3.1. FTIR Analysis
The FTIR spectra of the neat PLA, control, and blends prepared with various HDI contents are shown in
Figure 2. A peak corresponding to the –CH
3 of PLA at 2945 cm
−1, 2996 cm
−1 [
38] and the –CH– of PLA at 2997 cm
−1 (asym), 2946 cm
−1 (sym), and 2881 cm
−1 stretch [
39]. The –CH
2 of PBSeT was appeared at 2919 cm
−1, 2851 cm
−1 [
36]. However, it occurred nearby each other, so whole peaks were appeared in the range 2800–3000 cm
−1 as small round curves. In addition, the peaks at approximately 1700 and 1180 cm
−1 correspond to the carbonyl C=O and C–O vibrations, respectively [
40,
41,
42]. In the spectra of both PBSeT and PLA, peaks in the 1700 and 2800–2900 cm
−1 regions correspond to the C=O bond and aliphatic –CH
2 stretching, respectively [
43]. The spectrum of PBSeT mainly shows peaks in the 1200 and 1100 cm
−1 regions corresponding to the asymmetric and symmetric stretching of aromatic CO, respectively. In particular, the spectra of all the PBSeT/PLA blends showed a peak in the 720 cm
−1 region corresponding to the aromatic group, indicating that the PBSeT and PLA had blended well [
44]. With increasing HDI content, peaks corresponding to –C=OO
− were observed in the 1590 and 660 cm
−1 regions. In particular, the peak at 1590 cm
−1 (attributed to NH bending) indicates that the –N=C=O group had changed into the –NH–C=OO
− one [
32,
45]. In addition, a peak attributed to NH stretching appeared in the region at approximately 3300 cm
−1 when –N=C=O groups were bonded [
32,
46], and the intensity of the peak at approximately 2300 cm
−1 (corresponding to the HDI isocyanate group) was very weak (i.e., 0.1). Furthermore, the spectrum for the HDI 0.3 blend specimen clearly showed peaks attributed to the HDI isocyanate group [
47], suggesting that the HDI had reacted appropriately with the PBSeT and PLA.
3.2. GPC Analysis
Table 2 and
Figure 3 shows the GPC curves used to determine the
Mw,
Mn, and Đ of the control and blends prepared with various HDI contents. The control and all the blend specimens showed
Mw values in the range ~160,000 to 170,000 g mol
−1, lower than the known
Mw of PLA. In addition, PBSeT showed
Mw and
Mn of 154,900 and 64,600 g mol
−1, respectively. The control showed both the lowest
Mw and
Mn molecular weights of 114,000 and 45,400 g mol
−1, respectively, which were remarkably lower than those of all the HDI specimens, suggesting that the HDI had increased the molecular weights of the specimens. The control and the HDI 0.1, 0.3, 0.75, and 1.0 specimens showed
Mw molecular weights of 114,000; 148,200; 152,000; 146,300; and 131,300 and
Mn molecular weights of 45,400; 67,800; 57,100; 56,800; and 54,400 g mol
−1, respectively. The HDI 0.3 specimen showed the highest molecular weight and, therefore, high tensile strength and elongation in the tensile-property analysis. The results suggest that the physical properties of the blends were improved by adding HDI. Moreover, the additional peaks in the FTIR spectra of the HDI specimens suggest that the molecular weights increased owing to chemical bonding.
3.3. Tensile Strength and Elongation Properties
Figure 4 and
Table 3 shows the tensile-strength characteristics and stress–strain curves of the neat PLA, control, and blend specimens prepared with various HDI contents. According to previous research results, because PBSeT showed high elongation, it was expected to remarkably affect the tensile properties of the highly brittle PLA [
36]. The neat PLA and simple-blend control specimens showed tensile strengths of approximately 93 and approximately 52 MPa, respectively. The results of the one-way analysis of variance (ANOVA) with
post-
hoc Tukey honestly significant difference (HSD) showed a statistically significant difference between the tensile strengths of the control and PLA specimens. Furthermore, the control and PLA specimens showed elongations of approximately 43 and approximately 23%, respectively. However, because the standard deviation of the control specimen was relatively high in the significance validation, the difference between the elongations was negligible. The tensile strengths of all the HDI specimens were significantly different from that of the control specimen. However, among the HDI specimens, only the HDI 0.3 one showed a significantly different tensile strength of 64 MPa.
