**3.1. Long-term water absorption and thickness swelling**

Theoretically, the shape of the sorption curve plotted with experimental values is represented by the Fick's equation as log (*Mt* /*Mm*) = log(*k*) + *n*log(*t*), where *Mt* is the water absorbed at time *t*, *Mm* is the maximum water absorbed at saturation point and *k* and *n* are the diffusion kinetic parameters. The *n* values determined the diffusion behaviour cases: case I (*n* = 0.5) is Fickian diffusion, case II (*n* ≥ 1) is relaxation controlled, and case III (0.5 < *n* < 1). The *k* values represent the interaction between the material and moisture. Both coefficients of *n* and *k* are determined from the slope and intercept of log plot of *Mt* /*Mm* versus *t* which can be drawn from experimental data, respectively [1]. As shown in **Table 1**, the value of *n* is close to 0.5 which confirmed that all the investigated biocomposites exhibited Fickian diffusion behaviour. The *k* value was found to increase with the increasing wt% of RHB in the biocomposites irrespective of the TPB matrix types, in which a higher value of *k* indicates that the biocomposite attained the saturation point of WA in a shorter period of time. This was due to the increased hydroxyl and carbonyl groups as the RHB content increased [2].


**Table 1.** Diffusion kinetic parameters, coefficients and permeability of biocomposites.

The diffusion coefficient *D* represents the ability of water molecules to penetrate inside the composites, which can be computed with Eq. 3, in the case where the *Mt* value is < 60 % of the *Mm* value.

$$D = \pi \left[\frac{\theta}{4M\_m}\right] \tag{3}$$

where *D* is the diffusion coefficient, *θ* is the slope of the linear portion of *Mt* against ℎ / and *h* is the height (thickness) of composite panel [3]. In **Table 1**, the *Mm* and *D* increase with the RHB content which indicates that the biocomposites containing the higher wt% of natural fibre tended to absorb more water. This phenomenon could be explained by the components of RHB (cellulose, hemicellulose and lignin) that mostly contain polar hydroxyl groups which tend to combine with water molecules by forming hydrogen bond [4]. The difference in *D* between UTPB and CTPB matrixes was more predominant in the biocomposites at higher wt% of fibre. Especially at 80 wt% RHB, a remarkable decrease in *Mm* and *D* was observed for the biocomposites based on CTPB matrix as compared to UTPB ones. This confirmed the reduction of water absorption. This was probably because the free hydroxyl and carboxyl groups available in hydrophilic rPET play the role in absorbing water as there are lack of chemical interaction and bonding between polar rPET and nonpolar rHDPE chains in the case of without compatibilizer in the UTPB matrix [5]. Meanwhile, the presence of E-GMA compatibilizer tended to bind the rHDPE and rPET together, hence decreasing the hydrophilicity of CTPB matrix as well as reducing the WA.

The theoretical WA can be estimated using Fick's second law of diffusion. For the initial part of the curve where *Mt* /*Mm* < 0.6 [6], it can be predicted using Eq. 4. For the second half-sorption curve where *Mt* /*Mm*  > 0.6 [6], an approximation has been proposed with Eq. 5 [7]:

$$\frac{M\_r}{M\_m} = \frac{4}{h} \sqrt{\frac{Dt}{\pi}}\tag{4}$$

$$\frac{M\_{\rm \cdot}}{M\_{\rm \cdot \cdot \cdot \cdot}} = 1 - \exp\left[-7.3 \left(\frac{Dt}{h^2}\right)^{0.75}\right] \tag{5}$$

