Properties of Styrene–Maleic Anhydride Copolymer Compatibilized Polyamide 66/Poly (Phenylene Ether) Blends: Effect of Maleic Anhydride Concentration and Copolymer Content

Polyamide 66 (PA66)/poly (2,6-dimethyl-1,4-phenylene ether) (PPE) blends with a ratio of 50/50 (w/w) were produced by a twin-screw compounder. The immiscible blends were compatibilized using two different styrene–maleic anhydride copolymers (SMA) with a low (SMAlow) and a high (SMAhigh) maleic anhydride (MA) concentration of 8 and 25 wt%, respectively. Furthermore, the SMA content was varied from 0 to 10 wt%. The influence of MA concentration and SMA content on the morphological and thermomechanical properties of PA66/PPE blends was investigated. Herein, we established correlations between the interfacial activity of the SMA with blend morphology and corresponding tensile properties. A droplet-sea to co-continuous morphology transition was shown by scanning electron microscopy to occur between 1.25 and 5 wt% in the case of SMAhigh. For SMAlow, the transition started from 7.5 wt% and was still ongoing at 10 wt%. It was found that SMAlow with 10 wt% content enhanced the tensile strength (10%) and elongation at break (70%) of PA66/PPE blends. This improvement can be explained by the strong interfacial interaction of SMAlow within the blend system, which features the formation of nanoemulsion morphology, as shown by transmission electron microscopy. Very small interdomain distances hinder matrix deformations, which forces debonding and cohesive failure of the PPE phase as a “weaker” main deformation mechanism. Due to a lack of interfacial activity, the mechanical properties of the blends with SMAhigh were not improved.


Materials
A commercially available PA66 compounding grade, PPE powder, and SMA with various MA concentrations were used as provided. Most of the relevant material properties are shown in Table 1. The weight-averaged molecular weight (M w ) and polydispersity measurements of PA66 and PPE were performed via gel permeation chromatography (GPC). An instrument having four PSS-SDV gel columns (particle size = 5 µm) with a porosity range from 102 to 105 Å (PSS, Mainz, Germany) using a nonselective refractive index detector (Shodex, Techlab, Japan). Hexafluoroisopropanol (HFIP) and chloroform (CHCl 3 ) were used as eluents for PA66 and PPE, respectively. The eluent flow rate was set at 1.0 mL/min. The calibrations for PA66 and PPE were done with poly(methyl methacrylate) and narrowly-distributed polystyrene (PS) for PA66 and PPE, respectively.

Processing of Blends
Prior to processing, PA66 was dried overnight by using a dry-air granulate dryer (TLE 100, Gerco Technik GmbH, Enningerloh, Germany) at 80 • C. All materials were compounded at 270 • C and 300 rpm in a co-rotating twin-screw compounder (ZSK 26 MCC, Coperion GmbH, Stuttgart, Germany). Pellets were obtained by strand pelletizing after passing a water bath. For binary blends of PA66 and PPE, a single-step compounding was applied. The ternary blends of PA66, PPE, and SMA were melt-blended via two-step processing, where PA66/SMA blends were reactively compounded in the first stage. After overnight drying at 80 • C, the pellets were melt-blended with PPE.
The PA66/PPE blend ratio was set constant at 1:1 (w/w). Based on this, the SMA contents were varied from 1.25, 5, 7.5 to 10 wt%.
Specimens of overnight dried blends were prepared by injection molding (Arburg Allrounder 470H 1000-170, Arburg GmbH, Loßburg, Germany) with different geometries for further characterization. A nozzle and mold temperature of 290 • C and 100 • C and a cooling time of 20 s were applied.

Dynamic-Mechanical Analysis (DMA)
For the investigation of thermomechanical properties, a Gabo Eplexor 500N (NETZSCH-Gerätebau GmbH; Selb, Germany) DMA was used in tensile mode. Oscillatory stress (2.5 MPa) was applied at a frequency of 1 Hz while heating from 25 to 255 • C at a heating rate of 2 K/min. For the evaluation of the glass transition, tan δ values were plotted against temperature, wherein the peak values were considered. Each measurement was repeated three times to minimize the experimental errors.