The average tensile strengths of the neat PLA and HDI 0.1, 0.3, 0.75, and 1.0 specimens were 52, 59, 64, 62, and 60 MPa, respectively. Although the average tensile strength of the HDI 0.3 specimen was not significantly different from that of the HDI 0.75 one, it was significantly different from those of all the other specimens. In addition, the HDI 0.3 specimen remarkably showed approximately 151% elongation, which was significantly different from the elongations of all the other specimens. The neat PLA specimen showed the lowest elongation as 23%, and the elongations of the control, HDI 0.1, 0.3, 0.75, and 1.0 specimens were 43, 54, 151, 73, and 63% of elongation, respectively. Therefore, although adding HDI increased specimen elongation up to a point similar to the findings of Kim et al. (2012) [
48], adding more than 0.75 phr of HDI was an overdose, which decreased the specimen elongation.
Interestingly, with increasing HDI content from 0.3 to 1.0 phr, the specimen elongation and tensile strength both decreased, suggesting that 0.3 phr was the most appropriate HDI content for increasing the tensile strength and elongation of the blend specimens. Because HDI contents of 0.75 phr or higher caused binding reactions such as self- and branch bonding, the HDI 0.75 and 1.0 specimens did not exhibit appropriately high tensile strength or elongation, which may be why the HDI 0.75 and 1.0 specimens showed two melting temperature (
Tm) peaks in the DSC thermal analysis, as will be further discussed in
Section 3.5. Furthermore, such binding reactions may be why the HDI was not well dispersed throughout the matrixes of the HDI 0.75 and 1.0 specimens and, thus, why sufficient binding was not achieved. This finding means that when an appropriate amount of modifier is added to the blend, many functional sites can react and be efficiently combined with the modifier. However, it is expected that when a large amount of HDI is added, the binding efficiency is lowered because the HDI cannot be combined with all the functional sites, which had already reacted with each other.
Although the specimen strength and elongation both eventually increased owing to HDI-induced bonding, the two materials were not completely bonded together. Although HDI contents above 0.3 phr were excessive, the HDI is highly reactive and appeared to react quickly wherein all the reaction sites reacted with each other. Therefore, additional investigations (wherein the screw combination and velocity and temperature are varied) are required to elucidate the bonding mechanism and optimize the blending method.
Figure 4 was the stress–strain curves show that except for the neat PLA specimen, all the HDI blend specimens showed improved strains owing to the high PBSeT stretchability. However, some specimens showed necking, which is thought to occur past the PLA yield point, owing to the PBSeT, and the average strains of the HDI blends were slightly different depending on the HDI content.
Necking past the yield point and the phenomena leading to fracture are indicated by the initial slope in the stress curve as steep as the slope the PLA tensile-strength curve before the appearance of the PBSeT stretchability and ductility characteristics. The control, which consisted of only PBSeT blended with PLA, showed improved average strain compared with that of the neat PLA. Furthermore, adding HDI increased the average strains compared with those of the neat PLA and control specimens. In particular, the HDI 0.3 specimen showed the highest strain of approximately 150%.
3.4. Three-Point Flexural Strengths
Figure 5 and
Table 4 shows the 3-point flexural strengths and moduli of the neat PLA, control, and blend specimens prepared with various HDI contents. When PBSeT was blended with neat PLA, the flexural strength decreased from approximately 150 (for the neat PLA) to 95 MPa (for the blended control), suggesting that the relatively brittle PLA was made more ductile by adding the elastomeric PBSeT. However, all the HDI specimens showed higher flexural strengths than the control. Specifically, the HDI 0.1, 0.3, 0.75, and 1.0 specimens exhibited flexural strengths of 106, 108, 108, and 108 MPa, respectively, which although were higher than that of the control, were not remarkably different. However, the control and HDI 0.1, 0.3, 0.75, and 1.0 specimens showed yield strengths of approximately 81, 88, 91, 95, and 95 MPa, respectively. Although the yield strength increased with increasing HDI content, the yield strengths of the HDI 0.75 and 1.0 specimens were not remarkably different. The flexural strengths of all the HDI specimens are similar despite the increased yield-point strength because the experiment was conducted according to the ISO standard method of terminating the experiment when the strain reached 5%. Although the maximum flexural strengths of the specimens appeared similar up to a certain point, the yield-point strength slightly increased with increasing HDI content. In particular, the flexural modulus of the HDI 0.1 specimen was approximately 7 GPa, which is higher than that (6.5 GPa) of the control specimen, indicating that the HDI had increased the momentary modulus. However, the HDI 0.3, 0.75, and 1.0 specimens showed flexural moduli of 6.4, 6.6, and 7.1 GPa, respectively, indicating that the flexural modulus slightly decreased and then increased with increasing HDI content and that the HDI 0.3 specimen showed the maximum elongation of approximately 150%. Although the HDI 0.3 specimen showed a somewhat decreased flexural modulus, the yield-point strength was not reached during a longer strain, suggesting that the HDI 0.3 specimen could withstand strain without fracturing.
