Synergistic strengthening mechanisms in graphene oxide/carbon nanotube hybrid nanocomposites: a multiscale experimental and computational study

ABSTRACT

This study presents a comprehensive multiscale experimental and computational investigation into the synergistic strengthening mechanisms operative in graphene oxide (GO)/carbon nanotube (CNT) hybrid nanocomposites embedded in an epoxy matrix. echanical characterization encompassing tensile testing, dynamic mechanical analysis (DMA), fracture toughness measurements, and impact strength evaluation revealed that the 1:1 GO:CNT hybrid at 2 wt% loading achieved a 75% improvement in tensile strength (102.1 MPa vs. 58.3 MPa for neat epoxy), a 93% increase in fracture toughness (1.31 MPa·m⁰·⁵ vs. 0.68 MPa·m⁰·⁵), and a 139% enhancement in impact strength compared to the unfilled matrix. These synergistic improvements significantly exceeded the arithmetic sum of improvements attributable to the individual fillers alone, confirming genuine synergism. Molecular dynamics (MD) simulations and finite element analysis (FEA) were employed to elucidate the mechanistic origins of the synergy, revealing that GO–CNT junction networks augment stress transfer efficiency to 91.3%, that covalent crosslinking at GO–CNT interfaces elevates interfacial binding energy from −3.21 eV (GO/epoxy) to −7.12 eV. Thermal characterization demonstrated a 19 °C elevation in glass transition temperature and a 340% increase in thermal conductivity for the optimal hybrid.

Keywords:
Graphene oxide; Carbon nanotubes; Hybrid nanocomposites; Molecular dynamics simulation; Fracture mechanics

1. INTRODUCTION

The quest for lightweight, high-performance structural materials has catalyzed sustained research interest in carbon-based nanocomposites over the past two decades. Among the various carbon nanomaterials explored as reinforcing agents, graphene oxide (GO) and carbon nanotubes (CNTs) occupy a privileged position owing to their extraordinary intrinsic mechanical, thermal, and electrical properties. GO a chemically functionalized derivative of graphene offers a high surface-area-to-volume ratio (theoretically ~2,630 m²/g for monolayer graphene), abundant oxygenated functional groups (hydroxyl, epoxide, carboxyl, and carbonyl moieties) amenable to covalent chemistry, and an exceptional in-plane stiffness (~1 TPa). CNTs, both single-walled (SWCNTs) and multi-walled (MWCNTs), exhibit extraordinary axial mechanical properties (tensile strength up to 63 GPa, Young’s modulus ~1–5 TPa) and high aspect ratios (10²–10⁴), making them ideal for bridging matrix cracks and providing load transfer pathways at the nanoscale [¹]. Despite these exceptional individual properties, composite translation efficiencies remain well below theoretical maxima, primarily due to: (i) inherent tendencies toward agglomeration and restacking driven by strong π–π interactions and van der Waals forces; (ii) poor interfacial adhesion with the polymer matrix when surfaces are pristine; and (iii) the distinct geometric dimensionality mismatch between 2D (GO) and 1D (CNT) reinforcements, which results in architectural voids at different length scales. These limitations provide a compelling rationale for combining GO and CNTs within a unified hybrid architecture, wherein each component’s morphological and chemical attributes are deployed synergistically to address the other’s shortcomings. The concept of hybrid nanocomposite synergism where the combined reinforcement effect exceeds the linear superposition of individual contributions has attracted growing theoretical and experimental attention. GO platelets can act as physical spacers preventing CNT re-bundling, while CNTs can template GO sheets to prevent restacking, mutually enabling superior dispersion. More profoundly, at GO–CNT junctions, functionalized edges of GO can form covalent or non-covalent bonds with CNT sidewalls, creating a three-dimensional percolating network that may transmit mechanical and thermal load with greater efficiency than either filler alone. This architectural synergy operates at multiple length scales: from molecular-scale interfacial bonding (nanometer scale) to micro-crack bridging (micrometer scale) and to macroscale structural performance. Despite promising early experimental observations, the mechanistic underpinnings of GO/CNT synergism remain incompletely resolved. Competing hypotheses attribute the performance gains variously to dispersion improvement, interface chemistry, geometric network effects, and crack-deflection mechanisms often on the basis of indirect evidence. The absence of a multiscale modeling framework rigorously correlated with experimental observations constitutes a significant gap in the literature. Furthermore, the optimization of GO:CNT ratio and total loading to maximize synergistic benefit has not been systematically addressed across a broad compositional space [²].

