Interplay Between Octene Content and Grafting-Induced Molecular Weight Deviations and Their Effect on the Impact Toughness of Ethylene/1-Octene-Modified Polyamide 6

TL;DR

This paper studies how the octene content and grafting levels in ethylene/1-octene copolymers affect the toughness of polyamide 6 composites.

Contribution

The study reveals how octene content and grafting levels influence molecular weight changes and impact performance in polymer composites.

Findings

Higher octene content improves impact performance up to ≈0 °C but reduces performance at elevated temperatures.

Moderate grafting levels (0.5 wt%) and octene content (15 mol%) provide optimal impact strength and processability.

Stronger grafting increases polar interactions, reducing ductility and worsening impact performance above the BDTT.

Abstract

The impact modification of polyamide 6 (PA6) using maleic anhydride-grafted ethylene/1-octene copolymers (EOR-g-MAH) involves a trade-off between improved compatibilization, grafting-induced changes in modifier molecular weight MW, and melt processability. In this study, EOR modifiers with comparable initial MW but different octene contents (coct = 13, 15, and 16 mol%) were grafted to two MAH levels (cMAH = 0.5 and 1.0 wt%) and incorporated into PA6 at a fixed composition. The system was designed to maintain a comparable microstructure, enabling the isolation of grafting-induced changes in modifier properties from microstructural effects. MW distributions were analyzed by gel permeation chromatography, and the impact behavior was evaluated over a wide temperature range, using an instrumented Charpy impact test. The results reveal a strong, interrelated, coct- and cMAH-dependent competition between β-scission and cross-linking during grafting, which governs the modifier’s MW distribution and particle strength. Higher coct (15 and 16 mol%) enhances the impact performance up to ≈0 °C, well above the brittle–ductile transition temperature (BDTT), through increased elastic and plastic deformation capability of the modifiers. At elevated temperatures, however, successive melting of the modifiers leads to a loss of particle strength. At high coct and cMAH = 1.0 wt%, susceptibility to β-scission increases, leading to MW reduction that, for coct = 16 mol%, is detrimental to impact performance, particularly above the BDTT. This effect is further amplified by reduced ductility due to stronger polar intermolecular interactions at high grafting levels. A moderate cMAH = 0.5 wt% and coct = 15 mol% provides an optimal compromise between strength and ductility, delivering high impact strength across a broad temperature range. At this cMAH level, the number of PA6 chains covalently anchored to the modifier particles is moderate, resulting in lower compound viscosity and supporting favorable melt processability.

Figures

Tables

Table 2. Summary of the tensile test results of the modifiers: elastic modulus E, working capacity W, stress at yield σy, strain at yield εy, stress at break σb, and strain at break εb. Values are listed as mean values with standard deviations. Data for EOR-13-1 and EOR-15-1 are taken from ref. [41].

Designation	E[MPa]	W[J]	σy[MPa]	εy[%]	σb[MPa]	εb[%]
EOR-13-0	12.7 ± 0.7	4.9 ± 0.9	4.5 ± 0.1	41 ± 2	5.6 ± 0.4	695 ± 120
EOR-13-0.5	12.9 ± 0.2	4.3 ± 0.3	5.1 ± 0.1	34 ± 1	10.9 ± 0.3	327 ± 11
EOR-13-1	13.2 ± 0.2	4.3 ± 0.1	7.4 ± 0.1	63 ± 1	11.5 ± 0.2	291 ± 5
EOR-15-0	6.3 ± 0.2	6.2 ± 1.6	4.1 ± 0.1	54 ± 3	4.9 ± 0.2	942 ± 235
EOR-15-0.5	8.8 ± 0.4	5.8 ± 0.5	4.9 ± 0.1	82 ± 1	9.5 ± 0.4	496 ± 29
EOR-15-1	8.0 ± 0.4	4.3 ± 0.1	6.2 ± 0.1	84 ± 2	9.4 ± 0.2	348 ± 6
EOR-16-0	4.5 ± 0.2	12.0 ± 0.9	2.6 ± 0.1	114 ± 5	5.3 ± 0.1	1452 ± 67
EOR-16-0.5	6.0 ± 0.3	7.9 ± 0.4	4.0 ± 0.1	139 ± 3	7.3 ± 0.2	720 ± 25
EOR-16-1	4.2 ± 0.1	3.6 ± 0.2	4.8 ± 0.1	111 ± 3	5.4 ± 0.2	374 ± 13

Full text

1. Introduction

Melt blending of polyamide (PA) with soft (co)polymers exhibiting low stiffness and a low glass transition temperature $[eqn]$ , such as styrene-ethylene-butylene-styrene (SEBS) [1,2,3,4], acrylonitrile-butadiene-styrene (ABS) [5,6], ethylene-propylene rubber (EPR) [1,2,3,4,7,8,9,10], ethylene/1-octene rubber (EOR) [1,11,12,13,14,15], or ethylene-propylene-diene rubber (EPDM) [7,8,10,16], in amounts of 15–30 wt% is a well established approach to reduce the notch sensitivity of PA. This modification enables PA to achieve high notched impact strength $[eqn]$ , particularly at low temperatures $[eqn]$ [1,2,3,4,5,7,8,9,11,12,13]. The present study focuses on EOR modifiers due to their cost competitiveness and broad commercial availability in multiple grades offering a range of intrinsic parameters, including molecular weight $[eqn]$ , comonomer content $[eqn]$ , and crystallinity [17,18]. Their favorable cost arises from large-scale production for applications such as packaging and the use of simple olefin monomers, reducing the processing complexity. In addition, EOR modifiers exhibit UV stability, a favorable volatile organic compound (VOC) emission profile, and good resistance to automotive-relevant chemicals (e.g., cleaning agents), making them well suited for automotive components. These factors render EOR a technically and economically attractive alternative to styrene-based modifiers such as SEBS and ABS [19,20,21,22].

Due to the immiscibility arising from the different polarities of PA and copolymers, typically, a highly reactive unsaturated monomer, such as maleic anhydride (MAH), is grafted onto the copolymer backbone using an initiator, most commonly peroxide [19,23,24,25,26]. Peroxide-induced hydrogen abstraction from the copolymer chain enables covalent bonds with MAH [19,23,24,25]. Achieving higher MAH grafting degrees $[eqn]$ requires increased amounts of both MAH and peroxide [24,26]. Typical grafting levels range from 0.5 to 2.0 wt% [19].

During grafting, side reactions such as cross-linking, recombination, and β-scission influence the molecular weight distribution (MWD) of the modifiers [19,23,24,25,26]. Cross-linking or recombination of two chains via radical sites located on the polymer backbone leads to the formation of high- $[eqn]$ macromolecules. β-scission, which occurs predominantly at tertiary carbon atoms and increases proportionally with the $[eqn]$ , reduces the $[eqn]$ [19,25,27,28]. The addition of excess MAH can mitigate degradation [29], as it competes with the polymer backbone for reactive radicals, thereby suppressing β-scission. However, it may lead to an undesired release of unbound MAH during subsequent processing, which can be harmful due to outgassing, as well as its irritant and sensitizing effects.

During subsequent melt blending, a two-phase system forms, in which the modifier is dispersed as spherical particles within the PA matrix. The PA chains react with MAH on the particle surfaces, establishing connectivity through entanglements with the surrounding matrix. This ensures efficient stress transfer between the matrix and the modifier particles [30,31,32,33]. Below $[eqn]$ of the modifier, both phases deform predominantly elastically, and the compound fails in a brittle manner. Above the modifier’s $[eqn]$ , the modifier particles are capable of plastic deformation; however, a transition to ductile fracture occurs only at the brittle–ductile transition temperature (BDTT). At the BDTT, the particles elongate under deformation and undergo cavitation, thereby converting a triaxial stress state into a predominantly uniaxial one. This stress redistribution promotes shear yielding of the surrounding matrix and results in ductile fracture, which is accompanied by a significant increase in $[eqn]$ [2,7,10,30,32,34,35].

