2.3.7. Hydrophilicity Test

The water absorption rate of the test sample was tested in accordance with the plastic water absorption test method (GB1034-98). The blends were cut into small pieces with sizes of 10 × 10 × 2 mm and placed in a vacuum oven at 50 ◦C for 48 h. The initial weight m1 of a piece was measured using an analytical balance. Subsequently, the sample was soaked in deionized water for 24 h, wiped with filter paper, and weighed to determine the final weight m2. The water absorption rate of the blend materials was determined as [17].

$$\text{Water absorption} = \frac{m\_2 - m\_1}{m\_1} \times 100\% \tag{2}$$

## **3. Results and Discussions**

## *3.1. Mechanical Properties of PLA*/*PSt*/*PEG*/*CA Blends*

The mechanical properties of pure PLA and PLA/PSt/PEG/CA blends with di fferent PSt mass ratios are shown in Figure 1. The elongation at break and the notched impact strength of the PLA are 10.45% and 1.58 kJ·m<sup>−</sup>2, respectively, which reflects the brittleness of PLA. Compared with pure PLA, the elongation at break of PLA/PEG/CA blends increased slightly, while the tensile strength decreased, which may be due to the ability of low-molecular-weight PEG to plasticize PLA [18]. There is also an explanation that under the action of high temperature shear, a small amount of water and CA in PEG can promote the depolymerization of PLA [19], and the low-molecular-weight PLA produced by depolymerization can also have a plasticizing e ffect on PLA [20]. When the PSt content is less than 10wt%, the tensile strength of the blend decreases with the increase in PSt content, which is mainly due to the plasticization of PLA. The change in elongation at break is not obvious, which indicates that a new interaction force is generated between the blended components and a new structure is formed. This structure works together with the plasticized PLA to o ffset the defects caused by the incompatibility of PLA and starch, and this structure can improve the impact strength of the blend. The elongation at break and the impact strength of the blends show a significant upward trend as the PSt content increases and reaches the maximum at 30 wt%. This phenomenon can be attributed to the increased interaction between the components of the blend. Compared with the plasticizing e ffect of PEG and CA on PLA, the interaction force between the blending components at this time has a more significant e ffect on improving the toughness of the PLA/PSt/PEG/CA blend. When the PSt content exceeds 30 wt%, the elongation at break slowly decreases. The excess PSt forms an agglomeration, and concentration of stress occurs during the stretching process of the blend and results in increased defects. Conversely, the tensile strength of the blend monotonically decreases with the increase in the PSt content, which indicates that some components in the blend enhance the mobility of the PLA molecular chain. The results of the impact test are consistent with the trend of the elongation at break. The impact strength of the blend first increased and then decreased with the increase in PSt content, and reached the maximum when the PSt content was 30 wt%, indicating that the adhesion between PLA and starch was the largest at this time. When the PSt content exceeds 30 wt%, the possibility of PSt agglomeration in the blend increases, resulting in a decrease in the mechanical properties of the material. Therefore, a blend with a PSt content of 30 wt% was selected for follow-up research.

**Figure 1.** Mechanical properties of pure poly(lactic acid) (PLA) and PLA/premixed starch (PSt)/polyethylene glycol (PEG)/citric acid (CA) blends with di fferent PSt contents.

To explore the contribution of each component of the PLA/PSt/PEG/CA blend to the mechanical properties of the blend, several blends with di fferent components were fabricated. The test results are summarized in Table 2. Most raw materials blended with PLA alone caused the toughness of the blend to decrease, and only the PLA/PSt/CA blend and PLA/PSt/PEG/CA blend had a sudden change in toughness. This phenomenon indicates that the reason for the improved toughness of PLA/starch composites may be that the PEG and CA in the system improve the interface interaction between PLA and starch. Compared with pure PLA, the tensile strength and elongation at break of the PLA/PSt blends significantly declined. This phenomenon may be due to the weak interaction between the PLA and PSt components, which results in weak tangles among molecular chains and poor compatibility among components. After introducing a small amount of PEG into the PLA/PSt blend, the elongation at break of the new blend slightly increased and the tensile strength decreased. This result can be attributed to the plasticizing e ffect of PEG. PEG can enhance the mobility of the PLA molecular chain and increase the tensile toughness of the PLA/PSt/PEG blends. However, the PLA/PSt/CA blends showed obvious strain hardening, and their elongation at break reached 54.47%. After introducing PEG, the elongation at break of the PLA/PSt/PEG/CA blend significantly increased to 140.51%. The addition of CA increases the interaction force among the blend components, and at this time the negative e ffects of CA and water on PLA can be ignored. However, when CA and PEG are present in the blend at the same time, PEG acts more than just a plasticizer in the PLA/PSt/PEG/CA blend. Compared with that of the PLA/PSt/CA blend, the tensile toughness of the PLA/PSt/PEG/CA blend greatly improves. PEG may chemically react with other components during the blending process. This reaction can promote the movement of the PLA molecular chain and improve the compatibility among the blend components.



