Unlocking the optimization of electrical properties of epoxy-carbon nanotube composites modified with waste eggshell particles via the Taguchi-Grey method

Abstract

Advanced electronics demand materials that combine high thermal conductivity with enhanced electrical properties, yet achieving these improvements simultaneously poses significant challenges. This research employs the Taguchi-Grey methodology to explore the synergistic effects of dielectric elements and the high thermal conductivity of epoxy-based composites reinforced with carbon nanotubes (CNTs) and repurposed eggshell particles (ESp). Composite production involved solution blending, followed by evaluations of dielectric constant, thermal conductivity, and sample morphology. Under optimal conditions—1 wt% ESp, 2.5 wt% CNTs, curing at 90°C for 6 h—substantial increases in electrical and thermal conductivity of 19.130% and 94.27%, respectively, were achieved. These enhancements are attributed to the synergistic interaction between dielectric materials and CNTs, as well as the uniform CNT dispersion facilitated by the repurposed eggshells. The 95% confidence level confirmed a strong alignment between the predicted and experimental grey relational grades (GRG), validating the identified optimal parameters. This study demonstrates the potential of using repurposed eggshells to produce conductive polymers with uniformly dispersed CNTs, significantly enhancing thermal conductivity. These findings suggest a promising approach for sustainable, high-performance dielectric materials for electronic applications.

Introduction

The rapid miniaturization of electronic devices and the corresponding increase in power consumption have led to significant heat dissipation challenges, as most polymer dielectrics exhibit poor thermal conductivity. This limitation hinders their extensive use in advanced electronic applications [1–3]. A key challenge for engineers and researchers is achieving an optimal balance between the required insulating properties of dielectric materials and enhanced thermal conductivity. Studies have shown that the incorporation of thermally conductive fillers can greatly improve both the thermal conductivity and dielectric characteristics of composite materials [4–7]. A wide range of fillers, including carbon nanotubes (CNTs), metal powders, carbon fibers, graphite, carbon black, and metal fibers, have been investigated for use in conductive polymers [8–10]. Of these, CNTs have demonstrated exceptional potential in significantly enhancing the performance of conductive polymer composites [11].

Carbon nanotubes (CNTs) are typically formed by rolling graphene sheets with asymmetric helicity into concentric cylinders. The sp²-hybridized carbon atoms within CNTs provide them with remarkable chemical stability, along with exceptionally high electrical and thermal conductivity [11–14]. Due to their low density, excellent thermal properties, and high aspect ratio, CNTs have become increasingly prominent in the electronics industry, where they offer significant performance enhancements [12, 14]. However, a persistent challenge in fabricating polymer composites with CNTs is their tendency to segregate and aggregate, driven by strong Van der Waals forces [8, 9, 15, 16]. This aggregation compromises the mechanical, thermal, and electrical properties of the composite materials.

To mitigate these issues, researchers have explored hybridizing CNTs with inorganic materials, which has shown promising results in improving both the mechanical and electrical performance of polymer composites. Hao et al. [15], for instance, demonstrated significant improvements by decorating CNTs with silica and boron nitride nanosheets for epoxy composites, achieving a thermal conductivity of 0.68 W/m⁻¹ K⁻¹—an impressive 187% increase compared to pure epoxy. Similarly, Aigbodion [17] explored the use of green-synthesized silver nanoparticles (AgNPs) to enhance the dielectric properties of CNT-epoxy nanocomposites. By incorporating varying concentrations of CNTs (0.1%–0.5%) along with 0.5% AgNPs into the epoxy matrix, the study showed that the combination of 0.5% AgNPs and 0.5% CNTs yielded the highest electrical conductivity. Further work by Hao and Yu [18] demonstrated that polypyrrole-coated multi-walled carbon nanotubes (MWCNTs) resulted in notable increases in both the dielectric constant and dielectric loss. Wang et al. [19] showed that functionalized MWCNTs, when combined with diethylenetriamine and boron nitride nanosheets, exhibited superior thermal conductivity and reduced dielectric loss. In another study, Salehi et al. [20] highlighted that the surface functionalization of CNTs with nitric acid (HNO₃) and stearic acid (SA) led to enhancements in the thermal, mechanical, and morphological properties of high-density polyethylene (HDPE) nanocomposites, resulting in notable improvements in tensile strength and thermal performance. While these findings highlight the potential of various inorganic fillers, many require complex surface modifications to achieve proper CNT dispersion within the polymer matrix. This not only raises production costs but can also weaken the mechanical properties of the composite. Additionally, certain inorganic fillers can increase the stiffness or brittleness of the material, making them less suitable for applications that require both flexibility and durability [17, 21, 22].

