Sustainable end-chain polybenzoxazine functionalized polyethylene glycol for smart light-responsive shape-transformable structures

2.1. Materials

Ferulic acid (98%) and furfurylamine (98%) were sourced from Tokyo Chemical Industry Co., Ltd. in the United States. Paraformaldehyde (AR grade) along with PEGs having number average molecular weights (M_n) of 700, 1000, 1500 and 2000 g⸱mol^-1 were provided by Merck Co., Ltd. in Darmstadt, Germany. All chemicals were utilized in their supplied forms.

2.2. Synthesis of sustainable end-chain benzoxazine functionalized PEG

Sustainable end-chain benzoxazine functionalized PEG (Bz-PEG) was synthesized from ferulic acid, furfurylamine, PEG and paraformaldehyde based on two-step solventless method [23] as presented in Figure 1. First step, the substances, namely ferulic acid and PEG was continuously mixed at a molar ratio of 2:1 in an aluminum pan at 110°C for approximately 40 mins in order to achieve homogeneous mixture via esterification reaction. Then, paraformaldehyde and furfurylamine was gently added at molar ratio of 4:2 into the obtained mixture and stirred at 110°C for 2 h until a clear brown benzoxazine resin functionalized PEG was obtained. After that, the mixture was cooled down, a transparent brown resin was produced.

Close

Figure 1.

Synthesis of Bz-PEG via two-step solventless method and thermal curing reaction of the PBz-PEG.

Export to PPT

2.3. Preparation of sustainable end-chain polybenzoxazine functionalized PEG

The PBz-PEG was synthesized by polymerizing Bz-PEG (Figure 1), which was first placed into aluminum mold and then heated in the oven at 160°C for 1 h, followed by additional heating at 180°C for 1 h. Once the samples had cooled to room temperature (T_R) within the open mold, they were trimmed to the desired shape for characterization.

2.4. Characterization

Curing behaviour of the samples was analysed by differential scanning calorimetry (DSC) model from TA Instruments, specifically the DSC25 brand. The temperature testing profile was elevated by 10°C/min, ranging from 30 to 300°C, while sustaining a nitrogen (N₂) atmosphere purge flow rate of 50 mL/min. The sample mass is approximately 3-5 mg.

Network formation of samples was investigated by Fourier-transform infrared spectrophotometer (FTIR) model Nicolat iS5, Thermo Fisher Scientific (Thailand) Co. Ltd. Co-addiction scans of 128 in the wavenumber region of 4000–400 cm^-1 were used for FTIR spectrum analysis, with a resolution of 4 cm^-1.

Dynamic mechanical properties of the samples were analysed using a dynamic mechanical analyser (model DMA242 from NETZSCH, Inc.). The temperature-dependent profiles for storage modulus (E’) and loss tangent (tan δ) were displayed. The T_g was determined from the peak of the loss tangent curve during the temperature sweep test. The samples, sized at 30×5×0.5 mm³, were examined with a strain amplitude of 5 μm, in a tension mode at a frequency of 1 Hz, and a heating rate of 2°C/min, ranging from 0 to 200°C, in an atmospheric environment. The gage length of the samples was approximately 10 mm.

A scanning electron microscope (SEM, JSM-6510A, JEOL Ltd., Tokyo, Japan) was employed to examine the specimen morphology at an accelerating voltage of 15 kV. Prior to observation, all specimens were coated with a thin layer of gold using a JEOL ion sputter coater (JFC-1200 model).

Thermal stability of the samples was evaluated by thermogravimetric analyser (model TGA550 from TA Instruments, Thailand Ltd.). The samples, weighing 5–10 mg, were heated at a rate of 20°C/min from 30 to 800°C with N₂ flow rate of 50 mL/min.

Thermomechanical analyser (TMA) was used to measure the samples dimensional change (Bruker-AXS, TMA 4010). A fixed load of 5 mN (0.5 g) was applied to the sample in a N₂ atmosphere while the specimen, which had dimensions of 10×5×4 mm³, was heated at a rate of 5^oC/min between -10 to 200^oC.