## *3.2. Rheological Behavior of PLA*/*PSt*/*PEG*/*CA Blends*

By testing the rheological properties of the blend, it is helpful to analyze the changes in the compatibility of the blend components. Generally, after PLA is blended with poorly compatible materials such as polyamide11 [21], starch [22], and cellulose [23], the viscosity of the composite material decreases. The introduction of compatibilizers, coupling agents or reactive compatibilizers can increase the interaction force e ffectively between the interfaces of the components of the composite material, thereby increasing the viscosity of the composite material. The rheological properties of the blends, such as storage modulus, complex viscosity, and loss modulus, were analyzed to explore the e ffects of PSt, PEG, and CA on the melt strength of the blend. The presence of PSt can change the viscosity and degree of shear-thinning [24] as compared with neat PLA. The dependence of the complex viscosity of di fferent blends on frequency is illustrated in Figure 2a. The figure shows two types of the complex viscosity frequency dependence of the blends. In the low-frequency region, the viscosity of the blends after PSt addition decreases in varying degrees compared with that of pure PLA. Among the blends, the decrease in the viscosity of the PLA/PSt/PEG blends is the most obvious. This result can be attributed to the immiscibility of PLA and starch and the plasticizing effect of PEG on PLA [25]. The poorly compatible blend components are prone to phase separation, forming a defect structure with weak interaction force [26]. When external forces act on the blend, this defect structure cannot absorb energy, resulting in a decrease in the viscosity of the blend. PEG is

dispersed among the PLA molecular chains during the blending process, and such dispersion enhances the mobility of the PLA molecular chains and reduces the viscosity [27] of the blend. The viscosity of the blend without CA does not change with the change in the low-frequency region, thereby showing the properties of a Newtonian fluid, as well as an obvious shear thinning in the high-frequency region. The platform of the curve of the viscosity of the CA-containing blend in the low-frequency region disappears, suggesting that CA can increase the shear-thinning strength of the blend. The degree of entanglement among the components of the blend hinders the movement of the PLA molecular chain and therefore results in increased viscosity in the low-frequency region.

**Figure 2.** (**a**) Complex viscosity, (**b**) storage modulus, (**c**) loss modulus, and (**d**) loss tangent (tan δ) of PLA and different blends with PSt ratios of 30 wt%.

The behavior of the storage and loss moduli as a function of frequency is shown in Figure 2b,c, respectively. The storage modulus (G) of PLA in the low-frequency region is lower than its loss modulus (G"); such a discrepancy implies that PLA shows typical viscous characteristics. This relationship slightly changes after adding PSt and PEG to the blend. When the CA component is added, a significant increase in the storage modulus of the blend is observed in the low-frequency region. This increment indicates that the existence of CA can promote the transformation of PLA from a liquid-like state to a solid-like one [28]. The change in tan δ during frequency scanning is used to estimate the dispersion of the filler or the dispersed phase in the blended system. The tan δ of liquid-like materials linearly decreases with the increase in frequency. The changes in the tan δ of several samples with frequencies at 170 ◦C are displayed in Figure 2d. The finding shows that PLA has a typical liquid-like characteristic [29] which slightly changes after adding PSt and PEG. After adding CA, the tan δ value of PLA/PSt/CA blend and PLA/PSt/PEG/CA blend in the low-frequency region significantly decreases, which means that a gel-like substance that can help improve the compatibility of PLA and PSt may appear in the system. The frequency dependence of tan δ weakens, indicating that the blend tends to change from a liquid-like state to a solid-like one.