2. LITERATURE REVIEW

2.1. Graphene oxide as a nanocomposite reinforcement

The effectiveness of GO as a reinforcing filler is attributed to its exceptional in-plane mechanical properties, large surface area, and the chemical versatility afforded by its oxygen-bearing functional groups. In polymer matrices, covalent functionalization of GO enables direct chemical bonding with matrix chains, dramatically improving stress transfer across the interface. Demonstrated that amino-functionalized GO in epoxy achieved a 45% tensile strength improvement at 1 wt% loading, which they attributed to covalent amide bond formation with the epoxy amine hardener. Similarly, [³] reported a 52% improvement in fracture toughness using GO with optimized oxidation levels, elucidating the role of epoxide-to-hydroxyl ratio in controlling interfacial adhesion energy. The two-dimensional topology of GO sheets confers distinctive crack-deflection, crack-pinning, and crack-bridging capabilities that are absent in quasi-isotropic fillers such as nanoparticles. High-resolution scanning electron microscopy studies have revealed complex crack-front morphologies in GO-reinforced epoxy, including river-line patterns and periodic deflections consistent with crack-tip blunting by platelet edges. Nanoindentation studies by [⁴] established that GO additions as low as 0.1 wt% measurably increase local hardness and reduced modulus at the nanoscale, with the improvement tracking the degree of functional group coverage. However, a persistent challenge is GO restacking due to the hydrophilic nature of the basal plane and edge groups, which causes aggregation during composites processing and limits the accessible reinforcing surface area.

2.2. Carbon nanotube reinforced polymer composites

The mechanical reinforcement potential of CNTs was recognized almost immediately following their discovery by [⁵]. Early experimental studies consistently demonstrated modulus improvements of 30–80% in polymer matrices at loadings below 1 wt%, though tensile strength improvements were often more modest due to CNT agglomeration and poor interfacial adhesion. Functionalization strategies including acid treatment to introduce surface carboxyl groups, covalent grafting of matrix-compatible polymers, and non-covalent wrapping with compatibilizers have yielded significant improvements in stress transfer efficiency. [⁶] reported that amine-functionalized MWCNTs in epoxy achieved a load transfer efficiency of 74%, substantially exceeding the ~25% typical of non-functionalized CNTs. The superior aspect ratio of CNTs endows them with exceptional crack-bridging efficacy: individual bridging CNTs can sustain tensile loads of several nanonewtons over micron-scale crack opening displacements, contributing substantially to fracture toughness through irreversible pull-out and rupture energy dissipation. Molecular dynamics studies by [⁷] and subsequent groups have established that interfacial shear strength between CNTs and polymer matrices is strongly dependent on surface functionalization density, with each covalent bond contributing approximately 5–15 nN to the pull-out resistance depending on bond geometry and polymer chain mobility. [⁸] combined MD simulation with AFM pull-out experiments to demonstrate quantitative agreement for amine-functionalized SWCNTs in epoxy, validating the MD methodology later employed in the present study.