Tensile properties were assessed using a Universal Testing Machine (Model LR10K) from the brand LLOYD. The dimensions of the samples measured 100×10×0.5 mm³, and the testing was conducted at a crosshead speed of 10 mm/min. A minimum of five tests were conducted, and their results were averaged and reported.

Shape memory performance of the samples was evaluated using an Instron (Thailand) Co., Ltd. universal testing machine (model 5567) equipped with a heating chamber, employing a bending approach. Initially, a sample measuring 50×10×0.5 mm³ was heated to T_g and held at this temperature for 10 min within the heating chamber. Next, an external force of approximately 100 N was applied to bend the sample into a V shape, achieving a 90-degree angle. Subsequently, while maintaining this constant load, the sample was cooled down to T_R. After the load was released, the temporary shape of the sample was established. Following the unloading process, the shape fixity (R_f) was determined through deflection measurements, as specified in Eq. (1) [6].

Where: ${\theta }_0$ and ${\theta }_F$ are the sample angle at original storage state (90^o) and fix state, respectively.

The fixed temporary shape of sample was recovered to the original shape by light exposure to T_g in order to measure its shape recovery (R_N) by using Eq. (2) [6].

Where R_N represents the shape recovery ratio of the N^th thermomechanical bending cycle.

${\theta }_R$ is the sample angle at the recovery state.

The reported shape memory performances were averaged from five samples. A video camera was used to evaluate its recovery process, and the residual angle at the designated time was obtained immediately from a screen capture of the movie [10].

By tracking temperature change of the specimen as a function of time under light exposure, as determined by a thermocouple, the photothermal property of the sample was measured. For 130 seconds, an LED light with a 13.2 mW cm^-2 intensity and a wavelength of 600–650 nm was used, with the source and specimen positioned 20 cm apart. After that, the light was turned off for an additional 130 seconds while the sample temperature was being recorded.

The ultraviolet-visible (UV-Vis)-NIR spectrometer (UV-2600i, Bara Scientific Co., Ltd.) was used to measure the samples light transmission. Every specimen was examined between 200 to 1000 nm in wavelength.

Numerical simulation was used to study the recovery incident and occurred stress of the sample at each recovering time via a commercial ANSYS AUTODYN [24]. The sample having dimension of 50×10×0.5 mm³ was originally generated with the sample angle at fix state. After that, the sample was applied with temperature at T_g. Two edges of the sample model were fixed under the nonlinear orthotropic materials. Same conditions as in the experimental process were carried out in the simulation and compared carefully in order to achieve an accurate comparison. The generated structure at the initial fix state was thermally recovered under temperature of T_g to study the potential use and occurred maximum stress on the sample.

3.1. Curing behaviours and network formation of end-chain benzoxazine functionalized PEG at various PEG molecular weights

Curing behaviours of Bz-PEGs at various PEG molecular weights were investigated by DSC. Figure 2(a) shows the curing exotherms of the Bz-PEGs at various M_n of PEG of 700, 1000, 1500 and 2000 g⸱mol^-1. All Bz-PEGs showed a maximum exothermic peak at about 220^oC, which is indicative of an oxazine ring opening process [13]. With increasing PEG molecular weight of 700, 1000, 1500 and 2000 g⸱mol^-1, the heat of reaction as measured from the integrating area of the exothermic peak of the Bz-PEG increased marginally to be 15.9, 20.6, 21.8 and 24.0 J/g, respectively. The dilution caused by the presence of aliphatic PEG affected on the heat of reaction shifting it to the higher [25].

Close

Figure 2.

(a) DSC thermograms of Bz-PEG at various PEG molecular weight (i) Bz-PEG_700, (ii) Bz-PEG₁₀₀₀, (iii) Bz-PEG₁₅₀₀ (iv) Bz-PEG₂₀₀₀ and (b) FTIR spectra of (i) PBz-PEG₂₀₀₀ and (ii) Bz-PEG₂₀₀₀ .

