Effects of Carbon Fillers on Electrical and Mechanical Properties of Water-Based Polymer Nanocomposites

TL;DR

This study explores how adding carbon fillers to a water-based polymer affects both electrical conductivity and mechanical properties, showing that the right mix can create flexible and conductive materials for electronics.

Contribution

A novel water-based method for creating rGO-MWCNT/polymer composites with tunable electrical and mechanical properties is introduced.

Findings

MWCNT-rich fillers achieved the highest conductivity (up to 8.2 × 10−3 Sm−1) due to a segregated filler network.

Mechanical properties like elongation at break increased significantly with specific filler ratios and loadings.

The study highlights the need for careful optimization of filler content and ratios for flexible electronics applications.

Abstract

Both the electrical conductivity and tailored mechanical characteristics—showing flexibility and structural integrity—are key properties of polymer composites. In this work, a novel, simple, and water-based strategy for synthesizing rGO-MWCNT/polymer composites was developed. Namely, carbon nanofillers in a mixture of reduced graphene oxide (rGO) and multi-walled carbon nanotubes (MWCNTs) were incorporated in a waterborne methacrylic polymer matrix at loadings of 0.25, 0.5, and 1.0 wt.% nanofiller, and with rGO-to-MWCNT ratios of 10:1, 1:1, and 1:10 (w/w) at room temperature. Electrically conductive composites were obtained with all tested filler rates showing the highest conductivity (up to 8.2 × 10−3 Sm−1) for the MWCNT-rich filler due to the formation of a segregated network of the filler in the matrix. The mechanical properties of the composites—characterized by their Young’s modulus and elongation at break—strongly depended on both the filler incorporation rate and the rGO:MWCNT ratio. For instance, soft and flexible composites were obtained by incorporating 0.25 wt.% of the MWCNT-rich filler, which increased the elongation at break from 154.2% (neat polymer) to 252.4%. Overall, this study emphasizes the sensitive interplay between carbon filler introduction incorporating conductivity and the fillers’ impact on the mechanical properties of a polymer composite, both necessitating careful optimization for applications, e.g., in flexible electronics.

Figures

Full text

1. Introduction

Wearable electronics are pushing the limits of human–machine interfaces. Current devices still rely on conventional rigid electronics, but flexible and stretchable systems are increasingly investigated to enable reliable data acquisition without restricting natural body movement [1,2]. A major challenge is designing materials that can endure large deformations while maintaining functionality and conforming to the skin. Advances in soft materials have enabled the development of mechanically deformable electronic systems with promising applications in biomedical engineering and flexible electronics [3,4]. Careful material selection and mechanical analysis are therefore essential to achieve conformal integration for monitoring pressure, strain, and bio-potentials. Soft polymers play a key role in improving flexibility and adhesion in wearable devices. Their mechanical performance, including stiffness and modulus, can be tailored through structural design or material combinations [5,6]. Polymer-based nanocomposites, which integrate flexible polymer matrices with nanoscale fillers, have emerged as attractive platforms for flexible electronics [7]. In addition to mechanical compliance, electronic applications require sufficient electrical conductivity. The incorporation of conductive nanofillers into polymer matrices provides an effective strategy for achieving this functionality [8]. Carbon-based nanomaterials—such as graphene (G), reduced graphene oxide (rGO), and carbon nanotubes (CNTs)—are widely explored as conductive fillers due to their outstanding electrical conductivity and mechanical characteristics [9,10,11,12,13]. Combining different carbon nanostructures into hybrid systems offers opportunities to tailor network formation, charge transport pathways, and deformation behavior [14,15,16,17,18,19,20,21].

Graphene-based materials provide a high specific surface area, good electrical conductivity, and mechanical flexibility. However, graphene sheets tend to restack due to strong van der Waals interactions, which reduces accessible surface area and limits effective network formation. In contrast, CNTs offer one-dimensional conductive pathways, high aspect ratios, and the ability to form percolated networks at relatively low loadings, although their dispersion in polymer matrices can be challenging. Integrating graphene derivatives with CNTs may help mitigate these individual limitations by promoting spatial separation of graphene sheets while facilitating network connectivity. The properties of graphene strongly depend on synthesis and reduction routes. Graphene oxide (GO) contains oxygen-containing functional groups that improve compatibility with polymers but reduce electrical conductivity. Partial reduction to rGO restores conductivity while retaining some functional groups beneficial for interfacial interactions [22,23,24]. CNTs, characterized by their hexagonal carbon structure and high aspect ratio, can contribute to electrical conductivity and influence stress distribution within polymer matrices [25,26,27].