Calcium carbonate (CaCO₃), on the other hand, stands out due to its cost-effectiveness, ease of availability, and excellent compatibility with CNTs [23, 24]. Unlike other inorganic fillers, CaCO₃ is more naturally suited to improve CNT dispersion without necessitating extensive chemical modifications. Its fine particulate nature allows it to prevent CNT aggregation, thereby enhancing the overall mechanical, electrical, and thermal properties of the composite. Furthermore, CaCO₃ is highly stable at high temperatures and does not negatively affect the flexibility of the composite material, making it ideal for use in a wide range of electronic and electrical applications. Several studies have explored the use of CaCO₃ as a key inorganic filler to enhance the dispersion and performance of CNTs within polymer matrices, with particular attention to addressing CNT agglomeration. For instance, Shen and Zhu [24] investigated the effect of CaCO₃-coated CNTs in a polyethylene (PE) matrix, reporting a significant improvement in electrical conductivity at just 0.5 weight percent CNT loading. This suggests that CaCO₃ promotes better CNT dispersion, which in turn enhances the electrical properties of the composite. Similarly, Backes et al. [25] studied the incorporation of CNTs combined with various inorganic particles—such as calcium carbonate, sepiolite, and montmorillonite—into epoxy composites. Among these, CaCO₃ was found to substantially improve the electrical conductivity of the composite, while other fillers like sepiolite had the opposite effect, demonstrating the superior conductive potential of CaCO₃-based composites.

Li et al. [26] contributed further to this line of inquiry by developing polypropylene (PP)/CNT nanocomposites with CaCO₃ particles. Their results highlighted an increase in electrical conductivity and a reduction in the percolation threshold, reinforcing CaCO₃’s effectiveness in facilitating CNT dispersion and network formation within the polymer matrix. This finding is particularly critical in applications where lower percolation thresholds are required to maintain the composite’s insulating properties while improving conductivity. Landau et al. [27] provided additional insights by examining the synergistic effects of CaCO₃ and nickel nanoparticles on CNT composites. Their study revealed that the catalytic action of CaCO₃ promoted the partial combustion of carbon within MWCNTs, leading to the formation of nano-onion-like structures with enhanced conductive properties. This transformation underscores the role of CaCO₃ in improving CNT dispersion and in catalytically altering CNT structures to improve overall material performance. Furthermore, Md Saleh et al. [3] evaluated CaCO₃-filled CNT hybrid epoxy composites, concluding that CaCO₃ played a pivotal role in enhancing CNT transport capabilities, leading to superior dispersion throughout the matrix. This growing body of literature supports the role of CaCO₃ in ensuring the uniform distribution of CNTs, which is critical for the successful manufacture of conductive polymers [28]. However, one challenge associated with the widespread use of CaCO₃ lies in its significance to global cement production, which drives up the cost of creating conductive polymers when relying heavily on this resource [29, 30].

In response to this challenge, the present study aims to reduce the reliance on conventional limestone-derived CaCO₃ by exploring waste-based sources of CaCO₃, such as eggshells, which contain 5–15% calcium oxide (CaO), as observed by Heidari et al. [31]. The goal is to examine the electrical properties of epoxy/CNT composites using CaCO₃ derived from waste materials, such as leftover eggshell particles, to enhance both the thermal and electrical performance of the composites while improving CNT dispersion. This study focuses on optimizing process parameters for using CaCO₃ derived from waste sources, particularly eggshells, for the modification of CNTs in the production of conductive polymers. By leveraging waste materials, this approach not only offers a sustainable alternative but also holds the potential for improved composite performance.

Various methods have been employed to optimize the electrical properties of epoxy-carbon nanotube (CNT) composites. Traditional approaches often include mechanical mixing, solvent-assisted dispersion, and chemical modification of CNTs [32]. Mechanical mixing aims to enhance the uniformity of CNT distribution within the epoxy matrix, while solvent-assisted methods utilize organic solvents to improve CNT solubility and dispersion. Chemical modifications, such as functionalization of CNTs [17], have also been explored to enhance interfacial bonding between the CNTs and the epoxy, thereby improving electrical conductivity. However, these methods can be time-consuming and may not always yield optimal results in terms of performance and cost-effectiveness [32].