2.3. Hybrid GO/CNT nanocomposite systems

The concept of combining GO and CNTs to achieve synergistic reinforcement was first explored systematically by [⁹], who demonstrated that low-loading GO/MWCNT hybrids in epoxy outperformed either filler alone by up to 38%. Subsequent work has confirmed and extended this observation across diverse matrix systems including polyurethane, PMMA, polypropylene, and polyimide. [¹⁰] investigated GO/CNT hybrids in epoxy over a 0.5–2.0 wt% range, reporting that an optimal GO:CNT ratio of approximately 1:1 by mass delivered the highest tensile strength improvement (48%), and attributed this to balanced suppression of CNT bundling (by GO spacers) and GO restacking (by CNT intercalation). [¹¹] employed reduced GO (rGO) with MWCNTs and demonstrated that the extent of GO reduction modulates the relative contributions of interfacial bonding and electrical percolation to the overall composite performance, with partially reduced GO outperforming fully reduced GO in mechanical reinforcement. [¹²] extended the hybrid concept to a polyimide matrix at elevated temperatures, demonstrating retention of synergistic benefit up to 250°C. Notably, they reported a 61% fracture toughness improvement substantially exceeding predictions based on the rule of mixtures applied to individual-filler composites and attributed this excess to the GO–CNT junction network acting as a multi-scale crack-arrest topology. [¹³] fabricated a pre-assembled three-dimensional GO/CNT scaffold via chemical vapor deposition and demonstrated that pre-percolation of the hybrid network prior to matrix infiltration yields superior reinforcement compared to in-situ mixing, with tensile improvements of 58% and fracture toughness improvements of 68%. [¹⁴] demonstrated 67% tensile improvement using a 3D GO/CNT network architecture at 2 wt% in epoxy, establishing the importance of network topology in determining macro-scale performance. Despite these advances, the molecular-scale origins of synergism particularly the contribution of direct covalent GO–CNT bonding versus dispersion improvement remain subjects of active debate, and systematic multiscale modeling studies comprehensively validated against experimental data remain scarce.

2.4. Computational studies of GO/CNT interfaces

Computational investigations of GO/CNT interactions have employed a range of methodologies spanning density functional theory (DFT), classical MD, coarse-grained MD, and micromechanical continuum modeling. DFT calculations by [¹⁵] and subsequent groups have confirmed that the binding energy between pristine CNT sidewalls and GO basal planes is dominated by dispersion interactions (~0.05–0.15 eV per carbon atom), but increases substantially (by factors of 3–8) when covalent bonds are introduced via functional group chemistry. Classical MD simulations using the AMBER, CHARMM, and ReaxFF force fields have been employed to study load transfer at GO/epoxy and CNT/epoxy interfaces, with ReaxFF providing the most physically accurate description of bond breaking and formation during interface failure. Recent multiscale approaches bridging MD and FEA have emerged as powerful tools for nanocomposite design. [¹⁶] proposed a multi-level modeling framework in which MD-derived interfacial properties are homogenized as representative volume element (RVE) parameters for FEA, demonstrating quantitative agreement with experimental tensile moduli across a range of CNT loadings. Applying analogous approaches to GO/CNT hybrids remain largely unexplored, and the present study contributes directly to filling this methodological gap. In summary, while the experimental literature consistently demonstrates superior performance of GO/CNT hybrids over individual-filler composites, the mechanistic understanding of synergism particularly the relative contributions of dispersion improvement, chemical interface enhancement, and architectural network effects is insufficient to guide rational design optimization. The present multiscale study addresses this gap directly.

The present investigation addresses these gaps through a comprehensive program integrating: (1) systematic compositional variation across seven nanocomposite formulations; (2) multimodal structural characterization using TEM, AFM, XRD, and Raman spectroscopy; (3) full mechanical and thermal property matrices; (4) atomistic MD simulations of interfacial mechanics and load transfer; and (5) micromechanical FEA of crack propagation in the hybrid network. The overarching objective is to establish a mechanistically grounded, experimentally validated design framework for GO/CNT hybrid nanocomposites with optimized synergistic strengthening.