Hybrid rGO–MWCNT systems have been widely investigated for enhancing electrical conductivity and modifying mechanical performance in polymer composites [28,29,30,31]. Reported improvements are often attributed to improved dispersion, interfacial interactions, and network formation [32,33,34,35,36]. CNTs can act as spacers between graphene sheets, potentially reducing restacking and facilitating the formation of interconnected conductive pathways [36,37,38,39]. However, it should be emphasized that hybridization does not inherently guarantee a synergistic mechanical reinforcement effect. The final properties strongly depend on filler ratio, dispersion state, interfacial adhesion, and processing method. Systematic evaluation of filler composition is therefore necessary to understand how hybrid architecture influences composite behavior [40,41]. Existing synthesis approaches for incorporating mixed carbon fillers frequently involve high temperatures, organic solvents, or complex processing routes that reduce sustainability and scalability [40,41,42,43,44,45,46]. Developing simple, low-temperature, and environmentally friendly fabrication methods remains an important objective.

In this work, a water-based strategy was developed to prepare rGO–MWCNT/polymer composites using aqueous dispersions. A waterborne MMA/BA/GMA copolymer (49.5/49.5/1 wt.%) was synthesized via seeded semi-continuous emulsion polymerization. This composition provides a glass transition temperature below room temperature, enabling spontaneous film formation under ambient conditions [47,48]. Hybrid filler dispersions were incorporated at total loadings up to 1 wt.%, and the influence of rGO/MWCNT ratios (1:1, 1:10, and 10:1) on morphology and tensile behavior was systematically investigated [47,48].

Rather than presuming synergistic reinforcement, the present study aims to elucidate how hybrid carbon architectures influence deformation mechanisms and mechanical response in waterborne polymer composites intended for flexible electronic applications.

2. Materials and Methods

Materials: Multi-walled carbon nanotubes (MWCNTs) with a length between 5 and 15 μm and diameters of 10–30 nm purchased from Sigma-Aldrich (St. Louis, MO, USA), an aqueous dispersion of graphene oxide (GO, 5 mg/mL, 80%) from ACS material, and ascorbic acid (ASA, 99%) from Acros-Organics (The Hague, The Netherlands) were used to prepare the fillers. Polyvinyl pyrrolidone (PVP, 10,000 g/mol) for rGO stabilization was purchased from Sigma-Aldrich. Alkyldiphenyloxide Disulfonate (Dowfax2A1, Dow Chemical Co., Midland, MI, USA), methyl methacrylate (MMA, 99.9%), and butyl acrylate (BA, 99.5%) from Qumidroga (Barcelona, Spain), glycidyl methacrylate (GMA, 97%) from Acros-Organics, and initiator potassium persulfate (KPS, 99%) from Sigma-Aldrich were used for polymer synthesis. Milli-Q ultrapure water was used for the preparation of dispersions and emulsions as a solvent.

2.1. rGO/MWCNTs Filler Preparation

The schematic representation of the synthesis of fillers is given in Figure 1. Before synthesis, GO dispersions in different amounts (Table 1) were sonicated to exfoliate the aggregated platelets using a bar sonicator (Sonifier 450, Branson Ultrasonics Corporation, Brookfield, CT, USA) for 15 min at 80% power on a 50% duty cycle and agitation of 200 rpm (Figure 1a). Separately, varying amounts of MWCNTs (Table 1) were sonicated in air for 90 min at 70% power, and a 50% duty cycle under the agitation of 200 rpm was performed to separate any agglomerated MWCNT bundles (Figure 1b). More details on this process can be found elsewhere [49].

The MWCNT pre-sonicated in air was added in the respective amounts to the pre-sonicated GO suspension and sonicated for an additional 2.5 min at 80% power, with a 50% duty cycle under 200 rpm agitation (Figure 1c). To obtain reduced graphene oxide, the GO/MWCNTs filler was subjected to chemical reduction. The reduction was performed as follows: first, 5 wt.% PVP was added to the GO/MWCNTs filler to colloidally stabilize the reduced nanoparticles, and the mixture was stirred for 30 min (Figure 1d). Then ASA was added according to the following ratio: solid reductant to filler = 5:1 (Figure 1e). To reduce the water content, the mixtures of 10:1 w/w and 1:1 w/w (GO/MWCNTs) were kept in an oven at 90 °C for 60 min. The 1:10 mixture was only kept at room temperature for 12 h because of its already low water content due to the low amount of GO dispersion used (Figure 1f). After the reduction, all three rGO/MWCNTs fillers were subjected to dialysis to remove the excess of PVP (Spectrum Labs membrane, molecular weight cut-off membrane (MWCO) 12–14 kDa).