## *3.3. Crystallization Behavior of PLA*/*PSt*/*PEG*/*CA Blends*

PLA is a typical semi-crystalline polymer [30] whose mechanical properties are closely related to the thermal properties and crystallinity of a solid. The heat flow curves of several samples during the second temperature rise are shown in Figure 3 and Table 3. Compared with pure PLA, PLA/PSt blends have lower Tg and Tc, which may be due to the VOP entering the PLA matrix during the blending process destroying the crystallization of PLA. Secondly, the VOP attached to the surface of the starch blocks the interface between PLA and starch, inhibiting the heterogeneous nucleation [31] of PLA by starch. The results of analysis of mechanical properties and rheological behavior have explained the immiscibility of PLA and PSt. The crystallinity of PLA in the PLA/PSt blend calculated by formula (1) is reduced, which also confirms the reliability of the analysis. The Tg, Tc, and T m of PLA/PSt/PEG blends further decreased, which was mainly due to the plasticization of PLA by PEG. PEG can enter between PLA macromolecules during melt mixing and establish physical interactions such as hydrogen bonding or dipole-dipole interactions in between atoms. As a result, some of the rigid homogeneous PLA-PLA interactions were replaced by heterogeneous PLA-PEG interactions. This phenomenon enhanced the mobility of PLA and decreased the energy consumption during glass transition (decreased Tg) [32]. Compared with the PLA/PSt blend, the Tc and T m of the PLA/PSt/CA blend increased. This may be because the surface polarity of the starch changed after CA was introduced into the blend. The reason for the change may be that CA participated in the esterification reaction and formed a branched/crosslinked copolymer on the surface of the starch, which enhanced the interaction between the components of the blend [33]. Usually, Tg will increase as the interaction force between the components of the blend increases. However, the Tg of PLA/PSt/CA blends only increased slightly, which may be because CA tends to change the surface polarity of starch rather than acting on PLA. Moreover, the depolymerization of PLA by CA will destroy the crystallization of PLA and reduce Xc. Comparing PLA/PSt blends with PLA/PSt/CA blends, the result is that CA can slightly increase the Tg of the blends. However, when the PLA/PSt/PEG blend is compared with the PLA/PSt/PEG/CA blend, the result is that CA slightly reduces the Tg of the blend. This shows that when CA and PEG are contained in the blend, PEG may also participate in the esterification reaction in the melt-blending process. Some interesting facts have also been discovered from the perspective of crystallinity changes in Figure 4. After introducing PEG into the PLA/PSt blend, its crystallinity value increased by 4.32%; however, after introducing PEG into PLA/PSt/CA blend, its crystallinity value only increased by 2.05%. This may be because when the CA component is present in the system, CA restricts the dispersion of PEG in the PLA matrix, thereby attracting PEG to the CA side. Since PEG can plasticize PLA, the branched/crosslinked polymer containing PEG formed after the esterification reaction can more easily penetrate into the PLA matrix and form an amphiphilic bridging structure [34]. This structure is considered to be the "bridge" between PLA and starch, which can enhance the compatibility between PLA and starch. However, a part of the free branched/crosslinked polyester participates in the process of plasticizing [35] PLA and causes the Tg of the blend to decrease. The change of the interface structure between PLA and starch indicates that the compatibility of the two is increased. At this time, PSt can be regarded as a heterogeneous nucleating agen<sup>t</sup> of PLA and promotes the crystallization of the PLA matrix. The increase in e ffective nucleation sites in the blend enhances the interfacial adhesion between PLA and starch and allows the blend exhibit excellent toughness in the macroscopic view, and this has been confirmed from the mechanical performance test results.

**Figure 3.** Differential Scanning Calorimeter (DSC) heat flow curves of pure PLA and PLA/PSt, PLA/PSt/PEG, PLA/PSt/CA, and PLA/PSt/PEG/CA blends.

**Figure 4.** Surface morphology of several pretreated blend powder samples observed using SEM: (**a**) pure starch, (**b**) PSt, (**c**) PLA/PSt, (**d**) PLA/PSt/PEG, (**e**) PLA/PSt/CA, and (**f**) PLA/PSt/PEG/CA.