A homogeneous particle size distribution with controlled size and spacing is essential for achieving high $[eqn]$ and low BDTT [4,10,36,37]. Particles below a critical size do not cavitate, whereas excessively large particles increase the interparticle distance $[eqn]$ , thereby reducing the stress-transfer efficiency and restricting the plastic deformation of the matrix. Within an optimal particle size window, however, the $[eqn]$ has been reported to be largely insensitive to particle size [1,4]. For PA6.6- and PA6-based systems, an $[eqn]$ below a critical value of $[eqn]$ ≈ 300 nm is required, which is largely independent of the modifier chemistry and modifier content [36,37], because $[eqn]$ represents a matrix-controlled geometric criterion for overlapping stress fields and the onset of collective plastic deformation. Moreover, for EOR-modified PA6 compounds, an optimal weight-average particle diameter $[eqn]$ has been identified, with particle sizes of approximately 90–250 nm associated with a low BDTT and 150–800 nm associated with high $[eqn]$ values [1], with identical modifier content to this study. In contrast to $[eqn]$ , $[eqn]$ depends on modifier content, which controls the particle number density and thus the particle size required to provide a sufficient stress-affected volume to trigger cavitation.

In addition to an optimized microstructure, a low elastic modulus $[eqn]$ (i.e., low stiffness) of the modifiers, facilitating cavitation, and a low $[eqn]$ , which increases deformability at low $[eqn]$ , are required to achieve lower BDTT and high $[eqn]$ at low $[eqn]$ [2,8,10,38,39,40,41].

Previous studies have focused on adjusting the $[eqn]$ to attain high $[eqn]$ and low BDTT through an optimal microstructure [2,4,7,8,10,36,42,43]. As $[eqn]$ increases, the particle size decreases due to reduced interfacial tension [1,2,4,11,12,13,39]. Consequently, higher $[eqn]$ values have been considered advantageous for achieving low BDTT and high $[eqn]$ [1,2,4,8,11,13,16,39,44].

However, achieving higher $[eqn]$ requires larger amounts of MAH and peroxide, which intensify the side reactions, in particular β-scission, and thereby reduce the modifier’s $[eqn]$ [24,26]. In our previous study [41], we showed that low- $[eqn]$ modifiers exhibit reduced particle strength, promote less pronounced matrix shear yielding, and consequently lead to consistently lower $[eqn]$ and a higher BDTT than high- $[eqn]$ modifiers. Therefore, it is crucial to understand how $[eqn]$ affects $[eqn]$ , since $[eqn]$ not only improves compatibilization and reduces particle size but can simultaneously decrease the modifier’s $[eqn]$ , thereby diminishing particle strength and $[eqn]$ . Despite the strong interest in understanding the influence of $[eqn]$ on compound properties, earlier studies did not account for the effect of grafting on the modifier’s $[eqn]$ , even though variations in $[eqn]$ significantly affect the compound’s $[eqn]$ and BDTT, as demonstrated in our previous study [41]. Although numerous scientific publications exist, to the best of our knowledge, no study examined how deviations in the MWD—in particular, how different levels of $[eqn]$ and $[eqn]$ lead to distinct changes in this distribution—affect the $[eqn]$ .

Furthermore, the compound viscosity is higher than that of pure PA and rises further with increasing $[eqn]$ . Thus, while $[eqn]$ must be sufficiently high to obtain an optimal microstructure, excessively high levels should be avoided because they undesirably increase viscosity and may negatively impact subsequent forming processes [3,19,44,45].

The present study focuses on how $[eqn]$ and $[eqn]$ , individually and in combination, affect the MWD of the modifiers and, in turn, the $[eqn]$ of the compounds. The system was designed to keep the microstructure comparable, thereby isolating the effects of $[eqn]$ and $[eqn]$ .

To this end, EOR modifiers with comparable initial $[eqn]$ in the range of 34,000–42,000 g/mol were selected, as variations outside this range would significantly alter the viscosity ratio between the modifier and matrix, which governs droplet breakup and dispersion during melt mixing.

Modifiers were grafted with $[eqn]$ = 0.5 and 1.0 wt%, values that are both industrially relevant and widely reported in the literature [1,2,4,8,11,16,39,44]. Although MAH grafting may induce $[eqn]$ degradation, compensating for this effect by using substantially higher- $[eqn]$ base polymers is not a viable option in practice. EOR modifiers are predominantly produced for packaging applications, where low to moderate $[eqn]$ are required to ensure low melt viscosity, high processing stability at elevated throughputs, and cost-efficient large-scale manufacturing. As a result, high- $[eqn]$ EOR grades are rarely available at an industrial scale.

EORs with relatively high octene content ( $[eqn]$ = 13, 15, and 16 mol%) were chosen to ensure low $[eqn]$ and high deformability. Our preceding study [41] demonstrated that lower $[eqn]$ (8 mol%) results in a higher BDTT and reduced toughening efficiency. Higher $[eqn]$ grades beyond this range were not investigated, as excessive comonomer incorporation leads to pronounced material softening and tackiness, which can cause feeding and extrusion instabilities, such as pellet agglomeration, feeder blockage, or adhesion to the extruder screw. Moreover, such grades are associated with increased synthesis complexity, higher production costs, and limited availability at an industrial scale.

An instrumented Charpy impact test was employed to record force-displacement curves, enabling a more detailed analysis of fracture mechanisms than is possible with standard impact testing. Experiments were conducted over a broad $[eqn]$ range, and the interpretation of the impact behavior was supported by thermal and mechanical characterization of the modifiers. This approach allows for the evaluation of grafting-induced $[eqn]$ changes and their influence on $[eqn]$ in an industrially scalable PA6/EOR-g-MAH system with a controlled microstructure. The results indicate that MAH grafting influences strength and plastic deformability and, consequently, $[eqn]$ , with effects that depend on $[eqn]$ and $[eqn]$ .

Overall, an optimized combination of $[eqn]$ and $[eqn]$ in the modifiers not only enhances the impact resistance of the compound across a wide $[eqn]$ range but also improves processability by reducing its viscosity. While core–shell structures and more complex polymer alloys can offer superior performance in certain cases [46], they generally demand more intricate material design and processing compared to the approach presented here.

2. Materials and Methods

2.1. Materials

Three metallocene-catalyzed ethylene/1-octene copolymers (EORs) with different octene contents of $[eqn]$ = 13, 15, and 16 mol% and nearly identical initial molecular weights $[eqn]$ (prior to grafting) were grafted with maleic anhydride (MAH), resulting in MAH-functionalized EORs (EOR-g-MAH) with MAH contents $[eqn]$ of approximately 0.5 and 1.0 wt%. The chemical structures of EOR and EOR-g-MAH are shown in refs. [19,27,28]. These EOR-g-MAH were added at 20 wt% to a medium-viscosity, heat-stabilized polyamide 6 (PA6, AKROMID^®^ B3 1 natural) supplied by AKRO-PLASTIC GmbH (Niederzissen, Germany) to achieve impact toughening.

The nomenclature for the modifier follows the scheme EOR-“ $[eqn]$ [mol%]”-“ $[eqn]$ [wt%]”, while the corresponding PA6 compounds are designated as PA-EOR-“ $[eqn]$ [mol%]”-“ $[eqn]$ [wt%]”.

2.2. Modifier Characterization

The various characterization methods applied to the modifiers have been described in detail in a previous publication [41]. In this study, only the supplementary information that is required for interpreting the results is provided. The properties of the modifiers used in this work are summarized in Table 1.