Export to PPT

To study the network formation of the Bz-PEG sample, the sample at M_n of PEG at 2000 g⸱mol^-1 was representatively confirmed by FTIR as shown in Figure 2(b) due to the highest heat of reaction. The absorption peak of Bz-PEG at 1477 cm^-1 was associated with the tetra-substituted benzene ring [26,27]. The peaks at 930 and 1230 cm^-1 were also identified as being derived from the aromatic ether (C-O-C stretching) of the oxazine ring [28]. The PEG characteristics exhibited the peaks at 3340 (–OH stretching), 2870 (C–H stretching) and 1030 cm⁻¹ (primary alcohol C–O stretching and –OH in plane deformation). The peaks at 766 and 1590 cm⁻¹ were identified as furan group [29]. It was determined that C=O and C-O stretching of ester was showed the bands at 1680 and 1100 cm^-1, respectively.

As observed in Figure 2(b), after thermal curing, the absorbance bands at 930 and 1230 cm⁻¹ associated with the oxazine ring disappeared, signifying that the oxazine ring-opening polymerization of Bz-PEG took place [21]. The appearance of a shifted broad peak at around 3400 cm^-1, associated with the formation of phenolic hydroxyl groups (OH), suggests that a ring-opening reaction occurred ortho to the phenolic moiety of the benzoxazine monomer during thermal treatment [26,27]. The results indicated that the end-chain polybenzoxazine functionalized PEG was successfully prepared by thermal curing. Therefore, the curing condition of heating at 160°C for 1 h followed by curing at 180°C for 1 h was utilized for polymerization of all PBz-PEG at various PEG molecular weight. Curing condition of our sustainable PBz-PEG also exhibited lower values than that of petroleum-based bisphenol-A/aniline derived polybenzoxazine, i.e. curing temperature of ∼200^oC for 2 h [30].

3.2. Dynamic mechanical property of end-chain polybenzoxazine functionalized PEG at various PEG molecular weights

The influence of the M_n of PEG on dynamic mechanical property of the PBz-PEG was examined by using dynamic mechanical analysis (DMA). Figure 3(a) displays the E’ values for the samples at the glassy state (0°C), indicating the materials stiffness as a function of temperature. The E’ values of PBz-PEG were systematically decreased from 5.15 to 0.28 GPa as the molecular weight of PEG enhanced from 700 to 2000 g⸱mol^-1. Our hypothesis is that a very low rigidity of PEG (2-1300 kPa [31]) decreased the stiffness of the sample.

Close

Figure 3.

(a) storage modulus and (b) loss tangent of PBz-PEG at various PEG molecular weights.

Export to PPT

A key factor influencing SMP programming is the transition temperature. The T_gs of the PBz-PEG samples were identified from the peak of loss tangent curves as demonstrated in Figure 3(b). As M_n of PEG was varied from 700 to 2000 g⸱mol^-1, T_g of PBz-PEG decreased from 115 to 80°C. This reduction could be because of the increased flexibility of PEG, which facilitated the polymer chains movement and enhanced the polymer networks mobility [32]. The findings also showed that PBz-PEG specimen had a heterogeneous network because its T_g showed two separate peaks. The shape memory performance i.e., shape fixity and shape recovery of the PBz-PEG was significantly influenced by its heterogeneous network containing hard segment and flexible segment [33] represented from the higher T_g (ranging in 80-115^oC) and lower T_g (ranging in 10-25^oC) of the sample, respectively. Moreover, the crosslink density (ρ_x) of the polymer network could be determined from the rubbery plateau modulus based on Eq. (3) [34].

where E’ is rubbery plateau storage modulus at T_r. T_r is equal to T_g+30^oC (Kelvin; K) and R is the gas constant which equal to 8.314 m³⸱Pa⸱K^-1⸱mol^-1.

The crosslink density of PBz-PEG was calculated to be 6176, 1915, 1216 and 624 mol⸱m^-3 at PEG molecular weight of 700, 1000, 1500 to 2000 g⸱mol^-1, respectively. The ρ_x was systematically decreased with increasing M_n of PEG resulted to the decrease in T_g of PBz-PEGs. The findings implied that higher molecular weight of PEG in PBz-PEG leads to lower energy requirements for SMP programming and light-responsive recovering due to the lowered T_g and crosslink density.