Morphological Characterization
Morphological analysis was done via a field-emission scanning electron microscopy (FESEM) Zeiss LEO 1530 (Zeiss NTS GmbH, Oberkochen, Germany) at an acceleration voltage of 3 kV. The surfaces of cyrofractured tensile bars were etched with CHCl 3 to selectively dissolve the PPE phase.
Further analysis was carried out via transmission electron microscopy (TEM) at an acceleration voltage of 200 kV using a Zeiss EM922 OMEGA (Zeiss NTS GmbH, Oberkochen, Germany). Ultrathin sections (approximately 60 nm) were prepared from injection molded tensile bars using an ultra-microtome (Leica EM UC7, Leica Microsystems GmbH, Wetzlar, Germany) equipped with a diamond knife. The ultrathin sections were stained with ruthenium tetroxide for 15 min in order to enhance the contrast between the two major phases. The number-averaged PPE domain sizes were calculated from 100 droplets considering the largest diameter for each domain, as the shapes were rather ellipsoid. Here, we assume that the cuts have gone through the middle of each domain.
Fractographs after tensile testing were taken at an acceleration voltage of 3 kV. Representative tensile bars, with values closest to the average were sputtered with platinum (1.3-2 nm thickness) prior to the measurements.

Mechanical Characterization
Tensile testing was performed using a universal testing machine (Zwick Z020, ZwickRoell GmbH & Co. KG, Ulm, Germany) equipped with an extensometer. The measurements were conducted according to ISO 527-2 using 1A type specimen [47]. Since PA66 is hygroscopic, all samples were dried overnight at 80 • C under vacuum and subsequently vacuum-sealed to guarantee the absence of humidity. The samples were taken out of the sealed bags prior to the measurements.

Miscibility of SMA in PPE
To evaluate the miscibility of two polymers various methods, such as differential scanning calorimetry (DSC), DMA, or SEM/TEM can be applied. In DMA measurements, tan δ plots depict an easy method for the determination of thermal transitions. For non-miscible binary blends, two distinct peaks are visible, indicating independent transitions of each polymer. For miscible polymer blends, the two peaks coincide to obtain a single signal in between the two individual signals depending on the blend ratio [48]. For interacting polymers, either both peaks approach each other or one of the peaks reveals a shift. In Figure 1a,b the tan δ versus temperature plots are given for SMA low and SMA high , respectively. Starting from 5 wt% SMA low content (Figure 1a), a shift of the PPE glass transition temperature (T g ) at 219 • C is seen. At 10 wt%, a maximum shift of −4 • C is observed. This indicates an interaction between SMA low and PPE; however, it cannot absolutely be stated that SMA low is (partially) miscible in the PPE phase. The occurrence of the SMA peak at 130 • C agrees with the literature [49]. For SMA high (Figure 1b), no significant shift of the PPE peak is visible, indicating that neither an interaction nor a miscibility with PPE exists. The SMA high peak at 155 • C appears higher than the peak of SMA low , which is given by its higher concentration of MA.
Furthermore, it is observed that the T g of PA66 is shifted to lower values when either of the SMA is added. Since the reaction of anhydrides and amines eliminates water, PA66 is likely to hydrolyze. As a result, local chain scission of the PA66 lowers the M w and thus the T g [50]. The T g signals of the reference and 10 wt% SMA low and SMA high are given in Table 2. According to Table 2, higher MA concentrations result in a more pronounced peak shift in the PA66 signal. The MA concentration is proportional to the amount of water released by the anhydride-amide reaction, thus leading to a higher degree of chain scission.
To approve the DMA results and further clarify the miscibility of both SMA in the PPE, TEM micrographs were taken from PPE/SMA binary blends with a PPE/SMA ratio of 82/18 (w/w) as shown in Figure 2a  For SMA low (Figure 2a), no phase contrast, i.e., no phase separation is visible, which can be explained by a complete miscibility of SMA low with PPE. Figure 2b indicates a distinct phase separation of the SMA high (minor phase, shown in light gray) and PPE (major phase, shown in dark gray), resulting in a droplet-sea morphology. This strong phase separation is induced by the mismatch of the two components due to the high polarity of SMA high , leading to elongated large SMA high domains. These domains (diameter range from 400 to 1200 nm) result in a non-transparent binary blend.  According to Figure 3, up to 5 wt% SMA low blends show droplet-sea morphology. For 1.25 wt%, a qualitative reduction of the PPE domain size distribution is visible; however, it is coarsening beyond the distribution of the neat system for 5 wt% SMA low . Interestingly, a further increase in SMA low content to 7.5 wt% and 10 wt% cannot reduce the domain size of the PPE phase. The coarsening of the PPE domains has a maximum at 7.5 wt% SMA low , as for 10 wt%, a finer morphology is seen.