The Taguchi-Grey method presents a robust alternative for optimizing the electrical properties of composites [33, 34]. This approach combines the Taguchi method’s systematic experimental design with Grey relational analysis, which is particularly useful for dealing with multiple performance characteristics under uncertainty. The Taguchi method uses orthogonal arrays to minimize the number of experiments while maximizing information gained about parameter interactions. By integrating Grey relational analysis, researchers can effectively evaluate and rank multiple performance metrics simultaneously, simplifying the optimization process. This dual approach allows for a comprehensive assessment of how various factors influence the electrical properties of polymer composites, ultimately leading to enhanced performance with fewer experimental trials.

Materials and methods

Materials

The carbon nanotubes (CNTs) used in this study, with diameters ranging from 10 to 40 nm and lengths between 10 and 20 µm, were procured from Hongwu International, China, and employed in the experimental procedures (refer to Fig. 1). The LY556 epoxy resin, with a density of 0.98 g/cm³, and the HY951 hardener, with a density between 1.15 and 1.20 g/cm³, were utilized as shown in Fig. 2. Eggshells (ES), sourced locally from a tea vendor in Nsukka Town, Nigeria, were incorporated into the composite formulation. Additional chemicals used in the process included 30% hydrogen peroxide (H₂O₂) and 65% nitric acid (HNO₃).

Methods

ESp modification on CNTs

The ES were first washed with distilled water to remove the membranes and then dried for 24 h. After cleaning, the shells were calcined in a muffle furnace at 650°C for 5 h to produce calcium carbonate (CaCO₃) particles (ESp). These calcined eggshells were subsequently milled into fine ESp. The sol-gel method, as described by [35], was then applied to synthesize the ESp. Before decorating the ESp onto CNTs, the CNTs were modified by treating them with 100 ml of a mixture of 30% H₂O₂ and 65% HNO₃. The ESp was then attached to the CNTs using an ultrasonic sonicator (Sonics model VVC 505) at a power of 400 W for 60 min.

Composites manufacturing

To address the limitations of the traditional Taguchi method in handling multi-objective optimization, the Taguchi-Grey method was selected for this study’s design. The experimental trials were structured using Taguchi’s orthogonal array, while Grey relational analysis was employed to consolidate multiple responses into a single metric, optimizing overall performance. A Taguchi L9 design of experiments (DOE) was used to examine interactions among dielectric properties, electrical conductivity, and thermal conductivity. Key processing variables included eggshell particle weight percentage (wt% ESp) (Factor A), carbon nanotube weight percentage (wt% CNTs) (Factor B), curing temperature (Factor C), and curing time (Factor D). The response variables—dielectric properties, electrical conductivity, and thermal conductivity—were evaluated as critical outputs to enhance material performance for electrical applications, as indicated in [17, 36]. Table 1 outlines the specific factors and levels selected, informed by prior studies, while Table 2 shows the coded values for the L9 orthogonal array.

A refined solution stir-casting method was employed to produce the unique composites. A magnetic stirrer was used to combine the hardener and epoxy resin at a ratio of 1 : 10 and agitate the mixture for 15 min. During the composite fabrication process, air bubbles were eliminated using a roller press. The parameters provided in Table 2 served as the foundation for the composites’ formula. After that, the composite mixture was put into a heated wooden mold and left to cure for a full day at room temperature. The yield of CNTs (wt%), the ESp mixture (wt%), the curing temperature (°C), and the curing time (h) were all included in the Taguchi array responses. After the composites were cured, tests were conducted to evaluate dielectric properties, thermal conductivity, and electrical conductivity. The data obtained from these tests were then analyzed, and the signal-to-noise (S/N) ratio values were calculated from the results.

Composites characterization

The characterization of CNTs/ESp composites was performed using several analytical techniques to assess microstructure, electrical conductivity, dielectric properties, and thermal conductivity.

Transmission Electron Microscopy (TEM) was performed using the Joel JEM-2100F model to obtain high-resolution imaging of the composites’ internal microstructure, offering detailed insights into the nanoscale arrangement of the materials. The TEM imaging further revealed the surface morphology of the materials, shedding light on the textures and physical characteristics which play a significant role in enhancing electrical and thermal conductivity.