The densities $[eqn]$ of the ungrafted EOR modifiers were measured by the buoyancy (Archimedes) method, and $[eqn]$ was subsequently determined from these values, according to the methods described in ref. [47]. The degree of MAH grafting $[eqn]$ was quantified on vacuum-dried, pressed films using Fourier transform infrared spectroscopy (FTIR), following the procedures described in refs. [19,44,48]. Exemplary infrared (IR) spectra of the ungrafted (EOR-13-0) and MAH-grafted samples (EOR-13-0.5 and EOR-13-1) are shown in Figure S1 (Supporting Information), and procedures are further discussed in both the Supporting Information and ref. [41]. The number-average and weight-average molecular weight ( $[eqn]$ and $[eqn]$ ) were determined from the molecular weight distribution (MWD) obtained by high-temperature gel permeation chromatography (GPC). The glass transition temperature $[eqn]$ was measured by dynamic mechanical thermal analysis (DMTA). The melting temperature $[eqn]$ and the crystallinity $[eqn]$ of the modifiers were extracted from differential scanning calorimetry (DSC) thermograms. Both the overall $[eqn]$ and the $[eqn]$ corrected for the contribution of the process additive are reported in Table 1 (see also FTIR results in the Supporting Information, Section S1). The contribution of the additive appears as a shoulder in the DSC thermograms around 80 °C, quantified by fitting a Lorentzian, and subsequently subtracted from the overall $[eqn]$ . The corrected $[eqn]$ values allow us to unambiguously isolate the effect of the grafting process on $[eqn]$ . In contrast, in refs. [41,49], no correction was applied, as the effective uncorrected $[eqn]$ was considered decisive for the performance of the modifiers in the compound.

When interpreting the impact test results, it should be noted that the difference in strain rates between DMTA and impact testing leads to an upward shift in the apparent $[eqn]$ by approximately 9 °C. Even considering this shift, the effective $[eqn]$ < −41 °C of the modifiers remains well below the brittle–ductile transition temperatures (BDTT) (≈−20 °C) observed in the impact tests (see also Section 3.2.3). For further details, the reader is referred to ref. [41].

Finally, Table 2 summarizes the tensile properties of the modifiers, including the elastic modulus $[eqn]$ , working capacity $[eqn]$ , stress at yield $[eqn]$ , strain at yield $[eqn]$ , stress at break $[eqn]$ , and strain at break $[eqn]$ , obtained from uniaxial tensile testing.

2.3. Preparation and Characterization of the Compound

2.3.1. Compounding

The compounds were melt-blended in a co-rotating twin-screw extruder (FED 26 MTS, FEDDEM GmbH & Co. KG, Sinzig, Germany) at a set temperature of 250 °C, a screw rotational speed of 660 min^−1^, and a feed rate of 50 kg/h. A screw with an outer-to-inner screw diameter ratio $[eqn]$ / $[eqn]$ = 1.55 and a configuration incorporating both distributive and dispersive mixing elements was used to ensure a reproducible microstructure and fine dispersion. All raw materials were simultaneously fed into the extruder at a constant PA6/EOR-g-MAH ratio of 80/20 wt%. The moisture content of the PA6 used during compounding was maintained below 0.1 wt%.

2.3.2. Impact Test

The specimens for the impact tests on the PA6/EOR-g-MAH compounds were processed using an injection molding machine (ARBURG 420 C 1000-250, ARBURG GmbH & Co. KG, Loßburg, Germany) at barrel and nozzle temperatures between 240 and 255 °C and a mold temperature of 80 °C. Prior to processing, compound granules were dried in a vacuum oven at 80 °C to a residual moisture content of (0.05 ± 0.01) wt%. The moisture content was determined by Karl Fischer titration (C30 Coulometric KF Titrator, Mettler-Toledo GmbH, Greifensee, Switzerland).

The impact test specimens were prepared by injection molding Type A1 tensile bars according to the geometry specified in DIN EN ISO 527-2 [50]. The shoulders were subsequently removed, and a 45° V-notch with a notch radius of (0.25 ± 0.05) mm was introduced, in accordance with DIN EN ISO 179-1/1eA [51]. Prior to impact testing, the notched specimens were conditioned under standard climate conditions at 23 °C and 50% relative humidity for 16–24 h, in accordance with DIN EN ISO 179-1 [51]. For impact tests conducted at different temperatures $[eqn]$ , the specimens were conditioned at the respective test temperature prior to each measurement to ensure a uniform $[eqn]$ throughout the specimens. Conditioning was performed in a climate chamber for more than 3 h. The specimens were removed using tweezers and tested within 5 s, following the procedure described for metals in DIN EN ISO 148-1 [52].

The impact response of the compounds was measured using an instrumented 25 J hammer mounted on an impact testing device (HIT25P, ZwickRoell GmbH & Co. KG, Ulm, Germany). Tests were performed following DIN EN ISO 179-2 [53]. The instrumented striker recorded the force $[eqn]$ experienced by the specimen during impact. The area under the resulting force-deflection curves $[eqn]$ was separated into the energy dissipated up to crack initiation, $[eqn]$ , and the energy associated with crack propagation, $[eqn]$ . To determine the force maximum, a Lorentzian was fitted to the peak region of $[eqn]$ . For each material and $[eqn]$ , at least five specimens were tested. The reported $[eqn]$ values represent mean values, and the error bars correspond to the standard deviations.

2.3.3. Viscosity Measurement

The compound viscosities, $[eqn]$ , were measured at 240 °C, using a high-pressure capillary viscometer (Göttfert RG50, Göttfert GmbH, Buchen, Germany) over a shear-rate range $[eqn]$ = 3.5 × 10^2^–3.7 × 10^4^ s^−1^. Capillaries with a length-to-diameter ratio $[eqn]$ of 20/1 and 40/1 were employed. Measurements were initiated 10 min after the target temperature had been reached. All reported $[eqn]$ data were corrected according to the Bagley and Weißenberg–Rabinowitsch procedures. To ensure a constant moisture content of (0.05 ± 0.01) wt%, the compounds were dried in a vacuum oven at 80 °C prior to testing. The moisture content was confirmed by Karl Fischer titration (C30 Coulometric KF Titrator, Mettler-Toledo GmbH, Greifensee, Switzerland).

2.3.4. Scanning Electron Microscopy

The compounds’ microstructure was examined by scanning electron microscopy (SEM, Jeol JSM-7200F, JEOL Ltd., Tokyo, Japan) operated at an acceleration voltage of 15 kV. Cryo-fractured specimens were immersed in boiling xylene overnight to dissolve the modifier phase. Subsequently, a 6 nm thick iridium layer was sputtered onto the specimens, using a sputtering device (Safematic CCU-010, Safematic GmbH, Zizers, Switzerland).

Image analysis was performed using Image-Pro Premier Ver. 9.2 (Media Cybernetics Inc., Rockville, MD, USA) to determine the dimensions of the voids remaining after dissolution of the modifier, which were associated with the particle diameter. For each compound, at least 400 particles were evaluated. From the void-diameter distribution, the number-average particle diameters $[eqn]$ , the weight-average particle diameters $[eqn]$ , the volume-average particle diameters $[eqn]$ , and the interparticle distances $[eqn]$ were determined. The corresponding formulas are provided in our previous study [41]. The polydispersity index $[eqn]$ was calculated as the ratio $[eqn]$ / $[eqn]$ . The different average particle diameters, $[eqn]$ , and $[eqn]$ values, together with their standard deviations derived from the analysis of at least two SEM microphotographs, are summarized in Section 3.2.1.

3. Results and Discussion

3.1. Modifier Properties

3.1.1. Thermal and Temperature-Dependent Dynamic–Mechanical Properties

The thermal properties of the EOR-g-MAH modifiers with varying $[eqn]$ and $[eqn]$ were analyzed using DSC and DMTA. Figure 1a presents the DSC thermograms (heat flow from the second heating runs) for the ungrafted and grafted EOR copolymers ( $[eqn]$ = 0.5 and 1.0 wt%) with $[eqn]$ = 13, 15, and 16 mol%. The corresponding values for $[eqn]$ and $[eqn]$ are listed in Table 1. To provide a more complete picture of how thermal properties evolve with $[eqn]$ , the Supporting Information (Section S2) also includes DSC and DMTA results for EOR-8 (see also [41]).

