Poly (lactic acid) (PLA) with a grade of 2003D, obtained from Nature Works, USA, has a melt flow rate of 6 g/10 min (190°C/2.16 kg) and a density of 1.24 g/cm³. The Poly (ε-caprolactone) (PCL) used is Capa^TM6800, purchased from Perstorp, Sweden, with a melt flow rate of 4 g/10 min (160°C/5 kg) and a density of 1.13 g/cm³. Glycidyl methacrylate (GMA), obtained from Alfa Aesar, Germany, has a density of 1.08 g/cm³ and a boiling point of 189°C. Nano calcium carbonate (NCC) with a mean particle diameter of 80 nm and a density of 2.65 g/cm³ was purchased from Shiraishi (Austria) and treated with stearic acid.

Surface treatment of nano calcium carbonate with stearic acid involves coating the particles with a layer of stearic acid. This process is commonly used to improve the dispersion of nano calcium carbonate in polymer matrices, as well as to enhance its compatibility with various polymer systems. The surface treatment is typically carried out using a solution of stearic acid in a solvent like ethanol or toluene. The nano calcium carbonate particles are then added to the solution and stirred or sonicated to ensure uniform coating of the particles with stearic acid. The solvent is then evaporated, leaving behind a thin layer of stearic acid on the surface of the nano calcium carbonate particles. The stearic acid coating can improve the dispersion of the nano calcium carbonate in polymer matrices by reducing the agglomeration of the particles and promoting better interaction between the filler and the polymer. This can lead to improvements in mechanical properties, thermal stability, and other performance characteristics of the composite materials.

To prepare materials for compounding, they were dried in a vacuum oven at 70°C for 16 h. The compounding process was performed using an internal mixer (Brabender Plasticorder W50) with a capacity of 60 cm³. The components were mixed simultaneously in a single step at 170°C and 50 rpm for 10 min (However, to prepare nanocomposites, NCC was added after both polymer components were poured in the brabender). The resulting mixture was then compression molded at 190°C to form sheets of 2 and 3 mm thickness for subsequent measurements.

For the sample preparation, three steps were taken:

Simple blends of PLA/PCL with 15, 25, and 35 wt.% of PCL were prepared to study their mechanical and thermal properties. The blend with 25 wt.% of PCL (P75) showed the best mechanical properties and was selected for further investigations.

Compatibilized PLA/PCL (75/25) blends with varying amounts of GMA compatibilizer (1, 2, 3, and 4 phr) were prepared and their properties were studied.

PLA/PCL nanocomposites with different amounts of NCC (5, 7, 9, and 11 phr) were made and their properties were investigated as well. (The reason why these composition were chosen for NCC was the previous studied that had shown the proportion of NCC in which filler aggregation happened was higher than 15wt.%). We started making sample from 3 phr which was about 2 wt.% and TEM images showed no aggregate. Then we increased amount of NCC to 11 phr (9.7 wt.%) and there was no aggregated particle). So, the aggregate formation can be negligible, evet in sample containing 11 phr of NCC).

The Table 1 provides a summary of the names and compositions of the neat polymers, blends, and nanocomposites. For instance, P75G2 represents a nanocomposite sample containing 75 wt.% of PLA, 25 wt.% of PCL, and 2 phr of GMA.

Tensile properties and impact strengths are presented in Table 2 and reveal that incorporating ductile PCL into PLA improved its toughness. For example, 15 wt.% PCL increased the impact strength of neat PLA from 21.6 to 38.2 J/m (P85). Brittle fracture observed in neat PLA indicated a lack of yield phenomenon (Bhatia et al., 2007). However, the presence of PCL introduced a shear yielding mechanism that enhanced the toughness of PLA. Interestingly, the sample containing 25 wt.% of PCL exhibited the highest impact strength and elongation at break among the blends. This maximum is likely attributed to phase separation between PLA and PCL at higher PCL loadings, resulting in poor interfacial adhesion and a decrease in mechanical properties in the blend containing 35 wt.% of PCL (P65), as supported by previous studies (Yeh et al., 2009; Broz et al., 2003). The PLA/PCL blend demonstrates a lower critical solution temperature (LCST), in which the optimum PCL composition for the initiation of phase separation was estimated to be around 36% by volume. Consequently, the PLA/PCL blend with a ratio of 75/25 was selected for further experimentation.