Electrical conductivity measurements were performed with the Kaise SK5010 model, adhering to ASTM standards to ensure accuracy and reliability in evaluating the composite’s electrical properties. Composite samples, prepared in a standardized sheet format measuring 5 cm × 5 cm with a thickness of 2 mm, were placed between two electrodes, allowing for precise measurement of current flow through the material under controlled conditions. Electrical conductivity values (⁠$σ$⁠) were calculated using Equation (1), which provided a quantitative measure of the material’s ability to conduct electrical current, a critical parameter for applications in electronic devices. Additionally, the dielectric constant (⁠$ε1$⁠), a property indicative of the material’s insulating capabilities and response to electric fields, was determined using Equation (2).

Thermal conductivity measurements were conducted using the transient planar heat source (TPS) technique, employing the TPS2500s model to ensure precise evaluation of heat transfer properties. Standardized samples, each with a diameter of 12.7 mm and a thickness of 2 mm, were used to maintain consistency across tests. The transient planar heat source method allowed for rapid assessment of thermal conductivity by analyzing the heat dissipation within the sample when subjected to a controlled thermal pulse. Each sample was measured three times to account for any variability, with the average value recorded to enhance accuracy. This approach provided critical data on the material’s thermal management potential, which is essential for applications in heat-sensitive electronic and electrical systems.

Universal testing equipment, the PC-2000 Testometric testing machine, was utilized to gauge the samples’ tensile strength. The test was carried out according to ASTM C-633.

Results and discussion

Composition of the ESp

The chemical analysis presented in Fig. 3 reveals that calcium oxide (CaO) is the predominant constituent in ESp, with a concentration of 85.68 wt%. This high CaO content indicates that ESp is largely composed of calcium carbonate (CaCO₃), making it suitable for the functionalization of carbon nanotubes (CNTs). The findings align closely with those reported by prior studies, such as [37], confirming the high CaO composition of ESp and its potential utility in composite applications.

TEM images of ESp onto CNTs

Figure 4 illustrates the TEM image of the ESp/CNTs composite. A slight alteration in the TEM appearance of the CNTs was observed after the incorporation of ESp, as depicted in Fig. 1. The ESp helped to change the multi-network layer of CNTs. The black color was visible to show the ESp and CNTs in the TEM image. It was noted that when CNTs were adorned with ESp, there was negligible segregation. This enhanced the electrical channels and ensured greater CNT distribution and dispersion throughout the polymer. This corresponds to the findings of [38, 39].

Multi-response optimization

The results of the multi-response L9 experimental design are presented in Table 3. Grey Relational Generation (GRG) values were calculated using Equation (3), providing a basis for multi-response optimization. The signal-to-noise (S/N) ratios were evaluated within a range of 0 ≤ S/N ≤ 1, with higher values indicating improved performance outcomes. From the results, the electrical conductivity values exhibited substantial variation depending on the combined influence of CNT loading and curing parameters. For instance, the highest conductivity (1.54E-08 S/m) was observed in sample 6, with moderate eggshell loading (2 wt%) and maximum CNT loading (3 wt%), under low curing temperature but intermediate curing time. Furthermore, the dielectric constant values ranged from 2.22 to 4.94, peaking in sample 8 (3 wt% ESp, 2 wt% CNTs) with a dielectric constant of 4.94. This suggests that a higher ESp loading in combination with moderate CNT levels enhances the dielectric properties. Additionally, thermal conductivity reached a maximum of 0.98 Wm⁻¹K⁻¹ in sample 3 (1 wt% ESp, 3 wt% CNTs), where the highest CNT loading is paired with an elevated curing temperature and time. This suggests that heat transfer within the composite benefits from both high CNT content and increased curing, likely due to enhanced CNT dispersion and network integrity.

Where i ranges from 1 to 9. The computed results of deviation and grey relational generation are presented in Table 4. From Table 4, it was observed that sample 4 showed the highest deviation for electrical conductivity (1), indicating it had the least favorable conditions for electrical performance. Conversely, sample 6, with a deviation of 0, represented the best scenario for electrical conductivity. The grey relational generation for electrical conductivity revealed sample 6 as the most desirable (generation = 1), whereas sample 4 exhibited the lowest desirability (generation = 0), confirming that Sample 6 best met the target conductivity goals.