All materials exhibit a broad melting region that begins immediately above the glass transitions. The $[eqn]$ values, determined from the peak of the loss modulus from DMTA measurements (see Figure 2 and Table 1), are equal to (−55 ± 1) °C (EOR-16), (−51 ± 1) °C (EOR-15), and (−50 ± 1) °C (EOR-13) and are independent of $[eqn]$ within the experimental error. The hexyl side chains of the octene comonomer increase the free volume relative to polyethylene and simultaneously hinder crystallization. Consequently, increasing $[eqn]$ leads to lower $[eqn]$ , $[eqn]$ , and $[eqn]$ , shifting the entire melting region to lower $[eqn]$ , as also discussed in our previous study [41] and ref. [17]. Grafting slightly affects the melting behavior of the modifiers, as the MAH units interfere with polymer chain packing during crystallization [54,55,56,57]. This reduced packing efficiency results in smaller and imperfect crystallites, leading to lower melting peak heights, a shift of $[eqn]$ to lower $[eqn]$ , and a tendency toward reduced $[eqn]$ compared to the ungrafted materials (see Table 1 and refs. [54,55,56,57]). Increasing $[eqn]$ beyond 0.5 wt% does not further hinder crystallization.

In the DSC thermograms, a shoulder appears at approximately 80 °C, which is attributed to a process additive present in the peroxide- and MAH-containing masterbatch used for grafting, which was also observed in the IR spectrum (see Section S1, Supporting Information). $[eqn]$ were corrected for the contribution of the additive; however, the correction reduces $[eqn]$ by a maximum of only 0.8% and is therefore considered negligible for further discussion.

For interpreting the $[eqn]$ behavior of the compounds, the $[eqn]$ of the modifiers at the testing temperature is more relevant than the original $[eqn]$ , determined from the area under the entire melting peak. Figure 1b shows the uncorrected residual $[eqn]$ of the modifiers as a function of $[eqn]$ . Due to the early onset of melting, $[eqn]$ decreases continuously above $[eqn]$ , and at room temperature ( $[eqn]$ = 23 °C), it is substantially reduced. For EOR-16, EOR-15 and EOR-13, independent of $[eqn]$ , the reduction amounts to approximately (47 ± 1)%, (35 ± 1)%, and (28 ± 1)%, with residual $[eqn]$ of about (4 ± 1)%, (8 ± 1)%, and (11 ± 1)%, respectively. The melting of the modifiers, accompanied by a significant loss of crystalline material, affects their strength at elevated $[eqn]$ . Modifier strength is critical for the toughness of the compounds, as it governs the induction of shear yielding in the PA6 matrix (see discussion below).

Figure 2 shows the storage modulus $[eqn]$ and loss modulus $[eqn]$ as a function of $[eqn]$ , measured by DMTA, for the ungrafted and grafted modifiers investigated in this study: EOR-13, EOR-15, and EOR-16. The measurements cover the $[eqn]$ range that is relevant for interpreting the impact tests. According to ref. [8], the higher strain rates in an impact test compared to the low strain rates in a DMTA experiment can be accounted for by a $[eqn]$ shift $[eqn]$ of approximately 9 °C toward higher $[eqn]$ . Even after applying this shift, the glass transitions of all modifiers $[eqn]$ ≈ −41 °C (EOR-13), −42 °C (EOR-15), and −46 °C (EOR-16) remain well below the BDTT observed in the impact tests (see Section 3.2.3).

Consistent with the DSC results, the gradual softening of the materials above $[eqn]$ , as indicated by the decrease in $[eqn]$ , is attributed not only to the increased mobility of the amorphous phase but also to the onset of melting of the crystalline phase just above $[eqn]$ [17,58]. Above $[eqn]$ , $[eqn]$ is systematically lower and decreases more rapidly with increasing $[eqn]$ for EOR-16 compared to EOR-15 and EOR-13, reflecting the higher chain mobility (due to a larger free volume) and lower $[eqn]$ associated with higher $[eqn]$ , as also discussed in refs. [17,28]. For EOR-13 and EOR-15, no significant effect of $[eqn]$ on the DMTA results is observed. In contrast, for EOR-16, both moduli above $[eqn]$ are systematically higher for the grafted samples compared to the respective ungrafted ones. With increasing $[eqn]$ , the crystalline network weakens and contributes less to $[eqn]$ . At lower $[eqn]$ levels, the modulus is dominated by the crystalline phase, masking polar interactions induced by (MAH) and maleic acid (MA) moieties, whereas at $[eqn]$ = 16 mol%, the amorphous phase dominates and MA(H)-induced dynamic physical cross-links become mechanically significant, increasing both $[eqn]$ and $[eqn]$ . MA is formed via hydrolysis of MAH [59] during processing. The effect of the polar interactions on the mechanical properties of the modifiers is discussed in more detail in Section 3.1.3.

3.1.2. Gel Permeation Chromatography

In Figure 3, the MWD of the copolymers EOR-13, EOR-15, and EOR-16 are presented for different $[eqn]$ (ungrafted, 0.5 wt%, and 1.0 wt% MAH). Figure S4 in the Supporting Information shows the same data but compares the effect of varying $[eqn]$ at fixed $[eqn]$ . Values for $[eqn]$ and $[eqn]$ are summarized in Table 1.

The changes observed in the MWD are attributed to side reactions, including β-scission and cross-linking, that occur during the grafting process [19,21,24,56,60,61]. During grafting, the polymer is mixed with MAH and peroxide. The elevated $[eqn]$ associated with the mixing process induces thermal decomposition of the peroxide, leading to the formation of free radicals [19,56,60,61]. These radicals can abstract hydrogen atoms from the copolymer chains, generating macroradicals (EOR●) that are capable of undergoing a wide variety of subsequent reactions [19,28,56,60,61]. For the purpose of the present discussion, we focus on the most relevant reactions.

The reaction is initiated by the thermal decomposition of the peroxide, generating primary radicals that abstract hydrogen atoms from the polymer backbone and form EOR● macroradicals. In the desired reaction, EOR● reacts with MAH, forming an EOR-g-MAH● macroradical, which can then abstract a hydrogen atom from the same or another polymer chain, terminating the radical at that position while simultaneously generating a new macroradical at the site of abstraction [56,60,61]. In addition to this intended propagation, both EOR● and EOR-g-MAH● can react with each other, resulting in chain coupling (e.g., EOR-EOR or EOR-g-MAH-EOR) or cross-linked structures, in which large macromolecules are formed.

Hydrogen abstraction can also initiate β-scission, which may occur either directly at tertiary carbon atoms or initially at secondary carbon atoms, followed by radical migration along the polymer chain and eventual stabilization at a tertiary carbon site [19,24,28,56,61]. Consequently, β-scission predominantly occurs at tertiary carbons, due to the lower energy required for hydrogen abstraction and the higher stability of the resulting tertiary radicals compared to the secondary radicals [28,62].

To illustrate the extent of the respective side reactions as a function of $[eqn]$ and $[eqn]$ , Figure 4 presents the relative $[eqn]$ changes $[eqn]$ ( $[eqn]$ ), comparing $[eqn]$ and $[eqn]$ of the grafted copolymers ( $[eqn]$ ) with those of the corresponding ungrafted EOR materials ( $[eqn]$ ) for all modifiers.

Upon grafting, $[eqn]$ decreases for all modifiers, with the reduction becoming more pronounced at higher $[eqn]$ . Elevated peroxide concentrations are required to achieve higher grafting densities. Under these conditions, the increased radical concentration enhances the probability of β-scission relative to graft formation, even though the MAH concentration is increased to a greater extent than the peroxide concentration. β-scission is a unimolecular, intramolecular reaction that proceeds more rapidly than the bimolecular propagation involving MAH addition. Furthermore, the formation of relatively stable tertiary radicals during grafting kinetically favors chain cleavage over propagation. The dependency of $[eqn]$ on the $[eqn]$ and $[eqn]$ requires a more detailed discussion. At $[eqn]$ = 0.5 wt%, $[eqn]$ is lower than that of the ungrafted modifiers and decreases further with increasing $[eqn]$ . At this grafting level, the amounts of peroxide and MAH are insufficient to induce significant cross-linking, and β-scission predominates. With increasing $[eqn]$ , the number of tertiary carbons rises, further enhancing the probability of β-scission.