**Figures 1** and **2** depict the long-term WA of TPB/RHB biocomposites measured after a certain period of immersion in distilled water, respectively. The experimental data, as represented by the various symbols, show that the water absorbed by the biocomposites increased sharply with time and then gradually slowed until it reached an equilibrium state (*Mm*). The solid curves display the estimation of theoretical WA behaviour following Fick's second law of diffusion (Eqs. 4 and 5). All the curves in both **Figures 1** and **2** were found to follow the Fickiantype behaviour, as conformed to the coefficient of *n* which was close to 0.5 (**Table 1**). Thus, it

can be concluded that the experimental results will fit the Fickian mode of diffusion, especially in the initial stage of diffusion. This result is in agreement with the previous studies on polymer/natural fibre composites [6, 8]. **Figures 3** and **4** present the thickness swelling (TS) of immersed UTPB and CTPB biocomposites filled with RHB. It can be observed that TS was positively related to the WA where TS increased rapidly in the initial stage for all RHB contents and then remained constant. The higher the RHB loading, the more hydrogen bonding was created in the fibre cell wall by the adsorbed water, and thus, the higher TS of biocomposites was obtained.

The diffusion coefficient *D* represents the ability of water molecules to penetrate inside the

4 *<sup>m</sup>*

*h* is the height (thickness) of composite panel [3]. In **Table 1**, the *Mm* and *D* increase with the RHB content which indicates that the biocomposites containing the higher wt% of natural fibre tended to absorb more water. This phenomenon could be explained by the components of RHB (cellulose, hemicellulose and lignin) that mostly contain polar hydroxyl groups which tend to combine with water molecules by forming hydrogen bond [4]. The difference in *D* between UTPB and CTPB matrixes was more predominant in the biocomposites at higher wt% of fibre. Especially at 80 wt% RHB, a remarkable decrease in *Mm* and *D* was observed for the biocomposites based on CTPB matrix as compared to UTPB ones. This confirmed the reduction of water absorption. This was probably because the free hydroxyl and carboxyl groups available in hydrophilic rPET play the role in absorbing water as there are lack of chemical interaction and bonding between polar rPET and nonpolar rHDPE chains in the case of without compatibilizer in the UTPB matrix [5]. Meanwhile, the presence of E-GMA compatibilizer tended to bind the rHDPE and rPET together, hence decreasing the hydrophilicity of CTPB matrix as

The theoretical WA can be estimated using Fick's second law of diffusion. For the initial part

4 . *<sup>t</sup>*

*M Dt M h*

*m*

*m*

. 1 exp 7.3 *<sup>t</sup>*

é ù æ ö =- -ê ú ç ÷ ê ú è ø ë û

**Figures 1** and **2** depict the long-term WA of TPB/RHB biocomposites measured after a certain period of immersion in distilled water, respectively. The experimental data, as represented by the various symbols, show that the water absorbed by the biocomposites increased sharply with time and then gradually slowed until it reached an equilibrium state (*Mm*). The solid curves display the estimation of theoretical WA behaviour following Fick's second law of diffusion (Eqs. 4 and 5). All the curves in both **Figures 1** and **2** were found to follow the Fickiantype behaviour, as conformed to the coefficient of *n* which was close to 0.5 (**Table 1**). Thus, it

*M D t M h*

/*Mm*  > 0.6 [6], an approximation has been proposed with Eq. 5 [7]:

p

/*Mm* < 0.6 [6], it can be predicted using Eq. 4. For the second half-sorption

0.75

2

<sup>=</sup> (4)

*M* q

é ù <sup>=</sup> ê ú

value is < 60 % of the

against ℎ / and

(5)

ë û (3)

composites, which can be computed with Eq. 3, in the case where the *Mt*

*D*

where *D* is the diffusion coefficient, *θ* is the slope of the linear portion of *Mt*

p

*Mm* value.