**Table 3.** Thermal properties of the materials.

## *3.4. Morphology of the Dispersed Phase*

On the basis of the above analysis, CA is an important factor in toughening PLA/PSt/PEG/CA blends. To observe the effect of CA and PEG on the blend, a separate dispersed phase of PLA/PSt/PEG/CA blends was obtained by dissolving and washing. The surface morphology of the dried dispersed phase was observed using SEM. The morphologies of several samples examined under 5k× magnification are shown in Figure 4. Pure starch (Figure 4a) appears as an irregular sphere with a diameter of approximately 10 μm and has many small pits on the surface. As shown in Figure 4b, the surface of PSt is smooth. The starch in the pretreated PLA/PSt blends (Figure 4c) blend has similar size and surface morphology to pure starch. This similarity indicates that the compatibility between PLA and starch is still poor. Figure 4d shows that the size of starch particles in the PLA/PSt/PEG blend after the pretreatment increases to some extent, which may be due to the plasticizing effect of PEG on starch. The presence of numerous fine pits on the surface of starch (Figure 4d) indicates that PLA and starch are not effectively entangled, and starch only acts as a large particle filler in the PLA/PSt and PLA/PSt/PEG blends to reduce their macroscopic mechanical properties. However, although the particle size of starch in the PLA/PSt/CA blend does not significantly change after the addition of CA, the corresponding surface morphology greatly varies and several tiny protrusions form on the surface of the starch granules (Figure 4e). This phenomenon causes the agglomeration of starch, which is also evident in Figure 4f. In addition, the starch particles in Figure 4f seem to be affected by the plasticization of PEG, and the size of the starch particles increases to some extent. The addition of PEG promotes the plasticization of the PLA and starch phases, whereas the addition of CA enhances the compatibility among the components of the blend. The change of the starch surface structure improves the interface compatibility between PLA and starch, thereby increasing the effective nucleation sites in the blend. At this time, PSt can be regarded as a heterogeneous nucleating agen<sup>t</sup> of PLA, which enhances the interfacial adhesion between PLA and starch. The blends show excellent toughness in the macroscopic view, which has been confirmed from the mechanical performance test results.

## *3.5. Crystallization of the Dispersed Phase*

The XRD patterns of pure starch and pretreated PLA/PSt/CA and PLA/PSt/PEG/CA powder samples are shown in Figure 5. The shape of the XRD diffraction peaks of the three samples does not significantly change, which means that the crystalline type [36] of starch in the samples does not vary. The crystallinity of the three samples was obtained by fitting the XRD data. The crystallinity of the samples containing CA decreases compared with that of pure starch because the former destroys the amorphous area of starch to a certain extent, thereby enhancing the connection between PLA and the starch molecular chains and the compatibility between PLA and starch. In addition, the crystallinity of the blend further decreases after adding PEG, which suggests that PEG not only acts on PLA, but also plasticizes starch [37] to a certain extent. During the blending process, the crystalline area of starch is partly destroyed by the PEG and CA. This phenomenon facilitates the plasticization of starch and enhances starch dispersibility in PLA, and thus improves the compatibility between PLA and starch.

**Figure 5.** XRD patterns of pure starch and pretreated PLA/PSt/CA and PLA/PSt/PEG/CA powder samples.

## *3.6. FTIR Analysis of the Dispersed Phase*

Chemical reactions may a ffect the surface morphology of starch, and this e ffect will influence the compatibility of the components of the blend. The dispersed phase was extracted from several blends, and FTIR spectroscopy was performed to analyze the changes in the compatibility among the components. The FTIR test results of several samples are presented in Figure 6. The PLA/PSt, PLA/PSt/PEG, PLA/PSt/CA, and PLA/PSt/PEG/CA blends are powder samples obtained through pretreatment. As shown in the figure, the C=O absorption peak near 1743 cm<sup>−</sup><sup>1</sup> can only be detected in VOP and PSt. No C=O absorption peak is observed in the FTIR spectra of the PLA/PSt and PLA/PSt/PEG powder samples after pretreatment. This result indicates that the PLA and VOP in the blend are completely removed, and PEG may display a plasticizing e ffect. However, C=O absorption peaks are detected in the PLA/PSt/CA and PLA/PSt/PEG/CA powder samples after pretreatment, which means that ester groups are still present in the powder samples at this time. CA participates in the esterification reaction during the blending process, and a branched/crosslinked copolymer forms. This branched/crosslinked copolymer is coated on the surface of the starch to prevent the removal of the ester group during the blending process. Another obvious change in peak position occurs near 2863 cm<sup>−</sup>1, which can be attributed to the C–H symmetric stretching vibration absorption peak in VOP and PSt. The respective absorption peaks of pretreated PLA/PSt and PLA/PSt/PEG powder samples at 1743 and 2863 cm<sup>−</sup><sup>1</sup> disappear, implying that no chemical changes occur during the blending process. Starch only acts as a large-sized filler, and PEG acts as a plasticizer. The immiscibility between PLA and starch decreases the macroscopic tensile toughness of the blend. The esterification reaction might occur in the PLA/PSt/CA and PLA/PSt/PEG/CA powder samples after pretreatment, which might improve the surface activity of starch and increase the compatibility between PLA and starch. Furthermore, the blend shows a significant increase in tensile and impact toughness.