At $[eqn]$ = 1.0 wt%, the increased amounts of peroxide and MAH not only enhance β-scission but also promote cross-linking. $[eqn]$ is increased for EOR-13-1 and EOR-15-1 compared to the corresponding ungrafted modifiers, which is likely due to the higher macroradical concentrations, which reduce the average distances between radicals and thereby increase the probability of recombination of large and bulky macroradicals, as discussed in refs. [29,63]. As $[eqn]$ increases from 13 to 16 mol%, β-scission becomes increasingly dominant over cross-linking due to the higher proportion of tertiary carbons, which is consistent with refs. [28,57]. For EOR-16-1, a significant relative decrease in $[eqn]$ is observed, indicating that cross-linking is no longer significant. This is most likely a consequence of steric hindrance between macroradicals arising from the increased number of side chains, as noted in ref. [27], combined with a high probability of β-scission.

3.1.3. Tensile Properties

Figure 5 shows one representative stress $[eqn]$ –strain $[eqn]$ curve for EOR-13, EOR-15 or EOR-16, either ungrafted or grafted with $[eqn]$ = 0.5 or 1.0 wt%, respectively. The corresponding values of the elastic modulus $[eqn]$ , stress at yield $[eqn]$ , strain at yield $[eqn]$ , stress at break $[eqn]$ and strain at break $[eqn]$ of the pure modifiers are listed in Table 2.

To determine $[eqn]$ , the minimum of the first derivative of the stress–strain curve was used, and the stress at this strain was taken as $[eqn]$ . However, for EOR-13-1, EOR-15-1, and EOR-16-1, no minimum in the first derivative could be identified. In these cases, $[eqn]$ was identified as the intersection point of the slopes of the first derivative in the two approximately linear regions immediately before and after the pronounced change in slope. The corresponding stress at this intersection point on the stress–strain curve was taken as $[eqn]$ .

In the following, the influence of $[eqn]$ on the tensile properties of the ungrafted modifiers is discussed first, before the effect of $[eqn]$ is addressed.

The deformability in the low-strain region, i.e., the elastic regime, systematically increases with increasing $[eqn]$ from EOR-13-0 to EOR-15-0 and EOR-16-0. As reported in refs. [17,18], this behavior is predominantly governed by $[eqn]$ , which decreases from EOR-13 over EOR-15 to EOR-16, as confirmed by the DSC measurements (see Section 3.1.1 and Table 1). Lower $[eqn]$ increases the amorphous fraction, leading to enhanced deformability. These results also align with the trend in $[eqn]$ obtained from DMTA above $[eqn]$ (see Section 3.1.1).

As the strain increases, the ungrafted modifiers exhibit a distinct stress maximum followed by a pronounced stress drop, indicating the onset of necking. Specimens of EOR-13-0 and EOR-15-0 show strongly localized necking, whereas EOR-16-0, with higher $[eqn]$ , exhibits broader, less localized necking, reflected in a wider stress maximum [17,64,65].

Increasing $[eqn]$ , corresponding to a higher number of hexyl side chains, affects the deformation behavior through several mechanisms: (i) it reduces $[eqn]$ by disturbing the regular packing of the polymer backbone [17,64], and (ii) it introduces strong steric hindrance [65]. As a direct consequence of steric hindrance, the free volume in the amorphous regions increases (iii) [17,64], which reduces the probability of topological interchain entanglements, lowering the effective entanglement density (iv) [66]. Although reduced entanglement density (iv) tends to promote localized necking, steric hindrance from bulky side chains (ii), together with increased free volume (iii), limits the molecular orientation under tensile load and distributes deformation, suppressing localized volume contraction perpendicular to the stretching direction and thereby reducing necking [17,18,65]. Reduced $[eqn]$ also contributes by enlarging the amorphous phase, allowing deformation to be spread over a broader region [17]. Correspondingly, as $[eqn]$ increases, $[eqn]$ decreases, while $[eqn]$ increases, reflecting the delayed onset of necking and the more distributed stress over a larger effective cross-section.

In the subsequent cold-drawing regime, the neck propagates along the specimen and develops an oriented fibrillar morphology, stabilizing once it reaches the specimen shoulders, as previously described in ref. [17]. Following this, strain hardening occurs, where stress increases until failure as chains align and stretch, constrained by entanglements and tie molecules connecting crystalline domains [17,67]. Increasing $[eqn]$ reduces these constraints, weakening strain hardening. Higher $[eqn]$ is consistent with the larger amorphous fraction and the presence of fewer, smaller, less interconnected fringed micelles [17,18]. Despite the weakened strain hardening, the overall strength remains roughly constant because the reduced effect of necking (larger effective cross-section) offsets the expected weakening. Consequently, the working capacity $[eqn]$ , defined by the area under the stress–strain curve, is comparable for EOR-13-0 and EOR-15-0 within the error bars, but is significantly higher for EOR-16-0 due to its greater ductility, reflected in the higher $[eqn]$ at a similar apparent strength.

For the grafted modifiers, tensile properties differ significantly from those of the ungrafted ones. In the elastic region, grafting has only a minor influence on stiffness, as the slight reduction in $[eqn]$ is largely compensated, or even outweighed, by stronger polar interactions between the grafted chains in the amorphous phase, particularly for the high- $[eqn]$ modifiers (see Table 1, Section 3.1.1, and discussion below). For EOR-16-1, however, the pronounced reduction in $[eqn]$ and the associated loss of entanglements counteract this stiffening effect, resulting in a lower $[eqn]$ compared to EOR-16-0.5.

At higher strains, clear distinctions between the various $[eqn]$ levels emerge. Whereas ungrafted modifiers exhibited pronounced necking, necking formation is strongly suppressed at $[eqn]$ = 0.5 wt% and absent at $[eqn]$ = 1.0 wt%. Grafted modifiers display reduced $[eqn]$ and generally higher $[eqn]$ , $[eqn]$ and $[eqn]$ compared to their ungrafted counterparts, with the increase in $[eqn]$ and the reduction in $[eqn]$ becoming more pronounced at higher $[eqn]$ , consistent with refs. [55,68].

These effects can be primarily attributed to attractive, polar interactions between maleic anhydride (MAH) and maleic acid (MA) moieties along the grafted polymer chains [54,55,57,68]. MA units form via hydrolysis in the presence of ambient moisture during processing at elevated $[eqn]$ [59]. The resulting dipole–dipole interactions between MAH groups [69] and hydrogen bonding between MA groups [70] act as additional physical cross-links within the amorphous phase, increasing the resistance to plastic deformation and suppressing necking.

Compared to the ungrafted materials, grafting introduces additional physical constraints in the amorphous phase, which serve as load-transferring bridges between crystalline clusters and represent the primary site of plastic deformation. Attractive polar interactions between MAH and MA groups act as reversible physical cross-links, effectively increasing the density of intermolecular junctions [55]. Bulky grafted moieties impose steric constraints that reduce segmental mobility and increase resistance to chain sliding and disentanglement. Both effects are expected to intensify with increasing $[eqn]$ .

As a consequence, once the amorphous phase has stretched and the crystalline domains become aligned in the loading direction, stronger intermolecular interactions and steric hindrance prevent local rearrangement of the crystalline domains and limit chain disengagement from them [55,67]. This stabilizes the oriented network structure, increases resistance to plastic flow, and enhances the strain-hardening modulus, albeit at the expense of reduced $[eqn]$ . These two constraint mechanisms reduce the ability of both amorphous chain segments and crystalline domains to undergo further reorganization and alignment under continued deformation. As a result, strain localization is suppressed and necking is reduced or even absent, maintaining a larger cross-sectional area during deformation and thereby influencing the measured strength. Compared to the ungrafted counterparts, the combination of reduced necking and stronger intermolecular interactions increases both $[eqn]$ and $[eqn]$ , while $[eqn]$ decreases upon grafting. In contrast, $[eqn]$ tends to increase, relative to the ungrafted materials, because necking is hindered, causing $[eqn]$ to be reached at higher strains.