Samples displayed a range of deviations in the dielectric constant, with sample 8 achieving the highest generation value (1) and, therefore, most closely matching the desired dielectric constant, while Sample 7 displayed the greatest deviation (0). Notably, dielectric constant optimization is crucial for applications involving charge storage or dielectric performance enhancement. The consistency in high generation values for samples such as 2 and 5 suggests they are closer to the desired dielectric performance, affirming the effectiveness of moderate to high eggshell particle loading for enhancing dielectric properties.

For thermal conductivity, sample 3 achieved the highest value of grey relational generation (1), closely meeting the optimal heat transfer characteristics desired, while sample 8 shows a complete deviation (0).

Where: $(Δ0i(k) i,)$ represents the normalized S/N value deviation; $(Δmax(1))$ represents the mormalized S/N ratios maximum; $(Δmin)$ represents the normalized S/N ratios minimum; and $(ξ)$ identification coefficient. $ξ=0.5$ was utilized in the work.

GRG evaluation

Figure 6 presents the GRG plot, with the midline representing the mean of the GRG values shown in Table 6. Notably, the four variables under consideration displayed anomalous patterns. The weight percentage of ESp was found to significantly affect thermal conductivity, electrical conductivity, and the dielectric constant. Grey relational grades increased as the weight percentage of ESp moved from level 1 (0) to level 2 (1), then declined when it exceeded level 2. Similarly, the weight percentage of CNTs rose from level 1 (0) to level 3 (2.5). This increase in grey relational grades beyond level 1 for both CNTs and ESp is likely due to weak interfacial interactions between the reinforcement and epoxy resin, aligning with findings from [40]. In the epoxy matrix, the combination of CNTs and ESp formed dense 3-D structures and conductive network topologies, enhancing the grey relational grade. This enhancement was particularly notable with a rise in the weight percentage of CNTs at level 2 and ESp at level 3. The conductivity pathways within the polymer promoted the rapid movement of charge carriers. The direct contact between ESp and CNTs within the epoxy matrix allowed for conduction through both non-ohmic and ohmic mechanisms, facilitating indirect contact between ESp and CNTs within the polymer matrix [41].

The curing temperature plays a pivotal role in determining the final properties of the composite materials. In this study, the curing temperature was increased from level 1 (60°C) to level 2 (90°C), while the curing time was reduced from level 1 (6 h) to level 3 (4 h). This increase in temperature is crucial for enhancing the cross-linking process between the epoxy resin and hardener, resulting in improved thermal stability and setting characteristics of the composite [41]. A proper curing temperature ensures that the polymer matrix is fully formed, promoting a robust structure that binds the reinforcement materials—such as CNTs and ESp—to the matrix. The optimal curing temperature facilitates better particle dispersion and interaction with the matrix, leading to improved dielectric, thermal, and electrical properties [42].

At the higher curing temperature of 90°C, the dielectric constant, thermal conductivity, and electrical conductivity all exhibited improvements due to the more efficient cross-linking and enhanced integration of the reinforcement within the epoxy matrix. The curing temperature’s ability to influence these properties underscores its importance in the performance of the composite. However, an excessively long curing time, as observed at level 3, weakened the grey relational grade (GRG), which reflects the strength of the interface between the reinforcement materials and the epoxy matrix. As the curing period extended, the interface relationship began to deteriorate, suggesting that while a higher curing temperature improves certain properties, the curing time must be carefully controlled to maintain optimal composite performance [42]. Thus, balancing both curing temperature and time is essential to avoid over-curing and degradation of the interface.

Response surfaces analysis

Surface response analysis was conducted to examine the impact of GRG on the four criteria factors. Using Minitab 16, surface response plots were generated, as shown in Fig. 7. Specifically, Fig. 6a and b illustrate a slight decrease in GRG as the weight percentage of ESp increases from level 1 to level 3, suggesting that higher wt% ESp beyond the optimal level may offer limited benefits, contrasting with findings by [39, 42]. Figure 6c and d indicate that a minimal weight percentage of CNT is necessary to reduce thermal and electrical conductivity as well as the dielectric constant. Figure 7 highlights the significant effects of temperature and curing time, with the highest levels of ESp and CNTs enhancing both thermal and electrical conductivity. Figure 7e’s interaction plot illustrates the combined effects of the four factors on dielectric constant, thermal conductivity, and electrical conductivity. Optimal values for thermal and electrical conductivity and dielectric constant were observed at level 2 for CNTs (2.5 wt%), levels 2 and 3 for ESp (1 wt%), level 1 for curing time (6 h), and level 2 for curing temperature (90°C).