Blend Morphology
The irregular domains induced by 7.5 and 10 wt% SMA low indicate an incomplete transition from droplet-sea to co-continuous morphology. Figure 4 exhibits the SEM micrographs of PA66/PPE blends with 1.25, 5, 7.5 and 10 wt% SMA high content together with the reference. In contrast to SMA low , an earlier droplet-sea to co-continuous transition for SMA high between 1.25 and 5 wt% starts ( Figure 4). As observed for SMA low , further SMA high addition cannot decrease the domain size of the PPE phase. For 7.5 wt% SMA high , the transition proceeds and finishes at the maximum concentration of 10 wt% SMA high . As seen for SMA low , a coarsening of the PPE phases is obvious for SMA high blends, with a maximum at 5 wt%. A further increase of SMA high content results in a decrease of the PPE domain sizes and a more homogeneous structure.
The earlier transition of SMA high is explained by the shift of the viscosity ratio of PA66/PPE to lower values. The viscosity ratio λ stated by Utracki is given in Equation (1), where η d is the viscosity of the dispersed phase (PPE) and η m is the viscosity of the matrix phase (PA66) [2].
For a constant blend ratio of 50/50 PA66/PPE, a large λ value is calculated from Equation (1) due to the high intrinsic viscosity of PPE and low viscosity of PA66, which justifies the visible droplet-sea morphology with PPE domains dispersed in a PA66 matrix. With the introduction of SMA high into PA66, the viscosity of the binary blend (PA66/SMA high ) increases. With the addition of a sufficient amount of SMA, λ approaches a value close to 1, where equally viscous polymers tend to form co-continuous structures during melt blending.
As shown in Figures 3 and 4 both SMA low and SMA high lead to droplet-sea morphology for 1.25 wt%. A co-continuous morphology is only observed for 10 wt% SMA high with a transition between 1.25 and 7.5 wt% and co-continuous morphology for 10 wt% SMA content. To validate the morphological interpretations via SEM, TEM micrographs of the reference and blends with 1.25 and 10 wt% SMA (low and high) are shown in Figure 5. As expected, the blends with 1.25 wt% SMA (low and high) possess a droplet-sea morphology, while the blends with 10 wt% SMA contain irregularly shaped PPE domains. From Figure 4, we know that the 10 wt% SMA high compatibilized blend forms a bi-continuous phase separation.
Interestingly, with 1.25 wt% SMA (high and low), changes within the PA66 matrix are noticeable and even more pronounced at 10 wt%.
The apparent changes within the PA66 phases are highlighted by TEM micrographs at higher magnifications for 10 wt% SMA (low and high) given in Figure 6. Herein, sub-micron sized black spots are distributed all over the matrix having a diameter of 100 nm and smaller in size for both SMA. The size distribution of the small inclusions together with the PPE domain size distribution of the reference without SMA is given in Table 3. Table 3. Domain size distribution analysis of SMA compatibilized PA66/PPE ternary blends based on the TEM micrographs from Figure 6.

PA66/PPE Size Distribution of Matrix Inclusions
Reference 1027 ± 336 10 wt% SMA low 109 ± 31 a 10 wt% SMA high 24 ± 8 a a Domain size distribution of (un-)swollen SMA-g-PA66 micelles in the PA66 matrix. Table 3 shows that the nano-sized inclusions are approximately 10 times smaller than the large PPE domains. Interestingly, the inclusions with SMA low are four times larger compared to the inclusions of SMA high . As already discussed in Section 3.1, SMA high is not miscible with PPE due to its high polarity. This allows us to conclude that the matrix inclusions consist of SMA high -g-PA66 copolymer and possibly unreacted SMA high micelles. In contrast, the nano-inclusions of the ternary blend with SMA low seem to be swollen. We propose that these micelles have a core-shell like structure with either a SMA low -g-PA66 or unreacted SMA low shell and a PPE core. The balanced polarity of the SMA low enables a strong interfacial interaction within the PA66/PPE blend. With sufficient SMA low content, all interfaces between PA66 and PPE are saturated, and thus, the interfacial tension of the blend system is minimized. Consequently, the surface roughening of PPE followed by pinch-offs occurs, enabling SMA low to diffuse to the newly generated interfaces. The phenomenon of micelle formation and emulsification was already described for other blend systems by several work groups [51][52][53].