2.2. Polymer Latex Preparation

For the synthesis of polymer dispersions as shown in Figure 2, seeded semi-continuous emulsion polymerization was employed. To obtain 150 g of polymer seed dispersion with 10% solids content, Dowfax 2A1 (2 wt.% based on the monomers), 134 g water, 7.5 g MMA, and 7.5 g BA were stirred in a 250 mL reactor for 10 min. The temperature was raised to 70 °C, and the initiator KPS aqueous solution (0.5 wt.% based on monomers) was added. The reaction was performed in a nitrogen atmosphere in a batch mode for 3 h (Figure 2a). Then, the prepared seed was used for the synthesis of the final polymer latex dispersion in a semi-continuous process. To obtain 900 g of final dispersion with 50% solid content, MMA/BA seed (27 g) and water (110 g) were stirred in a 1000 mL reactor for 10 min. When the temperature of the reactor reached 70 °C, an aqueous solution of KPS (2.24 g in 15.7 g of water) was inserted into the reactor. Subsequently, feeding of an aqueous solution made of Dowfax 2A1 45% (2 wt.% based on monomer), water (300 g), and a monomer mixture of MMA/BA/GMA in amounts 221.4 g/221.4 g/4.5 g was performed at a rate of 4.2 g/min. The reaction was performed in a nitrogen atmosphere, in semi-batch for 3 h, and afterwards in batch for 2 h to achieve full monomer conversion (Figure 2b).

2.3. Preparation of Nanocomposite Dispersions

The rGO/MWCNTs/polymer composites were prepared by mixing the dispersions. Namely, appropriate amounts of the three rGO/MWCNTs (10:1 wt.%, 1:1 wt.%, and 1:10 wt.%) aqueous dispersions were added into polymer latex dispersion (0.25 wt.%, 0.5 wt.%, and 1 wt.% based on the solid content), so dispersions with different rGO/MWCNT concentrations were obtained. After the addition of the fillers, the mixtures were continuously stirred at room temperature for 3 h at 150 rpm.

2.4. Characterization

The method of specimen preparation and characterization of filler and composites is described in Figure 3.

The filler dispersions were drop-coated on a glass substrate and left to dry for 12 h before performing Raman and conductivity measurements. A confocal Raman microscope from Renishaw was used to perform measurements in backscattering geometry using a Labram HR 800 Horiba Jobin Yvon spectrometer (HORIBA, Kyoto, Japan) equipped with a 600 mm^−1^ grating and thermoelectrically cooled charge-coupled device detector for investigation of the chemical composition and structure of the rGO/MWCNTs filler samples. The Raman scattering was exciting with a 532 nm line of a doubled frequency Nd-YAG laser using laser power of ca. 2.5 mW. By passing the laser beam through a microscope objective (Olympus, Tokyo, Japan), the linearly polarized laser light was focused on the surface of the rGO/MWCNTs specimen. Each spectrum was recorded using 30 accumulations with an exposure time of 5 s. With a four-point handheld probe (FPP 5000 Miller, Veeco-Miller Design, Azusa, CA, USA), the conductivities of the different fillers were measured. Furthermore, Transmission Electron Microscopy (TEM, Tecnai G2 20 Twin, FEI Company, Waltham, MA, USA) was used to observe the morphology of the rGO/MWCNTs fillers drop-coated on TEM grids.

Solid samples from unfilled polymer and the composites were prepared using silicone molds in which the latexes were cast and dried at 25 °C and 55% relative humidity (standard atmospheric conditions). The obtained solid samples were cut with a die cutter for further analysis. Scanning Electron Microscopy (SEM, Quanta 250 e-SEM, Philips Tecna, Amsterdam, The Netherlands) was employed to observe the morphology of the neat polymer and nanocomposite samples. Cross-sectional cuts were prepared by breaking samples previously frozen in liquid nitrogen.