To improve compatibility of PLA and PCL, GMA was employed as a reactive compatibilizer in the blend. Notably, the addition of 1 phr of GMA to P75 increased the elongation at break from 35.1% to 76%. This improvement can be attributed to enhanced compatibility between PLA and PCL, as well as the uniform distribution of PCL droplets facilitated by GMA because of the reduction of interfacial tension (Guerrica-Echevarría et al., 2000; Ayirala and Rao, 2006). The positive impact of GMA on the mechanical properties of polymer blends has been reported previously (Chee et al., 2013; Juntuek et al., 2012; Kumar et al., 2010). The results presented in Table 2 reveal that increasing content of GMA up to 3 phr further improves the elongation at break and impact strength of the samples. However, exceeding the optimum amount of compatibilizer (3 phr) had an adverse effect on the mechanical properties, potentially due to phase separation. This behavior signifies that an extra quantity of compatibilizer within the blend did not undergo any reaction; therefore, its influence on softening of polymer phases decreased the mechanical properties. In the regions where this unreacted compatibilizer existed, the movement of polymer chains accelerated, as the induced free volume within the system increased. Moreover, a significant rise in elongation at break of polymer blends is attributed to the lubricating effect of GMA, and thus improved fluidity of molecular chain (Aliotta et al., 2020).

In a previous paper, incorporating more than 3 wt.% of GMA into PLA/PCL blends resulted in a decrease in flexural modulus, elongation at break, and impact strength (Chee et al., 2013). Also, Zhang et al. (2009) found that increasing the T-GMA content initially improved the elongation at break, but subsequent increases ultimately reduced this property.

In addition to investigating the effect of NCC on PLA/PCL blends, the impact of different amounts of NCC on uncompatibilized PLA/PCL (75/25) blend was examined. A decrease in tensile strength and elongation at break was seen as the quantity of NCC increased; however, the tensile modulus and impact strength of the samples were increased. The decline in tensile properties can be attributed to reduced chain mobility and the formation of voids during the tensile test, which arose from the discontinuity in load transfer across the chains. Despite this, the presence of NCC led to an increase in the impact strength of the P75 sample, potentially due to crack growth prevention. Notably, the use of 9 phr NCC yielded the most preferable mechanical properties, as the higher concentration of NCC increased rigidity. Liang et al. (2015) also observed that the tensile yield strength and un-notched Izod impact strength of PLA/PCL/NCC nanocomposite experienced minimal changes as the weight fraction of filler particles exceeded 3 wt.%. Figure 1A depicts the stress-strain plots of tensile test and Figure 1B shows four bar charts including mechanical properties of the selected superior samples from each group. As can be seen in Figure 1A, PCL does not change the brittleness of PLA because of their incompatibility. However, GMA swaps the fracture mode of PLA/PCL from brittle to ductile due to stress transfer along the interface. However, the addition of NCC in uncompatibilized blend results in a more brittle behavior compared to neat PLA.

The glass transition temperature (T_g) can be used to assess compatibility of polymer blend components. A single T_g indicates a blend that is a single phase. Additionally, variations in cold crystallization temperature (T_cc), melting temperature (T_m), and percentage of crystallinity (X_C) are often indicative of the interactions between different components (Li and Huneault, 2011). The obtained DSC results are summarized in Table 3 and selected thermographs are shown in Figure 2. To ensure accurate assessments, the thermal data was analyzed based on the second heating run, which eliminates any potential influence of process history.

The neat PLA demonstrates an orthorhombic structure with a crystallization peak at T_c = 132°C and a melting peak at T_m = 157.6°C. On the other hand, neat PCL only exhibits a melting peak at T_m = 55.2°C. P75 blend exhibits lower degree of incompatibility between PLA and PCL among other blend samples. For samples without NCC or GMA, T_cc values for all blends are lower than that of neat PLA, with the P75 blend having the lowest value, suggesting the nucleation effect of PCL. Therefore, the P75 sample composition is selected for subsequent samples.

The impact of varying amounts of GMA compatibilizer on the thermal properties of P75 blend was investigated. GMA decreases the T_g and T_m of P75 due to an increase in polymer chain mobility. This is attributed to the reactive epoxide group present in GMA, which can react with both PLA and PCL, facilitating the creation of covalent bonds that bridge the two polymer phases. Consequently, this covalent bonding strengthens the interfacial adhesion and promotes efficient stress transfer between the polymers, resulting in improved chain mobility (Kong et al., 2023). Additionally, as the T_g of GMA is lower than both PLA and PCL, its incorporation into the blend leads to an overall reduction in the blend’s T_g. This decrease in T_g enhances chain mobility, enabling polymer chain rearrangement and reducing chain entanglements.