3.3. Morphology of end-chain polybenzoxazine functionalized PEG at various PEG molecular weights

The morphology of the PBz-PEG samples is studied by using a scanning electron microscope (SEM). The SEM micrographs of the PBz-PEG surface at different PEG molecular weights are displayed in Figure 4. As shown in Figures 4(a-d), a rough fracture surface of the PBz-PEG was found and there was no phase separation in all specimen. An increase in the PEG molecular weight was accompanied by a noticeable enhancement in the surface roughness of the specimens, suggesting a correlation between chain length and morphological. Furthermore, because PEG is flexible, their rough surface showed that the PEG was comparatively evenly distributed in the PBz-PEG. The results indicated that the functionalization of PEG into polybenzoxazine enhanced the flexibility behaviour of the specimens.

Close

Figure 4.

SEM micrographs of PBz-PEG samples at 1000x magnification at various PEG molecular weights: (a) 700 g⸱mol^-1 (b) 1000 g⸱mol^-1 (c) 1500 g⸱mol^-1 and (d) 2000 g⸱mol^-1.

Export to PPT

3.4. Thermal stability of end-chain polybenzoxazine functionalized PEG at various PEG molecular weights

Thermal stability of the PBz-PEG samples at various molecular weights of PEG, i.e., T_d10 and char yield at 800^oC was studied by TGA, as illustrated in Table 1. The T_d10 of the PBz-PEG samples were enhanced with increasing M_n of PEG. This enhancement in thermal stability of the PBz-PEG samples might be due to the dilution caused by the presence of PEG resulted in T_d10 shifting it to the higher temperatures [25]. Moreover, the char yield at 800^oC of the PBz-PEG also systematically decreased with increasing PEG molecular weight. The reduced char yield might be attributed to the higher molecular weight of the PEG, which led to a decrease in the crosslinked density of the sample [25]. Thermal stability in term of T_d10 of our PBz-PEG was observed to be higher than that of PLA/PEG at PEG M_n of 2000 g mol^-1 (T_d10 ∼240-260^oC) [35]. These results offered compelling evidence for thermal stability improvement of sustainable PBz-PEG by the incorporation of the PEG.

3.5. Dimensional change of end-chain polybenzoxazine functionalized PEG at various PEG molecular weights

The PBz-PEG samples’ dimensional response to temperature is crucial for enhancing shape memory capacity. In order to assess the dimensional response to temperature change of the PBz-PEG samples at different PEG molecular weights as listed in Table 1, the coefficient of thermal expansion (CTE) was calculated from the slope changes of the dimension change curves between 0 - 150°C examined by Thermomechanical analysis (TMA). As the M_n of PEG increased, CTE values of the sustainable polymers increased systematically. This may be because the longer PEG chain decreased the crosslinked density of the sample, promoting polymer chain mobility and enhancing dimensional change ability of the PBz-PEG sample [36]. The findings implied that the dimensional change ability of sustainable PBz-PEG which could result in the improvement of shape recovery ability by the incorporation of higher molecular weight of the PEG.

3.6. Mechanical properties of end-chain polybenzoxazine functionalized PEG at various PEG molecular weights

Strength and modulus under tension mode of the PBz-PEG at various M_n of PEG of 700, 1000, 1500 and 2000 g⸱mol^-1 were examined as depicted in Table 1. From the results, tensile strength (σ_t) values of the PBz-PEGs were increased up to 1.41 MPa with the addition of PEG molecular weight of 1000 g⸱mol^-1. This might be due to the incorporated chemical crosslinking in the PBz-PEG, while the more pliable PEG contributes to the enhanced flexibility of the PBz-PEG, resulting in improved load transfer for the sample. Whereas, that of the PBz-PEG with PEG molecular weight beyond 1000 g⸱mol^-1 was slightly decreased to 0.26 MPa due to adequate PEG softening [37]. Moreover, as expected, tensile modulus (E_t) of the PBz-PEGs at various M_n of PEG in the range of 700 to 2000 g⸱mol^-1 were found to be decreased from 0.210 GPa to 0.061 GPa, respectively. This could be explained by the possibility that the stiffness of the polymer would be reduced if a softer, aliphatic PEG was incorporated [5]. This is consistent with earlier storage modulus findings.