Increasing $[eqn]$ from 0.5 to 1.0 wt% does not significantly change the values of $[eqn]$ and $[eqn]$ for EOR-13. Similarly, $[eqn]$ remains comparable between EOR-15-0.5 and EOR-15-1; however, $[eqn]$ decreases as $[eqn]$ increases from 0.5 to 1.0 wt% MAH. Increasing the MAH grafting level beyond 0.5 wt% does not lead to further enhancement in modifier strength. Although β-scission reduces $[eqn]$ , this effect is at least partially offset by an increase in $[eqn]$ , i.e., by the presence of additional high- $[eqn]$ chains formed by cross-linking, thereby preserving the density of tie-molecule-based junctions between crystalline domains. The observed reduction in $[eqn]$ arises because enhanced intermolecular interactions induced by MAH and MA hinder chain sliding in the amorphous regions and thus limit plastic deformation, causing $[eqn]$ to be reached at lower strains, as discussed in refs. [55,68], with steric constraints possibly further limiting chain sliding and disentanglement.

According to ref. [57], the impact of cross-linking and β-scission on the stress–strain behavior is smaller than the effects discussed above. For EOR-16-1, however, both $[eqn]$ and $[eqn]$ are significantly reduced compared to EOR-16-0.5. This behavior is primarily a consequence of the substantial decrease in $[eqn]$ caused by grafting and the dominance of β-scission over cross-linking (Section 3.1.2), which outweighs the positive contributions of enhanced intermolecular interactions through MAH and MA and diminishes the effective bridging between crystalline domains. This response is consistent with the premature failure of the low- $[eqn]$ modifier compared to the high- $[eqn]$ counterpart ( $[eqn]$ = 15 mol%, $[eqn]$ = 1.0 wt%) discussed in the previous study [41]. Furthermore, the effect of increased polar interactions and steric hindrance on $[eqn]$ is expected to become more pronounced at higher $[eqn]$ , due to the larger fractions of amorphous regions. This also explains why increasing $[eqn]$ has little effect on $[eqn]$ in EOR-13, as stronger polar interactions contribute less due to the relatively small amorphous fraction, while the highly crystalline domains act as the main load-bearing network during deformation.

Finally, $[eqn]$ is mainly determined by the trend in $[eqn]$ . $[eqn]$ is comparable for the ungrafted and grafted EOR-13 modifiers, indicating that the increased strength upon grafting compensates for the reduction in $[eqn]$ . For EOR-15, $[eqn]$ is comparable for EOR-15-0 and EOR-15-0.5, but is significantly lower at EOR-15-1, due to the noticeably reduced $[eqn]$ at higher $[eqn]$ . For EOR-16, $[eqn]$ decreases from EOR-16-0 to EOR-16-0.5 and, in particular, shows a strong drop for EOR-16-1, resulting from the reductions in both $[eqn]$ and $[eqn]$ caused by the substantial decrease in $[eqn]$ .

3.2. Compound Properties

3.2.1. Microstructure

Given the strong influence of microstructure on $[eqn]$ , particle diameters were measured and the corresponding $[eqn]$ calculated. Particles below a critical size cannot cavitate, whereas excessively large particles, depending on the modifier volume fraction, lead to $[eqn]$ s that are too large, preventing the initiation of shear yielding in the surrounding matrix [1,36,37]. Within this optimal size range, $[eqn]$ shows little dependence on the particle diameter [1]. For PA6- and PA6.6-based systems, effective toughening requires an $[eqn]$ below a critical value of $[eqn]$ ≈ 300 nm, largely independent of modifier chemistry and content [36,37]. For EOR-modified PA6, the optimal weight-average particle diameter $[eqn]$ lies between approximately 90–250 nm for low BDTT and 150–800 nm for high $[eqn]$ , at modifier contents that are identical to those in this study [1].

Table 3 summarizes the number-average $[eqn]$ , weight-average $[eqn]$ , and volume-average diameters $[eqn]$ , the corresponding polydispersity indices $[eqn]$ , and the calculated interparticle distance $[eqn]$ for all compounds investigated in this study. Figure 6 shows representative surfaces of cryo-fractured and xylene-etched specimens for the compounds PA-EOR-13-0.5 and PA-EOR-13-1. SEM micrographs for all remaining compounds used in this study are provided in the Supporting Information (Section S4). In all cases, the modifier is dispersed as spherical particles within the matrix.

At constant $[eqn]$ , all particle sizes and, consequently, the $[eqn]$ are comparable (see Table 3), and $[eqn]$ shows no notable influence on the microstructure, as discussed in our previous study [41].

Increasing $[eqn]$ from 0.5 wt% to 1.0 wt% decreases $[eqn]$ from (178 ± 4) nm to (153 ± 3) nm (see Table 3). This reduction is attributed to the lower interfacial tension resulting from higher $[eqn]$ , as suggested in ref. [37]. It should be noted that both $[eqn]$ and $[eqn]$ decrease significantly, indicating that the larger particles are particularly affected. This also results in a slightly narrower particle-size distribution, as reflected by the reduced $[eqn]$ . Please note that EOR-16-0.5 exhibits a slightly higher $[eqn]$ than EOR-13-0.5 and EOR-15-0.5 (Table 1), resulting in somewhat smaller particle sizes and lower $[eqn]$ .

Nevertheless, the particle diameters for $[eqn]$ = 0.5 wt% ( $[eqn]$ ≈ 226 ± 9 nm) and 1.0 wt% ( $[eqn]$ ≈ 184 ± 3 nm) remain within the optimal size range of $[eqn]$ ≈ 150–250 nm, which is suitable for achieving both low BDTT and high $[eqn]$ [1]. In addition, the $[eqn]$ of each compound stays well below the $[eqn]$ = 300 nm reported in refs. [36,37], ensuring that all investigated compounds meet the necessary conditions for effective toughening. Furthermore, the variation in particle size is relatively small, indicating that the subtle differences in microstructure are unlikely to account for the observed discrepancies in the impact test results.

3.2.2. Compound Viscosity

Both $[eqn]$ and flowability during processing are particularly important for the intended application of injection molding. Therefore, the $[eqn]$ of the compound was measured at 240 °C. Figure 7 shows $[eqn]$ as a function of $[eqn]$ for the representative samples PA-EOR-13-0.5 and PA-EOR-13-1. The corresponding results for PA-EOR-15 and PA-EOR-16 are presented in the Supporting Information (Figure S6). All compounds exhibit shear-thinning behavior as the PA6 chains orient along the flow direction under shear, thereby reducing the resistance to flow. For injection molding, the shear rates typically range between approximately 10^3^ and 10^4^ s^−1^, depending on the part geometry and flow rate [71]. Within this range, the $[eqn]$ of PA-EOR-13-0.5 is about 19% lower than that of PA-EOR-13-1. The corresponding decreases in $[eqn]$ are 12% and 11% for PA-EOR-15 and PA-EOR-16, respectively. This suggests that reducing $[eqn]$ lowers the number of covalently bound PA6 chains and thus the number of entanglements between EOR-g-MAH particles and the PA6 matrix, reducing $[eqn]$ . This is also discussed in refs. [72,73]. The $[eqn]$ decrease is further supported by the larger particle size at lower $[eqn]$ , which reduces the total interfacial area. Conversely, higher $[eqn]$ yields more entanglements and smaller particles with larger interfacial area—both effects act to increase $[eqn]$ .

3.2.3. Impact Toughness and Energy Dissipation

Instrumented impact testing distinguishes between energy dissipated during crack initiation $[eqn]$ and energy dissipated during crack propagation $[eqn]$ , providing detailed insight into fracture mechanisms. Below the BDTT, in the brittle fracture regime, there is essentially no resistance to crack propagation, and $[eqn]$ is primarily governed by $[eqn]$ . Energy dissipation occurs via elastic deformation of the matrix and, above the modifiers’ $[eqn]$ (here −55 to −50 °C, well below the low $[eqn]$ limit of the impact test), also through plastic deformation and, eventually, modifier particle cavitation [7,42].