In addition, the cold crystallization of PLA in the blend was accelerated after addition of GMA as evidenced by a decrease in T_cc and an increase in ΔH_m, in agreement with the findings of Sugih et al. (2009). GMA reduces the size and increases the number of PCL domains and the total area of PLA/PCL interface increases. As a result, the greater the number of PCL domains, the more effective it is as a nucleation agent. In addition, the acceptable effect of GMA on thermal properties of blends took place in the sample containing 3 phr of GMA. The limited interaction capacity between the polymer and compatibilizer led to unreacted GMA, resulting in incompatibility in the P75G4 sample (Chee et al., 2013).

Furthermore, the addition of NCC to the P75 increased the T_g and T_m. The T_cc of P75 decreased significantly, indicating that NCC enhances the rate of PLA crystallization. This can be attributed to the nucleating effect of the nanoparticle. NCC in 9 phr enhanced thermal properties, as evidenced by a distinct increase in the percentage of crystallinity compared to P75 (Balakrishnan et al., 2010; Salehiyan et al., 2015). Besides T_cc changes, it was understood that the P75N9 and P75N11 samples exhibit two distinct melting peaks. According to Figure 2D, low temperature melting peak (T_m1) should be associated with the fusion of the crystals grown by normal primary crystallization, whereas the high-temperature melting peak (T_m2) is related to the melting peak of the most perfect crystals after rearrangement during the second heating process (Tábi et al., 2010; Huang et al., 2009). This phenomenon happened because of the increment in NCC composition, which prompted reduction of chain mobility and prevented the lamellas from being completed. Therefore, some imperfect crystals with lower lamella’s thickness are formed; consequently, double peaks for T_m are seen. Moreover, it is found that the use of 11 phr of NCC in the P75N11 did not make favorable changes in T_cc and X_c compared to the P75N9. It seems that the nucleating effect of NCC is restricted due to nanoparticle aggregation. It is important to take into consideration that the process of crystallization is divided into two stages: nucleation and growth of lamellas (De Santis et al., 2011). The addition of both NCC and GMA enhances nucleation, but the effectiveness of NCC is superior. However, NCC hinders the growth of lamellas by restricting molecular chain movement and resulting in the formation of imperfect lamellas. On the other hand, GMA increases chain mobility and facilitates the arrangement of chains within the lamellas. So, GMA counteracts the formation of imperfect crystals and alters the characteristics of the samples at their melting point. (GMA does not have T_g as it is a material with molecular weight of 142 g/mol).

In Figure 3, SEM images of P75, P75G3, and P75N9 samples are displayed. All samples exhibit droplet-matrix morphology. The circular structures in the dispersed phase represent the PCL domains. Table 4 presents the results of particle size determined using ImageJ software. In Figure 3A, the micrograph of P75 suggests poor interfacial adhesion between the PLA matrix and PCL domains, indicating complete incompatibility of these two polymers, which is supported by previous studies (Yeh et al., 2009; Broz et al., 2003; Rodriguez et al., 2006). A similar structure is observed in P75G3 (Figure 3B), with smaller PCL droplets compared to those in P75, showing improved compatibility of polymers. The positive effect of GMA on modifying the interfacial adhesion of PLA/PCL was also reported (Chee et al., 2013). In addition, SEM analysis demonstrated that the incorporation of GMA into PLA/PCL resulted in enhanced compatibility by promoting multiple crazing and cavitation mechanisms. The same conclusion according to SEM analysis was made as an evidence of improved compatibility between maleated SEBS and PLA/cellulose nanocrystal (Diani et al., 2016).

In the case of the P75N9 sample (Figure 3C), the size of PCL domains is smaller compared to P75, but larger than those in P75G3. These results suggest that NCC can also act as a compatibilizer for the PLA/PCL. This incident may be due to the steric hindrance of nanoparticles in the merging of PCL droplets.