3.7. Photothermal property of end-chain polybenzoxazine functionalized PEG at various PEG molecular weights

The photothermal property of sustainable PBz-PEG at various M_n of PEG was examined under light exposure at wavelengths of 600–650 nm in order to develop light-responsive SMP, as illustrated in Figure 5(a). A thermocouple was used to measure the sample temperature at different exposure times. According to the findings, heat energy of pure PEG at M_n of 2000 g mol^-1 was slightly increased to only approximately 37^oC during light exposure suggesting very low light-response of the PEG. Whereas, those of the PBz-PEG samples increased steadily over the course of 130 seconds after being exposed to light, reaching about 110°C. Because of the π-π interaction between the polybenzoxazine chains, which allowed them to absorb light and transform photon energy into heat energy, the sustainable polymers were thermally induced [9,38]. The results suggested that the photothermal properties of the PBz-PEG was exhibited by the polybenzoxazine structure. Moreover, the findings also suggested that indirect heat from light might cause dimensional changes and a shape recovery mechanism in the sustainable polymers.

Close

Figure 5.

(a) Light induced temperature with exposure time and (b) the UV–Vis–NIR transmittance spectra of PBz-PEG at various PEG molecular weights.

Export to PPT

Photothermal property can be supported by UV–visible–Near infrared (UV–Vis–NIR) transmittance results. UV-Vis transmittance of all sustainable polymers is absent from the 200–700 nm range, as shown in Figure 5(b), suggesting total UV-Vis light absorption. The findings suggested that the π-π interactions in polymer chains could produce heat energy of PBz-PEG through UV-Vis light absorption [9,38]. In contrast, Vis-NIR transmittance performance of PBz-PEG at 700–1000 nm was decreased as M_n of PEG increases. The results showed that adding a longer PEG chain resulted in a noticeably lower Vis-NIR transmittance, indicating that NIR light was absorbed and blocked. This could be the result of a lengthy PEG chain that helps to enhance absorption capacity within the samples and block NIR transmission. According to the results, the sustainable PBz-PEG had a UV-Vis light-protective effect via UV-Vis light absorption mechanism [39], making it appropriate for use as a structural material that protects against UV-Vis radiation.

3.8. Shape memory properties of end-chain polybenzoxazine functionalized PEG

At certain particular compositions, our sustainable SMPs demonstrated shape memory effects (SMEs). Shape memory qualities were examined, as exhibited in Figure 6(a), in order to assess the shape memory performances of the developed PBz-PEGs at various molecular weights of PEG. An essential factor characterizing capacity of SMPs to commit a transient or permanent shape to memory is their shape fixity which related to the stiffness of the sample. Table 2 displays the shape fixity of the PBz-PEG samples. The R_f of the PBz-PEG samples diminished as the M_n of PEG increased, attributable to the longer PEG chains reducing the stiffness of the samples. This phenomenon was in a good agreement with the decrease in modulus as reported from DMA and tensile properties results.

Close

Figure 6.

(a) SMP programming and recovery set up (b) series of the recovery process by light exposure of the PBz-PEG samples at various PEG molecular weights.

Export to PPT

Shape recovery is another factor that reflects the effectiveness of retaining the original form of the SMP. Table 2 further demonstrates the relationship between shape recovery by light exposure and the molecular weight of PEG in the PBz-PEG samples. As the molecular weight of PEG increased, the R_N values of the PBz-PEGs also rose. It might be due to the chain flexibility and mobility of PEG which have ability to dimensional changes of the PBz-PEG supported from CTE results, resulted in the increase of their recovery [5]. In addition, the shape recovery of our PBz-PEGs was found to be greater than that of thermos-responsive PEG-based SMP (∼82%) [40] and light-responsive bio-derived SMP based on copolymer of benzoxazine/epoxy (∼82%) [41].

