with the results of the TGA discussed below in “Thermal properties.” The main component of WS/RPP surface was cellulose, which was whitish and comparatively stable against UV radiation (Kiguchi et al. 2007). These celluloses formed a gray fibrous layer on the surface of weathered samples and increased ΔL* (Beg and Pickering 2008). Stark and Matuana (2003) reported that whitening of WPCs during weathering originated mainly by photobleaching the wood fiber. At 1,200 hours, ΔL* and ΔE of WS/RPP-DB and WS/RPP-WB both declined. This was related to the surface cracks of the composite gradually extending to the interior as weathering time increased, which made the inner WSF exposed to the PP matrix.

View Figure

• Download as Powerpoint
• Download Figure

Surface chemistry

The FTIR spectra of WS/RPP-DB and WS/RPP-WB before and after weathering are shown in Figure 5. Calculated values of the carbonyl index and lignin index are shown in Figure 6. As shown in Figure 6A, the carbonyl index of WS/RPP-WB was higher than that of WS/RPP-DB before weathering, resulting from the outdoor working conditions of RPP-WB. The presence of carbonyl groups, hydroperoxide groups, and catalyst residues could be introduced during storage, which may initiate photochemical reactions (Matuana et al. 2011). At up to 480 hours of weathering, the carbonyl index of WS/RPP-DB and WS/RPP-WB quickly increased to 8.44 and 14.5 percent, respectively, compared with that of composites before weathering, indicating that PP was photodegraded under UV exposure, resulting in the main chain scission and accretion of carbonyl products such as ketone, ester, and carboxylic acid (Rajakumar et al. 2009). Meanwhile, WSF exposed from matrix and all main polymeric components of fibers (lignin, cellulose, hemicellulose, and extractives) underwent photodegradation, in which quite a few products containing the carbonyl group were generated (Stark and Matuana 2004a, Butylina et al. 2012b). The WS/RPP-DB experienced a small drop in carbonyl index during 720 to 960 hours, which was related to the degraded WSF and PP on the surface washed away by condensation water as shown in Figure 2c. After 1,200 hours of weathering, the carbonyl index of WS/RPP-DB composite rose slightly. The increase indicated that the surface cracks of composites gradually extended to the interior, as shown in Figure 2, which initiated further photodegradation of composites. The carbonyl index of WS/RPP-WB was higher than that of WS/RPP-DB. It may relate to the impurity mingled in with RPP-WB during use, which could accelerate carbonyl group formation (Wypych 2003).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

The lignin indices of WS/RPP-DB and WS/RPP-WB dropped during weathering (Fig. 6B). This probably arose from lignin photodegradation because lignin contains many UV light-absorbing groups that are very sensitive to UV light and that could form aromatic and other free radicals (Peng et al. 2014). Phenolic hydroxyl groups were changed by UV radiation rapidly to and from phenolic radicals, which transformed into quinone compounds and formation of carbonyl-based chromophoric groups (Fabiyi et al. 2008). In addition, the condensation water scoured away the WSF on the surface of the composite as shown in Figure 3, which aggravated the loss of lignin. The increase in the carbonyl index and the decrease in lignin index for WS/RPP implied that surface oxidation intensified upon weathering, and also suggested that the color change was closely related to the formation of carbonyl groups and degradation of lignin.

Thermal properties

TGA was carried out for WS/RPP-DB and WS/RPP-WB after weathering and is presented in Figure 7. The temperatures of different stages of decomposition, onset decomposition temperature (T_onset), temperature of maximum reaction rate (T_max), weight loss at T_max (WL_Tmax), and residual mass at 700°C of composites are shown in Table 1. The T_onset of WS/RPP-DB and WS/RPP-WB before weathering were 270.8°C and 277.8°C, respectively, compared with oriented PP and PP-WB (0 h) that were 433.9°C and 431.9°C separately, which indicated that the addition of WSF weakened the thermal stability of materials.

View Figure

• Download as Powerpoint
• Download Figure

Table 1 Decomposition data of WS/RPP-DB and WS/RPP-WB.^a

View Fullscreen

For WS/RPP-DB, there were three stages of weight loss: 200°C to 330°C, 330°C to 390°C, and 390°C to 500°C. The first two stages of weight loss were mainly from the thermal decomposition of WSF and the third stage was mainly from the thermal decomposition of PP matrix. The first stage of thermal decomposition was mainly attributed to hemicellulose (Alvarez and Vázquez 2004). The temperature at the first weight loss stage decreased from 307.3°C (0 h) to 303.4°C (1,200 h), which was attributed to the glycosidic bond between the monosaccharide groups being damaged by UV, and hemicellulose being degraded into small molecules, resulting in a decline of thermal stability of hemicellulose during weathering (Hosseinaei et al. 2012). It was interesting that the hemicellulose decomposition stage in WS/RPP-DB was more obvious than that in WSF (1,200 h), which was owing to the good property of WSF matrix interfacial bonding, leaving the RPP-DB matrix with better-protected WSF and reducing the external environment effect on WSF. The second weight loss stage of WS/RPP-DB was mainly caused by cellulose and lignin in WSF (Abu-Sharkh and Hamid 2004). The temperature at the second weight loss stage shifted to lower temperatures with increasing weathering time, indicating the reduction of molecular weight of cellulose and lignin because of the crystalline part of cellulose being damaged by UV radiation and the sensitive UV absorption functional groups in lignin (Peng et al. 2014). Meanwhile, many active groups like hydrogen peroxide were formed during PP degradation, which would have aggravated the photo-oxidation process of WSF, leading to decreased thermal stability (Rajakumar et al. 2009).