The main goal of the DMA is to study T_g and investigate the molecular interactions within the amorphous and crystalline regions of polymers (Hou and Qu, 2019). As can be seen in Figure 6, nanoparticles in P75N9 led to a higher modulus compared to P75G3. However, the modulus of P75N9 experienced a quicker decrease as the temperature increases. This can be attributed to the rearrangement of NCC particles and transformation of nanoparticle network. The introduction of nanoparticles served to restrict the mobility of polymer chains, thus impeding their movement. Consequently, this resulted in a decline in molecular motions, which in turn contributed to a rise in the overall modulus of the nanocomposite. As the temperature increased, the role of nanoparticles as physical barriers became less prevalent, hindering the deformability and movement of polymer chains. Thus, this led to a significant decrease in modulus. Generally, nanoparticles often possess a considerable surface area and can form robust interactions with polymer chains. These interactions facilitate load transfer between the nanoparticles and the matrices, boosting their bond formation. However, due to the increased temperature, polymer internal chain segments tend to detach from the surface of nanoparticles due to higher relative motional freedom (Tham et al., 2022). Moreover, because of conformational changes at higher temperatures, well-dispersed nanoparticles become less abundant within the polymer matrix, giving rise to formation of defects and regions of lower stiffness. Differences between steric (entropic) states with nanoparticle, does not allow polymer chains to replenish nanoparticles previous position completely. These localized regions characterized by reduced modulus contribute to an overall decline in modulus during DMA. Lastly, as the temperature increases, nanoparticles tend to aggregate or cluster, particularly if they possess a strong mutual attraction, such as NCC particles. These clusters of nanoparticles are susceptible for stress concentration, leading to a drop in the overall modulus (Banerjee and Ray, 2021).

In contrast, the glass transition region of P75G3 appeared in mild slope due to the increased compatibility and notable interaction between PLA and PCL compared to the P75N9. Additionally, the P75G3 exhibited a higher modulus and length in the rubbery region, which can be attributed to the effective performance of GMA in increasing chain extension, enhancing intermolecular entanglements, and promoting mutual attractions among the ends of PLA and PCL chains. (These are solid evidence to prove the formation of longer chains in P75G3).

The molecular weight distribution was determined using GPC with the Mark-Houwink-Sakurada parameters provided by Huang et al. (2003). The average molecular weight of polylactic acid is approximately 218,000 g/mol, while the average molecular weight of polycaprolactone is around 75,000 g/mol. In addition, weight-average molecular weight (M_w) and polydispersity index (PDI) were measured for P75, P75N9 and P75G3 (Figure 7). The results showed the desirable effect of GMA on molecular weight and destribution of P75. Surprisingly, M_w of P75N9 was significantly lower than that of the P75G3. Previous studies have also shown that the presence of GMA as a chain extender can increase the average molecular weight of the polymer formed (Pasanphan et al., 2021; Javadzadeh et al., 2020). The difference in average molecular weight between P75N9 (68,319 g/mol) and P75G3 (83,522 g/mol), and greater PDI value in P75G3, are attributed to the presence of longer and shorter polymer chains. The introduction of GMA led to longer and lower molecular weight chains, as well as a wider distribution. However, the presence of nanoparticles resulted in shorter chains with a narrower distribution.

GMA led to intermingling of adjacent chains, resulting in a higher molecular weight for P75G3 compared to P75N9. This also made molecular weight distribution wider. However, NCC nanoparticles might interfere with the chain extension process, potentially resulting in their scission during mixing (Castilla-Cortázar et al., 2019). This could be due to collision of the nanoparticles and chains.

Small amplitude oscillatory shear (SAOS) test were condocted to study the rheology of samples, with the aim of understanding the relationship between their molecular structures and flow properties. The plots of storage modulus (G′) and complex viscosity (η*) as a function of angular frequency are presented in Figures 8, 9, respectively.

Figure 8 demonstrates that both neat PLA and PCL exhibit a terminal behavior, where the storage modulus G′ is proportional to the angular frequency ω². Additionally, the storage modulus of PCL is significantly lower than that of PLA across all frequencies. When PCL is added to PLA, the storage modulus increases at low frequencies, while the storage modulus of P75 is lower than that of pure PLA at medium and high frequencies, indicating a weak interfacial interaction between polymers. However, GMA leads to higher storage modulus compared to uncompatibilized blend. Nevertheless, beyond 1 rad/s, the storage modulus of P75G3 becomes smaller than that of PLA, suggesting partially enhancement the interfacial adhesion between the polymers. Storage modulus of the nanocomposite (P75N9) is notably higher than that of the other samples, aligning with the findings of previous research on polymer blends incorporating nanoparticles (Najafi et al., 2012; Zachariah et al., 2014). Moreover, the appearance of G′ plateau at low frequencies in the nanocomposite sample indicates the formation of a three-dimensional network of nanoparticles, resulting in a solid-like behavior. The slope of G′ versus ω plots, denoted as (N), was investigated in the range of 0.01–0.1 rad/s and is shown in Table 5. The results reveal that the blend has a lower initial slope of G′ than PLA, showing the formation of entangled structures in PLA/PCL melts. GMA has a minimal effect on this slope, as the low molecular weight compatibilizer has a negligible impact on increasing the entanglement density. Furthermore, the outset slope of the G′ versus ω plot in P75N9 is much lower than that of neat polymers and compatibilized blends, suggesting nanoparticle aggregation (DEFENG WU et al., 2007; Huitric et al., 2009).