At the BDTT, the fracture mechanism changes, $[eqn]$ becomes dominant, and energy dissipation occurs primarily through matrix shear yielding [10]. Stress transfer from the matrix to the particles induces particle deformation and, at a certain stage, leads to cavitation at the particle centers, where the stress is concentrated [7,10]. Particle cavitation converts the triaxial stress state into a uniaxial one, thereby enabling extensive shear yielding of the surrounding matrix, provided that the particle strength is sufficiently high to hinder crack propagation [8,10] and that the particles exhibit high deformability [2,8,10,38,39,40,41], allowing for effective activation of shear yielding. It turns out that at low $[eqn]$ , the modifier’s deformability, as reflected by a low storage modulus $[eqn]$ , is essential for high $[eqn]$ , whereas at higher $[eqn]$ , a balance between strength and ductility, i.e., $[eqn]$ together with $[eqn]$ , also becomes important.

Figure 8 presents the impact test results as a function of $[eqn]$ for PA6 and the PA6 compounds PA-EOR-13, PA-EOR-15, and PA-EOR-16, with modifiers grafted at approximately $[eqn]$ = 0.5 or 1.0 wt%. Figure 8a shows the overall $[eqn]$ , whereas Figure 8b,c compare $[eqn]$ and $[eqn]$ for compounds containing modifiers with different $[eqn]$ at the same $[eqn]$ . Figure 8d–f, in turn, compare compounds containing modifiers with different $[eqn]$ but the same $[eqn]$ . Since all samples exhibit comparable microstructures (particle size, $[eqn]$ , and $[eqn]$ ; see Section 3.2.1) and adequate adhesion between the modifier and matrix [13], differences in $[eqn]$ cannot arise from these factors.

The discussion begins with the general trends of the overall $[eqn]$ in relation to $[eqn]$ and $[eqn]$ as presented in Figure 8a.

Pure PA6 exhibits brittle behavior across the entire $[eqn]$ range. The $[eqn]$ of the modified compounds is higher than that of neat PA6, even below the BDTT, due to energy dissipation via modifier particle deformation and cavitation. All impact-modified compounds exhibit a BDTT around −20 °C, marked by a sharp increase in $[eqn]$ , which corresponds to the onset of matrix shear yielding.

As previously shown, increasing $[eqn]$ from 8 to 13 mol% shifts the BDTT by approximately −5 °C toward lower $[eqn]$ (see ref. [41]). This shift is attributed to the lower $[eqn]$ and lower $[eqn]$ of the modifier with $[eqn]$ = 13 mol% ( $[eqn]$ = −50 °C, $[eqn]$ = 14.7 ± 0.1%) compared to the modifier with $[eqn]$ = 8 mol% ( $[eqn]$ = −44 °C, $[eqn]$ = 24.6 ± 0.4%). DSC and DMTA results for the 8 mol% modifier are provided in Section S2 (Supporting Information) for direct comparison. The increase in $[eqn]$ enhances particle deformability, as reflected by the lower $[eqn]$ values obtained from DMTA measurements, and promotes cavitation. This leads to a more efficient stress release, which in turn facilitates shear yielding of the surrounding matrix. Further increase in $[eqn]$ does not shift the BDTT, although using EOR-16 further improves the $[eqn]$ at the BDTT. This aligns with the fact that the $[eqn]$ of EOR-13 ( $[eqn]$ = −50 °C) and EOR-15 ( $[eqn]$ = −51 °C) are similar, whereas EOR-16 has a lower $[eqn]$ = −55 °C (see Table 1). However, the higher deformability of EOR-16 is likely partly offset by a reduction in particle strength, due to its lower $[eqn]$ and/or $[eqn]$ reduction resulting from β-scission (see also the discussion below).

For the subsequent detailed discussion, the influence of $[eqn]$ on the $[eqn]$ is analyzed for compounds containing modifiers at the same $[eqn]$ values.

According to Figure 8d, EOR-13-based compounds show no measurable influence of $[eqn]$ on $[eqn]$ and $[eqn]$ . This is consistent with the comparable $[eqn]$ obtained from DMTA of the pure modifiers and with their tensile test properties. The lower $[eqn]$ of EOR-13-1 compared to EOR-13-0.5, resulting from β-scission, appears to be compensated by the presence of the additional high- $[eqn]$ chains formed by cross-linking. Despite differences in yielding behavior, both materials exhibit similar deformability ( $[eqn]$ ) and comparable $[eqn]$ and $[eqn]$ , reflecting a balanced combination of strength and ductility. Overall, the differences in the MWD of EOR-13-1 and EOR-13-0.5 do not lead to notable changes in the $[eqn]$ of the corresponding compounds across the entire $[eqn]$ range.

For PA-EOR-15-0.5 and PA-EOR-15-1, the mechanical responses at low $[eqn]$ ( $[eqn]$ ≤ −10 °C) are essentially identical (see Figure 8e), which is consistent with DMTA results indicating that grafting does not measurably affect the deformability ( $[eqn]$ ) of the modifiers. At more elevated $[eqn]$ ( $[eqn]$ > −10 °C), however, PA-EOR-15-1 exhibits lower $[eqn]$ and $[eqn]$ than PA-EOR-15-0.5, in agreement with the tensile test results of the pure modifiers. While $[eqn]$ remains similar, $[eqn]$ and $[eqn]$ of EOR-15-1 are markedly lower than those of EOR-15-0.5. Consequently, the strength–ductility balance is compromised, leading to lower $[eqn]$ and $[eqn]$ . The lower $[eqn]$ at higher $[eqn]$ is attributed to increased polar intermolecular interactions and steric hindrance, which restrict chain sliding in PA-EOR-15-1 compared to PA-EOR-15-0.5, particularly in the amorphous phase. As a result, in the latter system, the triaxial stress state is more efficiently converted into a predominantly uniaxial one, enabling the formation of a more extensive shear-yielding zone ahead of the crack tip due to its higher deformation, until failure at comparable particle strength. This larger shear-yielding zone involves the deformation of a greater number of EOR particles, thereby increasing $[eqn]$ , while higher local stresses cause premature particle failure, limiting $[eqn]$ . We do not attribute the observed differences between PA-EOR-15-1 and PA-EOR-15-0.5 to changes in the MWD, as no such effect was observed for the PA-EOR-13 systems. The fact that an increase in $[eqn]$ has no effect on the $[eqn]$ in PA-EOR-13 is attributed to the higher $[eqn]$ of EOR-13. This results in a higher $[eqn]$ and a smaller contribution of the amorphous regions to $[eqn]$ .

In contrast, the differences in $[eqn]$ and $[eqn]$ observed between PA-EOR-16-1 and PA-EOR-16-0.5 (Figure 8f) are attributed to the significantly reduced $[eqn]$ of the modifier EOR-16-1. As demonstrated by GPC, the grafting process of EOR-16-1 is dominated by β-scission, rather than cross-linking, resulting in a pronounced decrease in $[eqn]$ , particularly in $[eqn]$ . Consequently, the mechanical performance is inferior, as reflected by lower $[eqn]$ , $[eqn]$ and $[eqn]$ . As a result, $[eqn]$ is significantly reduced, starting just above the BDTT (≈−15 °C) and $[eqn]$ is significantly reduced compared to PA-EOR-16-0.5 over the entire $[eqn]$ range. The combination of reduced $[eqn]$ and $[eqn]$ promotes premature particle failure in PA-EOR-16-1, limiting the development of an extensive shear-yielding zone in the matrix and thus reducing $[eqn]$ . Premature particle failure also lowers the energy dissipated during crack initiation, resulting in a decreased $[eqn]$ , even at lower $[eqn]$ , despite the fact that the modifier deformability, as indicated by $[eqn]$ from DMTA, remains largely unchanged.