There were mainly two weight loss stages for WS/RPP-WB: 220°C to 380°C for WSF thermal decomposition and 390°C to 500°C for PP thermal decomposition. The hemicellulose decomposition stage in WS/RPP-WB was not as obvious as that in WS/RPP-DB, which suggested that the decomposition rate of RPP-WB was faster than that of RPP-DB; thus degraded RPP-WB could not wrap and protect WSF, leading to surface WSF exposed to UV light directly. In the thermal decomposition stage of cellulose and lignin, the T_pk2 (peak temperature of second weight loss stage) of WS/RPP-WB shifted from 329.9°C (0 h) to 326.9°C (1,200 h), which was very close to the T_max of WSF after 1,200 hours of weathering (325°C). It could be inferred that RPP-WB did not wrap WSF completely, so more and more WSF were exposed to the WS/RPP-WB surface with increasing weathering time, which enhanced the surface ratio of WSF and PP matrix.

In the thermal decomposition of PP matrix, the TGA transitions for WS/RPP-DB and WS/RPP-WB both shifted to lower temperatures after 1,200 hours of weathering (Table 1). The decrease in thermal stability of composites during weathering results from the PP chain scission along with degradation of both WSF and the WSF–matrix interfacial bonding (Abu-Sharkh and Hamid 2004).

Mechanical properties

The effect of accelerated weathering on flexural strength and modulus of WS/RPP-DB and WS/RPP-WB are illustrated in Figure 8. In practical applications, flexural properties were critical for the performance of WPCs (Lopez et al. 2006). It can be clearly seen in Figure 8A that flexural strength of WS/RPP-DB decreased quickly up to 720 hours, maintained steady during 720 to 960 hours, and then markedly decreased at 1,200 hours. For WS/RPP-WB, flexural strength slightly decreased up to 480 hours and then sharply decreased after 720 hours. Stark and Matuana (2003) reported that the decrease of flexural strength was related to the occurrence and development of surface cracks on composites. After 240 hours of exposure, microcracks on the WS/RPP-DB surface were already distinguishable (Fig. 2b). When they occurred from flexural forces, these cracks could cause stress concentration and reduce the efficiency of stress transfer from WSF to a plastic matrix, leading to strength decrease (Assarar et al. 2011). However, the flexural strength decrease of WS/RPP-WB was not obvious at 240 hours, which was attributed to a relatively flat surface as shown in Figure 2f. The flexural strengths of WS/RPP-DB and WS/RPP-WB deceased by 55.6 and 67.5 percent, respectively, after 1,200 hours of exposure. The trends of changes in flexural modulus of WS/RPP-DB and WS/RPP-WB were similar to the flexural strength, clearly decreasing after 240 hours. The flexural modulus of WS/RPP-DB and of WS/RPP-WB deceased by 64.9 and 64.3 percent, respectively, at 1,200 hours. The reduction in flexural modulus observed was partly a result of the negative effects of moisture penetration during exposure (Hon and Shiraishi 2000). More cracks and deep holes were observed on surfaces of WS/RPP-WB (Fig. 3b), facilitating water penetration into the internal composite, which resulted in a smaller flexural modulus than WS/RPP-DB. It was reported that the crystallinity of the composites initially increases during exposure to UV light and water, and then decreases with continued exposure (Stark and Matuana 2006). When PP is exposed to UV light, chain scissions occur. The crystalline regions are affected and crystallinity decreases when chain scission continues further (Homkhiew et al. 2014). It was expected that after 1,200 hours of exposure, the crystallinity of PP would decrease, decreasing flexural properties.

View Figure

• Download as Powerpoint
• Download Figure

Impact strengths of WS/RPP-DB and WS/RPP-WB are illustrated in Figure 8B. After 1,200 hours of weathering, the impact strengths of WS/RPP-DB and WS/RPP-WB decreased by 22.4 and 27.4 percent, respectively. Decline in impact strength was due to poor stress transfer resulting from the degradation of the WSF–matrix interface (Stark and Matuana 2004b). The ways of energy absorption of natural fiber–reinforced composites mainly were fiber fracture, fiber pull-out, and matrix fracture. Fiber fracture could absorb the highest energy, whereas matrix fracture was the lowest (George et al. 2001). As shown in Figure 9a, a trace of WSF pull-out was observed, which suggested that the way of energy absorption of WS/RPP-DB after 1,200 hours of weathering was fiber fracture, although from the fracture surface morphology of WS/RPP-WB (Fig. 9b) the whole composite appeared to be loose and porous. As the severe degradation of RPP-WB led to poor interfacial adhesion and low efficiency of stress transfer from WSF to matrix, the main way of energy absorption of WS/RPP-WB was matrix fracture; for that reason, the impact strength of WS/RPP-WB after 1,200 hours of weathering was much lower than that of WS/RPP-DB.

View Figure

• Download as Powerpoint
• Download Figure