30 Composites from Renewable and Sustainable Materials

well as reducing the WA.

of the curve where *Mt*

curve where *Mt*

**Figure 1.** Water absorption of immersed UTPB/RHB biocomposites after a certain period.

**Figure 2.** Water absorption of immersed CTPB/RHB biocomposites after a certain period.

**Figure 3.** Thickness swelling of immersed UTPB/RHB biocomposites after a certain period.

**Figure 4.** Thickness swelling of immersed CTPB/RHB biocomposites after a certain period.

Theoretically, the swelling behaviour can be computed by the model suggested by Shi and Gardner [9], which gives Eq. 6 after being rearranged and taking the natural logarithm:

$$\ln\left(\frac{100T\_{\circ}}{\text{TS}(t) + 100} - T\_{\text{o}}\right) = \ln(T\_{\circ} - T\_{\text{o}}) - K\_{\text{SR}}t \tag{6}$$

where TS(*t*) is the TS at time *t*, *T*∞ and *T*0 are the equilibrium and initial thickness of sample, respectively, and *KSR* is the intrinsic relative swelling rate (a constant). As shown in **Table 1**, the *T∞* and *KSR* values increased with the RHB wt%, which indicates the higher swelling rate that the composite required a shorter time to reach the equilibrium TS [10]. The CTPB/RHB biocomposites exhibited lower *KSR* values than that of UTPB/RHB biocomposites. This was probably attributed to better compatibility of TPB matrix which led to better fibre-matrix adhesion, and thus the water was difficult to access in the cellulose [2]. From **Figures 3** and **4**, the experimental results fitted well with the predicted TS from swelling model (solid curve).

#### **3.2. Tensile properties**

**Figure 3.** Thickness swelling of immersed UTPB/RHB biocomposites after a certain period.

32 Composites from Renewable and Sustainable Materials

**Figure 4.** Thickness swelling of immersed CTPB/RHB biocomposites after a certain period.

æ ö

*t* ¥

Theoretically, the swelling behaviour can be computed by the model suggested by Shi and Gardner [9], which gives Eq. 6 after being rearranged and taking the natural logarithm:

0 0

¥

è ø <sup>+</sup> (6)

*<sup>T</sup> T T T Kt*

<sup>100</sup> ln ln( ) TS( ) 100 *SR*

ç ÷ - = --

The influences of the TPB matrix types and RHB contents on the tensile strength and elastic modulus of biocomposites are presented in **Figures 5** and **6**, respectively. In general, the compatibilization of incompatible rHDPE and rPET by E-GMA was capable to improve the tensile properties of CTPB-based biocomposites with and without the presence of RHB. This was because of the E-GMA copolymer that acted as compatibilizer in order to enhance the compatibility and adhesion between rHDPE and rPET components [11]. Therefore, the CTPB blend and composites possessed higher tensile strength and elastic modulus than the UTPB ones.

**Figure 5.** Tensile strength of TPB/RHB biocomposites.

**Figure 6.** Elastic modulus of TPB/RHB biocomposites.

From **Figure 5**, both composites based on UTPB and CTPB show a similar trend. Tensile strength increased with increasing RHB up to maximum values achieved at 70 wt%, which were 20.5 and 22.2 MPa, respectively, for UTPB- and CTPB-based composites. This phenomenon suggests that the adhesion and interfacial bonding between the hydrophilic fibre and hydrophobic matrix are enhanced via the surface modification by coupling agent (MAPE) in which RHB can effectively transfer stress to each other [12]. As RHB content increased up to 80 wt%, it is found that tensile strength of composite decreased but the values were still higher than the pure blend (UTPB and CTPB). This possible reason is that the fibres acted as defects when the content of RHB exceeded the limit, which was at 70 wt% in this study. At high fibre content, fibres were not sufficiently wetted by the matrix (lower content), and this resulted in the fibre agglomerations which blocked the stress distribution. This could be further explained by the fact that the high content of RHB will need higher amount of coupling agent in order to give better interfacial adhesion and reinforcement effect with polymer matrix [13]. Somehow, CTPB-based composites exhibited higher values of tensile strength than that of UTPBbased composites. On the other hand, the increase of elastic modulus (**Figure 6**) with the RHB content (0–70 wt%) can be related to the enhanced stiffness which was imparted by the intrinsic characteristic of RHB. This trend is logic for biocomposite containing low stiffness polymer matrix and high stiffness filler [12]. However, elastic modulus is reduced when 80 wt% RHB is added into the composite. This may be due to the weaker interfacial interaction of RHBpolymer matrix and rHDPE-rPET in UTPB matrix.