Additionally, the storage modulus plots of the blends exhibit a secondary plateau attributed to the shape relaxation of the PCL dispersed phase PCL (Elias et al., 2007). GMA widens this plateau, because of improved compatibility between PLA and PCL. The frequency dependence of complex viscosity for all samples is depicted in Figure 9. It is evident GMA influences the rheological behavior of P75, leading to an increase in complex viscosity due to enhanced connectivity between the PLA and PCL chains. Similarly, the incorporation of NCC particles into PLA/PCL blends significantly alters their rheological properties, resulting in a notable increase in complex viscosity, especially at low frequencies. Some researchers corresponded the observation of this plateau to the location of nanoparticles at the interface between two polymer phases (Vinckier et al., 1999a; Vinckier et al., 1999b). The outcomes of previous study demonstrated that a huge increase in the quantity of NCC particles present within the interface triggered the formation of aggregates along the polymer interface (Liu et al., 2024). This consequently impeded a consistent shape relaxation within the polymer interface owing to the substantial difference in the degree of molecular mobility between the polymer chains in the bulk and the interface. Subsequently, this phenomenon leads to the formation of a distinct separated phase in the proximity of the interface.

Additionally, it is obvious that interactions between nanoparticle and polymer, as well as particle-particle interactions, making a substantial increase in the complex viscosity of nanocomposite sample, particularly at low frequencies. High frequency measurements revealed that all samples exhibited pseudoplastic behavior, but neat PCL only displayed Newtonian behavior within a limited frequency range (up to 10 rad/s), on the other hand, P75N9 has a smaller frequency range for Newtonian behavior. The viscosity reduction after the Newtonian region, as determined by the degree of shear-thinning (n), known as power-law index, was more pronounced in the nanocomposite sample compared to other samples (Wagener and Reisinger, 2003). η = A ω n (2)

The value of n was calculated using Eq. 2 in its logarithmic form, and the results are presented in Table 5. It was observed that the P75N9 exhibit more intense shear-thinning behavior compared to the other samples. At higher frequencies, this shear-thinning behavior becomes more evident due to the dispersion of nanoparticle clusters along the direction of shear.

Rheological study of the samples was developed using the Palierne model. This model provides valuable information on the interfacial properties of incompatible PLA/PCL blend and nanocomposite (Lacroix and Carreau, 1997). In the Palierne model, complex modulus of matrices and polymer interface have been utilized to predict the complex modulus of samples (Genoyer, 2018). The aim of this study is to elucidate the relaxation mechanisms based on polymer chain movements, droplet shape evolution, and relaxation caused by Marangoni stresses (Jacobs et al., 1999). The plots obtained from SAOS test were compared with the plots generated by the Palierne model to assess their overlap. This model is expressed as Eqs 3, 4: G * ω = G m * ω 1 + 3 φ Η ω 1 − 2 φ Η ω (3) H ω = 4 α R 2 G m * ω + 5 G d * ω + G d * ω − G m * ω 16 G m * ω + G d * ω 40 α R G m * ω + G d * ω + 2 G d * ω + 3 G m * ω 16 G m * ω + G d * ω (4) where α represents the interfacial tension, ϕ represents the volume fraction of dispersed phase, R represents the dispersed particle radius, and represent the complex modulus of the dispersed phase and matrix, respectively.

The simplified Palierne model with the approximation of Graebling et al. (Graebling et al., 1993) was used, where the surface dilatation modulus and surface shear modulus were ignored. The fitting program was assigned an initial value of 100 for α/R, and the optimum value was found to be 76.31 N/m². Using Table 4, the interfacial tension was calculated to be 0.29 mN/m for PCL droplets with an average diameter of 7.6 μm for P75. Also, P75N9 and P75G3 were compared with R_v extracted from Table 4 and the same interfacial tension (Fenni et al., 2020).

As shown in Figure 10, the Palierne predictions for P75N9 and P75G3 showed a discrepancy in the actual value. It is assumed that in the P75N9, the NCC particles are located both at the interface and in polymer phases, which does not mean that particles saturated the interface, but rather that excessive amounts of NCC were located in the droplets, hindering their shape relaxation (Liao et al., 2020). In the case of sample P75G3, there was more similarity between the fitted plot and the experimental one. Therefore, the effect of GMA on the compatibility of polymers is more remarkable, partially increased their interfacial adhesion during the mixing procedure.