Finally, we compare the influence of $[eqn]$ on the $[eqn]$ for compounds with the same $[eqn]$ . At $[eqn]$ = 0.5 wt% (Figure 8b), increasing $[eqn]$ leads to a slight increase in $[eqn]$ , up to $[eqn]$ = 0 °C. In the same $[eqn]$ range but above the BDTT, a measurable increase in $[eqn]$ is also observed, with the effect being most pronounced for the highest $[eqn]$ of 16 mol%. Since the differences in MWD at $[eqn]$ = 0.5 wt%, particularly with respect to the high- $[eqn]$ fraction, are small (see Table 1 and Figure S4, Supporting Information), the enhanced energy dissipation cannot be attributed to MWD effects. Instead, this behavior is attributed to the increased deformability, as indicated by the lower $[eqn]$ obtained from DMTA, which increases with decreasing $[eqn]$ . In addition, the higher $[eqn]$ with higher $[eqn]$ , likewise associated with reduced $[eqn]$ , appears to compensate for the concomitant reduction in $[eqn]$ , assuming that the tensile properties are representative of the mechanical behavior in this $[eqn]$ range.

Starting at approximately 0 °C, EOR-16-0.5 begins to lose its effectiveness in impact toughening, as reflected by reductions in both $[eqn]$ and $[eqn]$ . At 23 °C, a slight decrease in these contributions to energy dissipation is observed for all compounds, including those with $[eqn]$ = 1.0 wt%. This decline is attributed to a reduction in particle strength caused by the progressively increasing fraction of molten modifier. This effect becomes more pronounced with increasing $[eqn]$ , which is consistent with the DSC results.

At $[eqn]$ = 1.0 wt% (Figure 8c), $[eqn]$ of PA-EOR-15-1 and PA-EOR-16-1 shows only a slight increase, just above the BDTT, and falls below that of PA-EOR-13-1 from 0 °C onward. $[eqn]$ of PA-EOR-16-1 is reduced to values that are comparable to PA-EOR-13-1 up to −15 °C and becomes lower at higher $[eqn]$ . For PA-EOR-15-1, $[eqn]$ remains higher than that of PA-EOR-13-1, as observed at $[eqn]$ = 0.5 wt%, but also drops below it above 0 °C.

At $[eqn]$ = 0.5 wt%, changes in $[eqn]$ as a function of $[eqn]$ can be mainly explained by the decrease in $[eqn]$ with increasing $[eqn]$ . At $[eqn]$ = 1.0 wt%, however, the differences between compounds with varying $[eqn]$ originate from different mechanisms. Up to approximately 0 °C, the behavior of PA-EOR-15-1 is similar to that of PA-EOR-15-0.5. At higher $[eqn]$ , the response of PA-EOR-15-1 is dominated by a reduction in $[eqn]$ , which is attributed to increased polar intermolecular interactions and steric hindrance that restrict chain sliding, while the particle strength remains comparable to PA-EOR-15-0.5. The inferior performance of PA-EOR-16-1 can be particularly attributed to a reduced $[eqn]$ combined with insufficient $[eqn]$ , resulting from $[eqn]$ degradation during modifier grafting together with restricted chain sliding in the amorphous phase.

In summary, while for $[eqn]$ between 8 mol% and 15 mol% at a constant $[eqn]$ = 1.0 wt%, it was sufficient to consider $[eqn]$ -dependent differences in elastic deformability to explain the changes in $[eqn]$ [41]; this approach is no longer sufficient to explain the $[eqn]$ -dependent changes at $[eqn]$ ≥ 15 mol%. In particular, the changes observed from the BDTT onward, upon increasing the $[eqn]$ content from 0.5 wt% to 1.0 wt%, can only be explained by a reduction in plastic deformability resulting from enhanced polar interactions, together with the consideration of the adequacy of the modifier strength.

4. Conclusions

This study demonstrates that the $[eqn]$ of PA6/EOR-g-MAH compounds is governed by the interplay between $[eqn]$ , $[eqn]$ , and the competition between β-scission and cross-linking reactions during grafting. By keeping the initial modifier $[eqn]$ comparable and the blend microstructure similar, the influence of grafting-induced changes in the MWD and modifier properties on impact performance could be isolated and clearly evaluated. The instrumented impact test was valuable, as it distinguished $[eqn]$ from $[eqn]$ , revealing the parameters and their correlations controlling compound toughness at different $[eqn]$ .

MAH grafting shifts the MWD toward lower $[eqn]$ due to β-scission, with degradation becoming more pronounced at higher $[eqn]$ and $[eqn]$ . At $[eqn]$ = 0.5 wt%, chain scission dominates for all modifiers, while cross-linking remains negligible. At $[eqn]$ = 1.0 wt%, the balance becomes strongly $[eqn]$ -dependent: EOR-13 and EOR-15 show some compensation of degradation by cross-linking, whereas for EOR-16, β-scission clearly prevails, leading to a pronounced reduction in $[eqn]$ . These $[eqn]$ changes are important because the original materials cannot be easily replaced due to economic and availability constraints. Moreover, shifting the MWD to higher $[eqn]$ may not compensate for β-scission as effectively as the small fraction of very long chains that preserve the modifier performance.

These $[eqn]$ scenarios directly reflect the observed impact performance. While, for $[eqn]$ between 8 mol% and 15 mol% at a constant $[eqn]$ = 1.0 wt%, the changes in $[eqn]$ can be adequately rationalized by $[eqn]$ -dependent differences in elastic deformability [41], this approach is no longer adequate to explain the $[eqn]$ -dependent changes at $[eqn]$ ≥ 15 mol%. In particular, the changes in toughness observed from the BDTT onward upon increasing the $[eqn]$ from 0.5 wt% to 1.0 wt% require considering both the reduction in plastic deformability caused by enhanced polar interactions, particularly restricting chain sliding in the amorphous phase, and the adequacy of the modifier strength. As β-scission is partially compensated by cross-linking in EOR-15, the impact response of PA-EOR-15-1 is primarily determined by the loss in ductility. In contrast, for EOR-16, $[eqn]$ degradation dominates and, together with reduced ductility, results in the inferior toughness of PA-EOR-16-1. Additionally, reduced $[eqn]$ and the earlier onset of successive melting further contribute to the diminished performance of PA-EOR-16-1 and PA-EOR-15-1 compared to PA-EOR-13-1 and -0.5. Both PA-EOR-13 variants exhibit nearly $[eqn]$ -independent performance, as grafting-induced cross-linking maintains strength, while ductility remains largely unaffected by $[eqn]$ due to the higher $[eqn]$ .

At $[eqn]$ = 0.5 wt%, changes in $[eqn]$ as a function of $[eqn]$ can be mainly explained by the decrease in $[eqn]$ with increasing $[eqn]$ , thereby enhancing the modifiers’ elastic and plastic deformation capability and their ability to promote cavitation-driven shear yielding. This effect increases the $[eqn]$ of PA-EOR-15-0.5 and PA-EOR-16-0.5 at low and intermediate $[eqn]$ up to about 0 °C above the performance of both PA-EOR-13 variants. Only for PA-EOR-16-0.5 at $[eqn]$ does the loss in strength due to successive melting produce a significant drop in $[eqn]$ compared to the other two compounds.

Overall, the impact performance of PA6/EOR-g-MAH compounds results from a balance between $[eqn]$ , $[eqn]$ , and grafting-induced $[eqn]$ changes. A moderate $[eqn]$ = 0.5 wt% and $[eqn]$ = 15 mol% offers an optimal compromise between strength, elastic deformability, and ductility, delivering low BDTT and high $[eqn]$ across a broad $[eqn]$ range, while also maintaining good compound processability by avoiding excessive viscosity caused by an increased number of PA6 chains that are covalently anchored to the modifier particles. While core–shell and other sophisticated polymer alloy concepts might achieve higher performance in certain applications, the approach presented here is aimed at reducing the material design complexity and processing requirements at a high cost efficiency.