**Figure 6.** FTIR spectra of vegetable oil polyols, starch, PSt, and pretreated PLA/PSt, PLA/PSt/PEG, PLA/PSt/CA, and PLA/PSt/PEG/CA blend powder samples.

## *3.7. Hydrophilicity Test of PLA*/*PSt*/*PEG*/*CA Blends*

The water absorption rates of several blends are shown in Figure 7. The addition of hydrophilic starch to PLA improves the water absorption rate of the blend materials. On this basis, this rate will be further improved by adding PEG or CA. These materials contain a large amount of hydrophilic -OH, which helps the blend to combine with H2O molecules to form hydrogen bonds. From the results shown in Figure 7, PEG in the blend system seems to have other functions besides acting as a plasticizer. The water absorption rates of the PLA/PSt/CA and PLA/PSt/PEG/CA blends sugges<sup>t</sup> that the hydrophilicity of the blend material decreases after the addition of hydrophilic PEG. This finding reveals that PEG may also participate in the esterification reaction along with CA during the reactive blending process. The polarity of PEG decreases and the content of hydrophilic groups decreases after the esterification reaction after the esterification reaction, resulting in a certain decrease in the water absorption rate of the blend. Not only that, under the combined action of PEG and CA, the interface between PLA and starch becomes more stable, and the interfacial gap becomes smaller, which results in starch not being able to be swollen fully. Secondly, while the crystalline and amorphous regions of starch are destroyed, PEG and CA will be dispersed near the active hydroxyl groups [38] of the starch, rendering these locations unable to accommodate more water molecules and reducing the water absorption of the blend.

**Figure 7.** Water absorption rates of pure PLA and PLA/PSt, PLA/PSt/PEG, PLA/PSt/CA, and PLA/PSt/PEG/CA blends.

3/\$

3/\$36W

3/\$36W3(\*

3/\$36W3(\*&\$

3/\$36W&\$

WLPHV

 

FRQWDFWDQJOHV˄e˅

˄ ˅

In view of the change of the hydrophilicity of the blends, the dynamic contact angles of several blends were analyzed. As shown in Figure 8, the contact angle of pure PLA is the largest in the initial state, indicating that PLA is relatively hydrophobic, which is determined by the structural properties of the macromolecular PLA itself. After introducing PSt, PEG and CA into PLA, the contact angles of the blends all decreased, which indicates that the blends are more hydrophilic. However, the contact angle of the PLA/PSt/PEG/CA blend increased compared with the PLA/PSt/CA blend, which is consistent with the conclusion of water absorption. This change further indicates that PEG may participate in the esterification reaction in the PLA/PSt/PEG/CA blend system.

**Figure 8.** Dynamic contact angles of water in the first 10 min on the surface of pure PLA and PLA/PSt, PLA/PSt/PEG, PLA/PSt/CA, and PLA/PSt/PEG/CA blends.

In addition, low-molecular-weight PEG acts as a plasticizer to PLA and exerts a certain plasticizing effect on starch. This feature of PEG allows the branched/crosslinked copolymer with PEG chains that formed after the esterification reaction to penetrate the interface between PLA and starch and act as a bridge between the two to increase their interaction force. Meanwhile, the branched/crosslinked polyester forms a highly stable coating on the starch surface, which facilitates the compatibilization between PLA and starch.