Tensile Properties
The tensile properties of SMA compatibilized PA66/PPE blends are shown in Figure 7a-c wherein the Young's modulus (E) is displayed together with the tensile strength (σ m ) and elongation at break (ε b ), respectively. In comparison to the reference blend, Figure 7a shows a reduction in Young's modulus with 1.25 wt% SMA low (2360 MPa), whereas a recovery is seen with a maximum at 5 wt% (2620 MPa). The further incorporation of SMA low causes lower modulus values of 2540 MPa (10 wt%), which still is in the range of the reference. For SMA high , the modulus is found to be independent of the SMA content, showing no significant change.
For σ m (Figure 7b) and ε b (Figure 7c), the influence of SMA content is more pronounced. For SMA low , 1.25 wt% leads to deteriorated properties due to the disordered interfaces between PA66 and PPE, increasing local stress concentrations. However, we would expect better mechanical performance as the PPE domain sizes decrease (Figure 3), indicating a reduction of interfacial tension and successful compatibilization [25]. The further addition of SMA low results in an increase of σ m up to 11 % for 7.5 wt%, facing a plateau with no further increase at 10 wt%. It seems that the occurring morphology transition at 5-10 wt% (Figure 3) competes with the compatibilization effect of SMA low , resulting in only moderate improvements. With 1.25 wt% SMA low , the elongational properties of the blend do not differ from the reference. Starting from 5 wt%, a constant increase in elongation at break is observed, reaching its maximum at 10 wt% with a total increase of 70% compared to the reference. From Section 3.2, it is known that the morphology transition from droplet-sea to co-continuous structures happens at 5 and 10 wt% SMA low . As for tensile strength results, a change in morphology seems to overpower the compatibilization effect, leading to detrimental tensile properties compared to a clear droplet-sea type of phase separation. This phenomenon is also valid for the results of SMA high compatibilized blends. For 1.25 wt% SMA high , values of σ m and ε b are within the range of the reference. With the transition to co-continuous structures, a decrease in both σ m and ε b is seen at 5 wt% SMA high . Interestingly, the partial recovery of both values is evident for higher amounts of SMA high . These results agree with the findings of [34,35], where a similar behavior of SMA with 8 wt% MA concentration was observed for PA6/PPE. It is confirmed that SMA low reveals saturation content between 5 and 10 wt% where no further improvement of tensile properties is reported. SMA with high MA concentration (21.8 wt%) was shown to steadily increase the tensile properties with amounts up to 10 wt% [39]. It is noteworthy that the researchers used a PA6/PPE blend ratio of 70/30 where a droplet-sea structure was achieved independent of the SMA content. This led to a basic understanding of the effect of SMA as a compatibilizer without any further influences, such as the change in morphology. Differently, in our systems, the change in morphology suppresses the combatibilizing effect of both SMA (low and high), leading to moderate improvements (for SMA low ) or even worsening (SMA high ) in tensile properties.