GRG optimal setting

Table 7 presents the optimal levels and parameters, while Table 6 displays the mean values of the grey relational grades (GRG). The delta value for curing temperature, ranked first, is 0.0586, with a higher GRG value at level 2 (0.5232). The weight percentage of ESp follows closely, ranked second, with a peak score at level 2 (0.5013). The curing time ranks fourth, with a delta value of 0.0441 and an optimal setting at level 1, while the weight percentage of CNTs ranks third, achieving its highest score at level 3 (0.4993). According to the ranking, curing temperature, wt% ESp, wt% CNTs, and curing time significantly influence thermal conductivity, dielectric constant, and electrical conductivity, contributing to improved electrical and thermal conductivity. Table 7 identifies the optimal parameters as wt% ESp at level 2, curing temperature at level 2, wt% CNTs at level 3, and curing time at level 1. Thus, the ideal values for this study are 1 wt% ESp, a curing temperature of 90°C, 2.5 wt% CNTs, and a 6-h curing duration.

The results obtained for the four elements in the optimal experiment are as follows: a dielectric constant of 4.25, thermal conductivity of 0.75 W/m·K, and electrical conductivity of 1.25 × 10⁻⁹ S/m. These optimal values fall within the acceptable range for microelectronic materials [11, 15]. The observed increase in thermal conductivity is attributed to a reduction in the heat barrier between the ESp/CNTs and the epoxy matrix. Additionally, the strong synergistic interaction between CNTs-ESp and the epoxy at the ideal configuration enhanced both thermal and electrical conductivity. This improvement is confirmed by the experimental GRG value of 0.560, aligning with the predicted GRG of 0.5056. Notably, the 95% confidence interval of the experimental GRG (0.560) encompasses the predicted optimal GRG (0.5056), validating the effectiveness of the identified optimal processing settings.

XRD examination

Microstructure evaluation

Figure 9 presents SEM images that highlight the differences between the epoxy and its composites, with the reinforcement appearing as white phases. As the CNT content increases to 2 wt% and ESp to 1 wt%, the white phases representing CNTs in the composite become more pronounced. In the sample with 1 wt% ESp, the ESp distribution is uniform and well-organized, enhancing the interfacial bonding between the reinforcement and the polymer (Fig. 10b). This improvement in bonding is comparable to the effects seen with CaCO₃ from limestone, as reported in previous studies [3, 24]. However, when the ESp content is further increased to 1.5 wt%, minor segregations and agglomerations of CNTs are observed, as shown in Fig. 9c. Such CNT agglomerations, as indicated in Figs 5 and 6, may negatively impact electrical conductivity due to clustering.

Figure 10 presents the TEM images of the composites. In Fig. 10a, disordered and agglomerated CNTs are observed in the epoxy with 2.5 wt% CNTs. In contrast, Fig. 10b shows a uniform dispersion of CNTs within the epoxy matrix for the composite with 1 wt% ESp. The presence of ESp facilitates the incorporation and uniform dispersion of CNTs in the epoxy. This finding corroborates the earlier observations, indicating that 1 wt% ESp is crucial in preventing CNT segregation and promoting effective electron passage.

Figure 11 presents the stress-strain curves for the composite samples, revealing significant differences in strength and toughness based on the composition and functionalization of each sample. Notably, the sample functionalized with 1 wt% eggshell particles (ESp) exhibited the highest tensile strength and the largest area under the stress-strain curve, indicating both improved strength and toughness. This enhancement is largely attributed to the strong interfacial bonding between the epoxy matrix and the carbon nanotube (CNT) phases, alongside the uniform dispersion of ESp within the epoxy. The result is a well-integrated composite structure that effectively distributes applied stress, enhancing resistance to deformation and breakage under load. In contrast, the unfunctionalized sample containing 2.5% CNTs showed comparatively lower tensile strength, likely due to less effective bonding and dispersion within the epoxy matrix, which can create points of weakness under stress.