To determine the mechanical properties of the nanocomposite materials, tensile specimens (15 mm × 3.5 mm × 0.5 mm) were cut, and the measurements were performed using TA.HD plus texture analyzer (Stable Micro Systems Ltd., Godalming, UK) with a constant strain velocity of 1.5 mm s^−1^. The length between the jaws was 15 mm, and the experiment’s nominal strain rate was 0.1 Hz. Young’s modulus and Offset Yield stress were estimated.

The unfilled polymer and the composites were spin-coated (600 rpm for 20 s followed by 3000 rpm for 6 s) on gold-coated interdigitated electrodes. Afterwards, all composite films were dried under ambient conditions for 2 h. The thickness of composite films was measured at ~10 ± 0.3 µm using a 3D laser microscope (LEXT OLS4000-Olympus, Tokyo, Japan), while their conductivities were determined with a handheld multimeter. All the measurements were carried out at room temperature.

3. Results and Discussion

3.1. Filler Properties

In the preparation of rGO/MWCNTs fillers, three different weight ratios of rGO:MWCNTs, 10:1, 1:1, and 1:10 (w/w), were selected. As the aqueous dispersions of the rGO/MWCNTs fillers were stabilized with PVP, no precipitation was observed even after longer storage times (Figure 4a). A similar approach to rGO stabilization in aqueous dispersion was demonstrated by Arzac et al. [46] using the same stabilizing agent (PVP).

The investigation of the chemical composition and structure of the rGO/MWCNTs fillers before composite formation was carried out with Raman spectroscopy. The Raman spectra of rGO/MWCNTs fillers with different ratios of the rGO and MWCNTs are shown in Figure 4b. Raman results for all rGO/MWCNTs fillers demonstrate the presence of four different vibration peaks G, D, 2D, and D + D′ that correspond to sp^2^ hybridized carbon (G peak at ~1590 cm^−1^), sp^3^ hybridized carbon (D and 2D at ~1350 and 2702 cm^−1^), and combination mode (D + D′) at 2945 cm^−1^ as a sign of atomic insertions. The 2D band of rGO/MWCNTs may be attributed to the presence of rGO on the MWCNTs, and the D′ band is caused by defects generated by an internal double-resonance process.

The intensity ratio of the signal D and G (I_D_/I_G_) is presented in Table 2; the ratios reflect the level of defects or disorder versus graphitic order in carbon materials, with higher values indicating more defects and lower values indicating a more ordered structure. In our case, the values are similar for all samples, and they are relatively high, which is likely due to a high number of defects in both carbon structures of the fillers [50]. This fact simply illustrates the significant presence of still non-reduced-oxygen functional groups [51].

The electrical conductivities of the dried filler mixtures are given in Table 2. Strong π-π interfacial coupling between the MWCNTs and the rGO that promotes charge carrier mobility between both rGO and MWCNTs, as well as van der Waals interactions between individual rGO nanoparticles and MWCNTs nanoparticles, presumably occurs. The conductivity increases with increasing MWCNT content, presumably due to higher conductivity and longer length of the MWCNT compared to the rGO flake size, both enabling more and efficient percolation pathways.

To evaluate the structure of the fillers, they were subjected to TEM analysis. Figure 5 presents the TEM images of the rGO:MWCNTs fillers with different ratios of rGO and MWCNTs. The result clearly illustrates that rGO platelets are well exfoliated in the dispersion, often single-layered, and CNTs are mostly disentangled and individual. The MWCNTs probably interact with the rGO by building stacking π-π interactions or even connecting a few rGO platelets. By increasing the MWCNT quantity (from Figure 5a–c), the nanotubes are more aggregated and entangled, covering the rGO platelets. The better interaction between them and the most likely higher conductivity of the MWCNT are presumably responsible for the observed increasing conductivity of the filler with increasing MWCNT quantity.

3.2. Nanocomposite Films

The nanocomposites containing 1 wt.% fillers with different rGO:MWCNT ratios (1:10, 1:1, and 10:1(w/w)) were studied. Furthermore, in the case of the 10:1 (w/w) rGO:MWCNT, the composites were prepared to contain different loadings of the fillers of 0.25, 0.5, and 1 wt.%. Figure 6 shows photos of all the composites. The darker color of the composites was observed with increasing the MWCNTs quantity compared to the other films due to the homogenous dispersion of rGO and MWCNTs into the polymer. It is worth mentioning that after mixing polymer nanoparticles and the filler mixtures in colloidal aqueous dispersion, the platelets and the MWCNTs are adsorbed on the surface of the polymer particles. This configuration avoided or at least minimized the aggregation of the inorganic fillers during the composite films’ formation with water evaporation and preserved a homogenous filler distribution in the dried composite, as shown by SEM investigations below.