Fracture Analysis
In order to correlate the tensile properties with the fracture surface, SEM analysis of the blends after tensile testing was performed. SEM fractographs of 1.25 wt% and 10 wt% SMA low and SMA high are displayed together with the reference PA66/PPE blend in Figure 8a-e.
The reference (Figure 8a1,a2) shows a rough fracture surface with strong crack deflections. No plastic deformation is seen, as PA66 is rather brittle when in a dry state. The deformation behavior is predominated by pull-outs of the PPE droplets having generally insufficient interfacial bonding to the PA66 matrix. Occasionally, bound PPE domains with a low degree of plastic deformation are observed. Whenever PPE domains are elongated, matrix deformations coexist at the interface expressed by fibrillation (white arrows). Therefore, we propose that the PA66 is able to reactively couple to the hydroxyl-terminated PPE polymer. The covalent bonds allow good energy dissipation at the interfaces by debonding and fibrillation.
The fractographs of a blend with 1.25 wt% SMA low (Figure 8b1,b2) show a similar surface to the prior discussed material with the brittle fracture of the PA66 matrix. In addition, pull-out of the PPE phases coexists with the fibrillar deformations of the matrix (blue arrows) where elongated PPE phases appear. Nevertheless, the tensile strength is found to be lower than the reference. Considering the high interfacial activity of SMA low (see Section 3.2), we assume that an incomplete coverage of the PA66/PPE interfaces results in a disorder with lower tensile strength. More precise, the covalent bond formation between PA66 and PPE ( Figure 8a2) seems to be interrupted when low amounts of SMA low are added. For 1.25 wt% SMA high (Figure 8c1,c2), a smooth fracture surface is seen for having a low plastic deformation of the PA66 matrix. The PPE phase is mainly pulled-out, revealing cavities with very smooth surfaces, indicating weak interfaces (orange arrows). As SMA high has a low affinity to PPE, it remains in the PA66 phase and does not disturb the formation of covalent bonds between PA66 and PPE. As for the reference, elongated PPE domains are seen, indicating locally strong interfacial bonds.
With increasing contents of SMA, a change in the morphologies is expected to result in different fracture mechanisms and thus surfaces. In Figure 8, the fractographs of 10 wt% SMA low (d1, d2) and SMA high (e1, e2) and are depicted. For SMA low (Figure 8d1), the fracture surface with an intermediate roughness is observed. In contrast to SMA high (Figure 8e1), step-like PA66 deformations (crack deflection) together with low levels of PPE elongation are very pronounced. Finding a significantly high amount of matrix fibrillation at the interfaces (Figure 8d2, blue arrows), it can be concluded that the stress transfer between both phases is very efficient. With this, the stress is deflected strongly with formation of sharp-edged steps parallel to the direction of force applied. . SEM fractographs after tensile tests of the reference binary blend (a1), 1.25 wt% SMA low (b1) and 1.25 wt% SMA high (c1), 10 wt% SMA low (d1), 10 wt% SMA high (e1), and their corresponding graphs at higher magnifications on the right column (a2, b2, c2, d2 and e2). The white arrows in Figure (a2) indicate strong matrix fibrillations, while the blue arrows in Figure (b2 and d2) indicate intermediate fibrillation, and the orange arrows in Figure (c2, e2) indicate weak interfaces. According to Figure 8e1 again, PA66 reveals a brittle behavior with a smooth fracture surface having weak crack deflections, whereas PPE shows a ductile behavior with cohesive failure. Typical pull-out effects are not apparent for co-continuous structures, as the phases are interpenetrating each other, acting as mechanical anchors. The PA66 exhibits further embrittlement due to dispersed SMA high -g-PA66 and SMA high micelles. Interestingly, no interfacial bonding is observed with higher SMA high content (Figure 8e2), meaning that the mechanical strength is only upheld by mechanical anchoring of the ductile PPE phase.
Cohesive failure of the individual blend phases is the predominant fracture mechanism for SMA low , whereas SMA high does not show sufficient stress transfer due to weak interfaces resulting in lower tensile performance, as summarized in Figure 9.

Conclusions
In this study, we reported that PA66/PPE blends can effectively be compatibilized by using SMA copolymers. A correlation between MA concentration and SMA content and the resulting morphology and tensile properties was successfully established. Herein, it was found that SMA high with higher than 8 wt% MA concentration is not miscible with PPE; thus, no interfacial interaction is observed. In contrast, SMA low (8 wt% MA) revealed a complete miscibility in PPE and high interfacial activity in PA66/PPE blends. The location of the SMA, tuned by its MA concentrations, controls the morphology of the blend systems. For the immiscible SMA high , micellar nanostructures within the PA66 phases in the diameter range of 24 ± 8 nm were observed. These lead to an increase of the PA66 phase viscosity, thus shifting the viscosity ratio of PA66 and PPE close to 1 (for 10 wt% SMA high ). With this, a transition from droplet-sea to co-continuous morphology for low contents of SMA high (between 1.25 and 7.5 wt%) was observed. For lower MA concentrations (SMA low ), the morphology transition is shifted to higher SMA contents starting from 7.5 wt% and ongoing for 10 wt%. With increasing SMA low content, a larger number of swollen micelles were seen in the TEM micrographs. These nano-emulsions were identified to be PPE pinch-offs covered by SMA low -graft-copolymers with PA66. For both SMA, a droplet-sea morphology is preferred because a co-continuous morphology leads to either a plateau or a decrease in the tensile strength of the blends. In terms of tensile properties, SMA low revealed the highest tensile strength of 72.5 MPa (7.5 wt%) and elongation at break of 5% (10 wt%). Enhanced tensile properties are explained by strong interfacial interaction and thus bonding between PA66 and PPE, which is expressed by the cohesive failure of the PPE phases together with strong matrix fibrillation at the interfaces. For SMA high , high contents were necessary to compensate for the tensile property loss. The partial recovery of the results is explained by mechanical anchoring of the PPE and PA66 phases, as a co-continuous morphology was observed for 10 wt% SMA high .