#### **3.3. Flexural properties**

**Figure 6.** Elastic modulus of TPB/RHB biocomposites.

34 Composites from Renewable and Sustainable Materials

polymer matrix and rHDPE-rPET in UTPB matrix.

From **Figure 5**, both composites based on UTPB and CTPB show a similar trend. Tensile strength increased with increasing RHB up to maximum values achieved at 70 wt%, which were 20.5 and 22.2 MPa, respectively, for UTPB- and CTPB-based composites. This phenomenon suggests that the adhesion and interfacial bonding between the hydrophilic fibre and hydrophobic matrix are enhanced via the surface modification by coupling agent (MAPE) in which RHB can effectively transfer stress to each other [12]. As RHB content increased up to 80 wt%, it is found that tensile strength of composite decreased but the values were still higher than the pure blend (UTPB and CTPB). This possible reason is that the fibres acted as defects when the content of RHB exceeded the limit, which was at 70 wt% in this study. At high fibre content, fibres were not sufficiently wetted by the matrix (lower content), and this resulted in the fibre agglomerations which blocked the stress distribution. This could be further explained by the fact that the high content of RHB will need higher amount of coupling agent in order to give better interfacial adhesion and reinforcement effect with polymer matrix [13]. Somehow, CTPB-based composites exhibited higher values of tensile strength than that of UTPBbased composites. On the other hand, the increase of elastic modulus (**Figure 6**) with the RHB content (0–70 wt%) can be related to the enhanced stiffness which was imparted by the intrinsic characteristic of RHB. This trend is logic for biocomposite containing low stiffness polymer matrix and high stiffness filler [12]. However, elastic modulus is reduced when 80 wt% RHB is added into the composite. This may be due to the weaker interfacial interaction of RHB-

**Figures 7** and **8** depict the modulus of rupture (MOR) and modulus of elasticity (MOE) of biocomposites based on two types of TPB with different contents of RHB.

**Figure 7.** Modulus of rupture (MOR) of TPB/RHB biocomposites.

**Figure 8.** Modulus of elasticity (MOE) of TPB/RHB biocomposites.

The results showed that the presence of RHB improved the flexural properties of biocomposites. In **Figure 7**, CTPB-based biocomposites had higher MOR values than the UTPB-based biocomposites. This was attributed to the compatibilizing reaction for inherently immiscible TPB matrix which then improved adhesion between fibre and polymer blend matrix besides the enhanced compatibility between polymer blend components. Meanwhile, the incorporation of E-GMA into the CTPB matrix seemed not to provide a significant effect in the MOE of biocomposites (**Figure 8**). Comparing to UTPB-based composite, the MOE for CTPB-based composite reinforced with 50–60 wt% RHB was slightly lower. This phenomenon is associated to the weak intrinsic mechanical properties of E-GMA [14]. However, for the composites with higher content of RHB (70–80 wt%), this factor can be negligible due to relatively low content of the matrix (30 wt% and lower).

#### **3.4. Impact properties**

**Figure 9** presents the impact strength of TPB/RHB biocomposites. For neat blends without the presence of RHB, compatibilization reaction of E-GMA provided a reinforcement effect and led to a significant increase of impact strength for CTPB matrix. However, this large improvement decreased when RHB was added to the blend matrix, which was because rice husk is one kind of stiff inorganic fillers [15]. The impact strength is reduced with RHB content. At high fibre content, there were many interactions between fillers as a result of the filler agglomerations in the composite which were more susceptible to the cracks than the fibre-matrix interface. This indicates that the cracks spread more easily in the biocomposites with high content of rice husk, thus decreasing the impact strength [16].

**Figure 9.** Impact strength of TPB/RHB biocomposites.