Quantitatively, the tensile strength of pure epoxy was measured at 81.5 MPa, while the epoxy sample containing 2.5% CNTs reached a strength of 110.5 MPa. However, when the sample was further functionalized with 1 wt% ESp, the tensile strength increased to 123.6 MPa, demonstrating the considerable reinforcing effect of ESp. This study thus confirms that functionalizing epoxy composites with both CNTs and a small addition of ESp can significantly enhance mechanical properties, producing a material that is not only stronger but also more resilient under stress. Such findings suggest promising applications for these functionalized epoxy composites in the development of conducting polymers with superior strength and durability, suitable for various advanced engineering applications where enhanced mechanical performance is critical.

Improvement mechanisms of electrical and thermal conductivity with ESp addition

As earlier noted, the incorporation of CNTs into an epoxy matrix introduces pathways for enhanced electrical and thermal conductivity, critical for applications in advanced electronics. However, due to the strong Van der Waals forces among CNTs, they tend to aggregate, creating non-uniform conductive pathways that limit their potential. The addition of ESp to the composite addressed this challenge by enhancing CNT dispersion, thus enabling the development of continuous, stable pathways for electron mobility and heat transfer.

First, ESp played a pivotal role in improving electrical conductivity by promoting better CNT dispersion within the epoxy matrix. The calcined ESp particles, composed largely of calcium carbonate, acted as a physical spacer that reduced CNT clustering and increases the percolation threshold. This reduction in clustering allowed more consistent and continuous pathways for electron flow, effectively lowering electrical resistivity. The composites containing ESp exhibited significantly higher electrical conductivity compared to those without ESp, indicating the successful facilitation of electron mobility through better CNT distribution.

Additionally, the thermal conductivity of the composite was enhanced by the improved filler-matrix interaction facilitated by ESp. ESp’s fine particulate structure aided CNT distribution and enhanced matrix thermal stability. Results showed that the composites with ESp achieved higher thermal conductivity values, likely due to the reduction in thermal interface resistance and more efficient heat dissipation. The calcium carbonate in ESp further stabilized the matrix under thermal load, allowing heat to transfer effectively across the composite.

ESp’s unique composition and structure enabled it to serve a dual purpose in enhancing both electrical and thermal properties. As a sustainably sourced calcium carbonate material, ESp provides a cost-effective, eco-friendly approach to enhancing composite performance. The porosity and particulate nature of ESp ensured that CNTs remained uniformly dispersed within the epoxy matrix, forming a consistent network that prevented agglomeration.

Conclusion

This study utilized the Taguchi-Grey method to uncover new insights into the synergistic interactions between dielectric materials and the high thermal conductivity of epoxy composites enhanced with carbon nanotubes (CNTs) and decorated with eggshell particles (ESp). Findings indicate that the optimal composite configuration includes 1 wt% ESp, a curing temperature of 90°C, a curing duration of 6 h, and 2.5 wt% CNTs. Under these optimal conditions, the ESp-CNT composite exhibits notable electrical, dielectric, and thermal properties due to enhanced electron mobility, significantly improving the polymer’s conductivity. The predicted ideal grey relational grade (GRG) of 0.5056 falls within the 95% confidence interval of the experimental GRG of 0.560, validating the effectiveness of the identified processing parameters. The values achieved for thermal conductivity (0.75 W/m·K), dielectric constant (4.25), and electrical conductivity (1.25 × 10⁻⁹ S/m) are within an appropriate range for microelectronic applications. This study demonstrates that repurposed eggshells can be effectively utilized to create uniformly dispersed, high-strength, conductive polymer composites with CNTs.

Acknowledgements

The authors hereby appreciate and acknowledge the Africa Centre of Excellence for Sustainable Power and Energy Development, ACE-SPED, University of Nigeria, Nsukka; Energy Materials Research Group, University of Nigeria, Nsukka, Nigeria; and Faculty of Engineering and Built Environment, University of Johannesburg, Auckland Park, South Africa for their support.

Author contributions

Ekele Dinneya-Onuoha (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Resources [equal], Validation [equal], Writing—original draft [equal], Writing—review & editing [equal]), Victor Sunday Aigbodion (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Resources [equal], Writing—original draft [equal], Writing—review & editing [equal]), and Alfred Ogbodo Agbo (Conceptualization [equal], Resources [equal], Writing—original draft [equal], Writing—review & editing [equal])

Conflict of interest: None declared.

Funding

The authors declare that this research was conducted without any external funding support.

Data availability

The authors confirm that the data supporting this study’s conclusions is included in the publication.

Consent to publish

The authors give the publisher consent to publish the work.

References