To evaluate this possibility and the size and interactions of the rGO:MWCNTs nanoparticles in the composites, cross-section cuts of the prepared composite films were subjected to SEM analysis (Figure 7, Figure A1 and Figure A2 (please check Appendix A)).

Figure 7 shows the SEM images of the composites. As can be seen in all composites, the carbon nanofillers with embedded white-grey structures are homogenously distributed into the polymer matrix with a dark continuous appearance. Nevertheless, MWCNTs appear as thinner and larger structures, as is shown in Figure 7e, where the filler contains a tenfold higher amount of MWCNTs than rGO. This morphology of a segregated network, like the distribution of the graphene-based filler into the polymer matrix, is typical for waterborne composites [52]. However, it is worth noting that at a low filler amount (0.25 and 0.5 wt.%, shown in Figure 7a,b), no aggregation was observed in the morphology of rGO-MWCNTs. As a result, the distribution of the fillers is much more homogeneous, whereas the composites containing 1 wt.% filler platelets have a more inhomogeneous structure and larger areas of neat polymer films and aggregates. The exception to this may be observed in the case of a 1:10 (w/w) ratio filler (Figure 7e), showing a very nice, segregated network with an abundant amount of MWCNTs presented.

The electrical conductivities of the composites are given in Table 3. Compared to the neat polymer, the conductivity of all composites was enhanced by several orders of magnitude. Interestingly, increasing the filler content of rGO:MWCNTs (10:1) within the composites from 0.25% to 1% does not significantly influence their conductivity. Possibly, the higher filler content cannot generate more percolation pathways due to filler aggregation as observed in SEM. Obviously, at a content of 0.25%, most of the percolation pathways have been formed. When increasing the MWCNTs concentration, the electrical conductivity of the composites strongly increased by an order of magnitude comparable with the conductivities of the filler alone.

Figure 8a presents the tensile stress–strain behavior of the nanocomposites in dependence on the loading range of 10:1 filler into a polymer under stress, whereas in Figure 8b, the effect of different ratios of rGO and MWCNTs within the filler (10:1, 1:1, and 1:10 (w/w)) in the polymer matrix is given. The mechanical properties determined from these curves are given in Table 4.

Figure 8 shows the significant strain hardening for all samples according to the results given in Table 4. Surprisingly, the neat polymer presents a very high value of Young’s modulus, high offset yield and maximum stress, and low strain rate at the maximum. The reason may be the presence of epoxy groups of GMA monomers with the ability to react with many functional groups, which makes the polymer stiffer. At low loading of the fillers (0.25 and 0.5 wt.%) in the polymer matrix, surprisingly, the composites are less stiff but softer than the neat polymer and much more stretchable, showing elongation of 304.6% and 238.4%, respectively. By increasing the loading range from 0.25 wt.% to 1 wt.%, the elongation decreased continuously to 154.9%, which confirms the stronger interaction between filler and matrix. Furthermore, the elongation at break significantly increased to 252.4% and 243.6% by incorporation of MWCNT-rich fillers of 1:1 and 1:10 into the polymer, respectively.

At lower filler loadings, the composite exhibits a softer mechanical response due to the limited polymer–filler interfacial area and the absence of a percolated filler network. The reduced interfacial contact results in a smaller fraction of immobilized polymer chains, allowing greater matrix chain mobility. In addition, isolated filler particles at low concentrations are less effective in stress transfer, causing the mechanical behavior to be dominated by the polymer matrix rather than filler reinforcement.

A lack of interaction between the filler and the polymer phases is usually the reason for the mechanical weakening of the polymer composites compared to the neat polymer.

Nevertheless, the impressive improvement in the flexibility of the same composites, almost threefold, eliminates this possibility and indicates excellent interaction.