The process of shape recovery for the PBz-PEG samples subjected to bending load is illustrated in Figure 6(b). Shape recovery as a function of recovering time and temperature generated by light exposure of PBz-PEG at various PEG molecular weights was also illustrated as shown in Figures 7(a and b), respectively. Recovering times of the PBz-PEG samples were rapid as PEG molecular weights increased. This phenomenon might also be affected by the PEG chain mobility and dimensional change under heating by light exposure which could promote the shorter recovering time of the PBz-PEG sample. Moreover, it could be noticed that when the generated temperature by light was reached T_g of the PBz-PEG samples, the R_N tended to meet the maximum recovery value of the samples. This might be thanks to the dimensional response of the sustainable polymers to T_g that help completely recovery their shape. The findings indicated that the addition of PEG with M_n of 1000 g⸱mol^-1 to the PBz-PEG sample demonstrated excellent shape memory characteristics, achieving a relatively high R_f of 90%, a notable R_N of 94%, and a rapid recovering time of 115 s.

Close

Figure 7.

Shape recovery as a function of (a) recovering time and (b) temperature generated by light exposure during the shape recovery process of PBz-PEG at various PEG molecular weights.

Export to PPT

A possible shape memory effect of sustainable end-chain PBz-PEG₁₀₀₀ is presented in Figure 8. The findings indicated that SMPs could be distorted at T_g. In this stage, the cross-linking point of polybenzoxazine network was stored the stress from applying external load and the flexible PEG chain was expanded during heating to transform their shape into desired shape. After cooling while applying load to temporary shape, the shape could be fixed. When SMP is further heated by light exposure, the shape of materials autonomously changes to its original shape. In this stage, the stress in polybenzoxazine network was released as driving forced, while, the PEG chain was squeezed macroscopically (push spring) and stored potential energy, which caused the sample to change back to its original shape during heating by light exposure.

Close

Figure 8.

Shape memory effect of sustainable end-chain PBz-PEG₁₀₀₀.

Export to PPT

3.9. Feasibility application for shape-transformable structure of end-chain polybenzoxazine functionalized PEG by numerical simulation

Numerical simulation was employed to examine the crucial shape recovery process of sustainable end-chain PBz-PEG₁₀₀₀ to its original form at different recovery times as seen in Figure 9(a). The material properties of specimen are presented in Table 3. The experimental specimen dimension was also input to create numerical specimen model. From Figure 9(a), the numerical angle of the sample was found to be 99^o, 125^o, 164^o and 174^o at recovering time of 0, 15, 50 and 115 s, respectively. The experimental shape memory process of sustainable end-chain PBz-PEG₁₀₀₀ at the same recovering time was also shown in Figure 9(b) to accurately and qualitatively compare with those from numerical simulation. The numerical and experimental R_N were calculated by using Eq. (2) and quantitatively compared as reported in Table 4. According to the findings, there was less than 3.2% discrepancy between the experimental and predicted shape recovery. The numerical results are highly correlated with the experimental results of shape-transformable PBz-PEG₁₀₀₀. The results suggested that numerical simulation model is represented an accurate shape-transforming mechanism of the PBz-PEG₁₀₀₀ sample.

Close

Figure 9.

Comparison of (a) numerical and (b) experimental recovery process of SMP based on sustainable end-chain PBz-PEG₁₀₀₀.

Export to PPT

Furthermore, the predicted stress occurred during recovery process of the sample was found at bending region with the maximum value of 1.8514 Pa. The maximum stress occurred in the sample was significantly lower than the strength of the PBz-PEG₁₀₀₀ (1.41 MPa) which implied that the shape-transformable sample based on PBz-PEG₁₀₀₀ has ability to repeat the recovery process without any damage. The results revealed that sustainable end-chain PBz-PEG₁₀₀₀ having good shape memory performance is suitable for using as shape-transformable structure application such as shape-transformable window as shown in Figure 10.

Close

Figure 10.

Shape-transformable window based on sustainable end-chain PBz-PEG₁₀₀₀.

Export to PPT