3.5. Thermal Properties
The thermal properties of the control and blend specimens were investigated by DSC. The DSC curves of the specimens heated a second time (i.e., the “second heating curves”) were used to determine the melting points and characterize the thermal properties of the specimens because the second heating curves provide more-accurate data independent of the thermal histories of the specimens.
Figure 6 shows the DSC curves of the control and blend specimens prepared with various HDI contents, and
Table 5 shows
Tg, the cold-crystallization temperature (
Tcc),
Tm, and the other thermal properties of the specimens. A peak corresponding to
Tg appeared in the DSC curves of all the specimens, and the peak shifted depending on the HDI content of the specimen. The DSC curve of the control showed a
Tg peak at approximately 53 °C, which is slightly lower than the
Tg previously known for neat PLA. The DSC curve of the control also showed a peak corresponding to
Tcc at 87 °C, which is lower than the
Tcc of the other specimens, indicating a fairly low
Tcc. It previously was thought that PBSeT affected the PLA crystallization rate and that the PLA cold-crystallization heat capacity was 6.2 J g
−1, which is lower than those of the other specimens. However, in the current work, although the
Tcc of the control was 87 °C, the
Tcc peak shifted to higher temperatures and the peak sizes increased with increasing HDI content. For example, although the
Tcc of the HDI 0.1 and 0.3 specimens was approximately 105 °C, that of the HDI 0.75 and 1.0 ones increased to approximately 115 °C, which is associated with increased crystallinity and is thought to originate from the increasing size and number of crystals with increasing HDI content. This finding suggests that the HDI and PBSeT contents both affect the crystallization of the main PLA domain because HDI and PBSeT both are almost amorphous and in particular, it suggests that the HDI-centered crystal structure affects the PLA crystallization.
The control and HDI 0.1, 0.3, 0.75, and 1.0 specimens, showed 7, 18, 30, 31, and 32% crystallinity, respectively. Although the crystallinity increased with increasing HDI content and the crystallinity of the HDI 0.1 specimen did not increase significantly to 18%, the crystallinity of the HDI 0.3 specimen did significantly increase to 30%. However, the crystallinities of HDI 0.75 and 1.0 specimens were not significantly different.
The control specimen showed the lowest Tm of 163 °C, and a relaxation zone appeared before Tm in the DSC second-heating curve of the control owing to insufficient PBSeT and PLA mixing. However, the relaxation zone disappeared when HDI was added to the specimens. The Tm of all the specimens were measured in the range ~163–167 °C and were all slightly lower than the Tm previously known for neat PLA, except HDI 0.75 and 1.0 which showed two Tm peaks.
The DSC second-heating curves of the HDI 0.75 and 1.0 specimens showed similar Tm peaks at 159 and 167 °C, respectively, unlike the curves of the other specimens. The Tm measured for the neat PLA is affected by either changes in the PLA crystal shapes and sizes owing to the addition of a nucleating agent or by the increasing thicknesses of external and internal crystals. In this study, the Tm was affected by the formation of additional crystals when HDI was added to the specimens.
The mechanical and physical properties of blends consisting of an amorphous polymer (such as PBSeT) blended with a semicrystalline polymer (such as PLA) are greatly affected by the crystallization kinetics, degree of crystallinity, and microstructure of each of the individual blends in the same structure. The miscibility, properties, and internal structural interactions of amorphous/semicrystalline blends greatly influence the crystallinity and physical strength. Usually, the higher the polymer crystallinity is, the higher the physical stiffness (i.e., elasticity modulus) and tensile strength are, which decrease ductility. Although most of the specimens in this study followed that general rule, the HDI 0.3 specimen showed drastically increased tensile strength and elongation.
Figure 7 shows the TGA curves of the control and blend specimens prepared with various HDI contents. All the specimens were sufficiently dried before testing, and none of them showed any weight loss at approximately 100 °C. All the specimens were pyrolyzed above approximately 340 °C, and the TGA curves of all the specimens exhibited peaks indicating two-stage decomposition because PLA and PBSeT show different decomposition temperatures. Previously reported main decomposition temperature of PBSeT was around 400 °C [
36].
However, the main peaks of all the HDI specimens were at decomposition temperatures higher than that of the main peak of the control (PBSeT/PLA blend) specimen. That is, the control specimen mainly decomposed at approximately 370 °C, while the HDI 0.1, 0.3, 0.75, and 1.0 specimens decomposed at 372, 371, 372, and 371 °C, respectively. The decomposition and decomposition-onset temperatures of the HDI specimens were not remarkably different from those of the control specimen, indicating that the HDI had only slightly changed the heat resistance of the blends.