Compared to the neat polymer, for all rGO-MWCNTs (10:1) composites, the stress at break and the strain at break increased, while Young’s modulus and the offset yield stress decreased. This behavior is opposite to that obtained in similar materials with individual rGO [52] and MWCNTs [53], in which cases Young’s modulus of polymer always increased with the loading of the fillers [53,54]. In our case, as observed in the SEM and postulated during blending, the filler containing mainly rGO has finely dispersed and binds presumably by a small number of hydrogen bonds to the polymer latex surface before and after drying as described above. This may hinder polymer–polymer interaction, especially the entanglement of the polymer chains, in which the rGO-rich filler behaves in small concentrations as a plasticizer. This fact leads to a decreased strength of the matrix and an increased elongation at break, which confirms the ductility of the material and its ability for deformation without a break. By increasing the loading weight ratio to 1 wt.% rGO-MWCNTs (10:1) within the composite, the value of all tensile characteristics is close to that of the neat polymer again. As the SEM image indicated in this case, the distribution of the rGO-rich filler is not homogenous for 1 wt.%, as in the 0.25 wt.% and 0.5 wt.% samples, indicating agglomeration of the filler. Therefore, this sample shows similar tensile properties to the neat polymer and higher stiffness. However, the elongation at break is not reduced by the addition of a higher filler ratio, as we expected; the reason is due to the polymer–filler interaction and stress transferring.

Also, when going from rGO-rich fillers to MWCNTs-rich fillers, a plasticizing effect can be observed, which makes the film softer and more flexible. Here, it can be assumed that, even when the filler material is finely distributed, the disturbance of the polymer–polymer interaction (i.e., entanglement) is probably smaller due to the 1D structure compared to the 2D structure of rGO. However, MWCNTs are assumed to form a weak interfacial interaction with the polymer compared to rGO, due to fewer hydrophilic groups on their surface. This fact induces the debonding of nanoscale fillers from the polymer matrix during loading, possibly weakening the polymer matrix as indicated by the decreased strain at the break with increasing the MWCNTs content [55,56,57,58,59]. While pristine polymer exhibits a modulus of 0.19 MPa, the modulus of all composites except 1 wt.% rGO:MWCNTs 10:1 (0.21 MPa) decreased. A more relevant increase in stress and strain at break has been observed in the case of samples containing rGO:MWCNTs 1:1 (w/w) filler, with its stress at break of 14.26 MPa and strain at break of 3.18 MPa. From the enhancement in tensile properties of the polymer after the addition of all filler samples, it can be concluded that all samples behave as soft and ductile materials. Although all samples demonstrated close values, the samples containing 0.25 wt.% rGO:MWCNTs 10:1 (w/w), 0.5 wt.% rGO:MWCNTs, and rGO:MWCNTs 1:10 (w/w) fillers demonstrated a combination of lower Young’s modulus and higher stress and strain values, which can be considered as a softer material with higher viscoelastic properties among other introduced composites.

4. Conclusions

Conductive nanocomposite films based on a waterborne methacrylic polymer matrix and hybrid carbon nanofillers were successfully prepared using a fully aqueous processing route. The mechanical and electrical properties of the films were tailored by varying both the rGO/MWCNT weight ratio and the total filler loading (0.25–1 wt.%). Morphological investigations (TEM and SEM) demonstrated that increasing the MWCNT content led to more pronounced nanotube entanglement and partial coverage of rGO platelets, influencing the formation of conductive pathways. The electrical conductivity of both the mixed fillers and the composite films increased with higher MWCNT fractions, reaching the highest values for the MWCNT-rich system (rGO:MWCNT = 1:10, w/w). These results indicate that MWCNTs play a dominant role in establishing effective charge transport networks within the polymer matrix. In terms of mechanical behavior, the incorporation of hybrid fillers modified the deformation response of the polymer rather than producing classical stiffness reinforcement. Composites containing 0.25 wt.% and 0.5 wt.% rGO-rich fillers (10:1) exhibited a reduction in Young’s modulus accompanied by an increase in elongation at break. Similar trends were observed for MWCNT-rich systems, resulting in softer and more flexible films compared to the neat polymer. The observed mechanical response suggests that the hybrid rGO/MWCNT architecture influences stress distribution and fracture propagation within the matrix. While rGO sheets provide extended interfacial contact with polymer chains, MWCNTs may act as bridging elements that alter filler–filler interactions and reduce rigid aggregation (Figure 9). This structural arrangement likely promotes energy dissipation and enhanced chain mobility, leading to decreased stiffness and improved ductility.

Overall, the results demonstrate that hybrid carbon nanofillers can be used to tune the balance between electrical conductivity and mechanical flexibility in waterborne polymer composites. The developed films exhibit promising characteristics for flexible and wearable electronic applications, where mechanical compliance and conductivity must be simultaneously achieved. Future studies involving direct comparison with single-filler systems will further clarify the extent of hybrid effects in such materials.