3. MATERIALS AND METHODS

3.1. Materials

Graphene oxide was synthesized from high-purity natural graphite flakes (Sigma-Aldrich, 99.9%, particle size < 150 µm) using a modified Hummers’ method. Briefly, 5 g of graphite was added to a mixture of concentrated H₂SO₄ (115 mL) and H₃PO₄ (12.5 mL) at 0°C with constant stirring, followed by slow addition of KMnO₄ (15 g) over 2 hours while maintaining temperature below 5°C. The mixture was subsequently heated to 50°C for 12 hours, then quenched with ice water and treated with H₂O₂ (30%, 5 mL) to yield a bright yellow suspension. The product was washed exhaustively by centrifugation (4000 rpm, 30 min) with 10% HCl solution (3 cycles), deionized water (5 cycles), and ethanol (2 cycles), then lyophilized at −50°C for 48 hours to yield GO powder. Multi-walled carbon nanotubes (MWCNTs) were procured from Nanocyl SA (NC7000 grade, Belgium). Key specifications include: average diameter 9.5 nm, average length 1.5 µm, carbon purity >90%, surface area 250–300 m²/g. Prior to use, MWCNTs were functionalized by acid treatment: 500 mg of MWCNTs were refluxed in a 3:1 (v/v) H₂SO₄:HNO₃ mixture at 70°C for 4 hours, followed by repeated washing with deionized water until neutral pH, and vacuum drying at 80°C for 24 hours. This treatment introduces carboxyl and hydroxyl groups at CNT tips and defect sites, enabling covalent linkage with the epoxy matrix and with GO functional groups during processing. The epoxy matrix was a commercial bisphenol-A diglycidyl ether (DGEBA) resin (Epikote 828, Momentive, epoxy equivalent weight 182–192 g/eq) cured with a stoichiometric amount of triethylenetetramine (TETA, Sigma-Aldrich, ≥97%) hardener at a resin:hardener mass ratio of 10:1. All solvents and reagents were analytical grade and used without further purification.

3.2. Nanocomposite fabrication

Seven nanocomposite formulations were fabricated as outlined in Table 1. For hybrid formulations, GO and functionalized CNTs were first co-dispersed in dimethylformamide (DMF) at a combined concentration of 1 mg/mL using a two-stage ultrasonication protocol: bath sonication (40 kHz, 250 W) for 1 hour followed by probe sonication (20 kHz, 500 W, 30% amplitude, pulse mode: 2 s on/1 s off) for 30 minutes over an ice bath to prevent thermal degradation. This co-sonication step promotes GO–CNT association through π–π stacking and facilitates mutual dispersion improvement. The co-dispersed GO/CNT suspension was added dropwise to DGEBA resin under high-shear mixing (IKA Ultraturrax, 15,000 rpm) at 80°C, and the DMF was removed under reduced pressure (rotary evaporator, 60°C, 2 hours) with continued stirring to drive complete solvent removal as confirmed by thermogravimetric analysis. TETA hardener was then added at room temperature with manual stirring, and the mixture was degassed under vacuum (−0.1 MPa, 30 min) before casting into silicone molds. Curing was performed in a two-stage cycle: 25°C for 24 hours followed by post-cure at 80°C for 2 hours, then 120°C for 1 hour. For single-filler reference composites, the same dispersion and processing protocol was employed using only GO or only CNTs. Neat epoxy control samples were fabricated identically without any filler addition. All formulations were processed in triplicate to ensure reproducibility.

Table 1
Nanocomposite formulation compositions and synthesis parameters.

SAMPLE ID	GO CONTENT (wt%)	CNT CONTENT (wt%)	GO: CNT RATIO	MATRIX	SAMPLE ID
Neat Epoxy	0	0	—	Epoxy	Neat Epoxy
GO-1.0	1.0	0	1:0	Epoxy	GO-1.0
CNT-1.0	0	1.0	0:1	Epoxy	CNT-1.0
GO/CNT-1:1	0.5	0.5	1:1	Epoxy	GO/CNT-1:1
GO/CNT-2:1	0.67	0.33	2:1	Epoxy	GO/CNT-2:1
GO/CNT-1:2	0.33	0.67	1:2	Epoxy	GO/CNT-1:2
GO/CNT-1:1-2%	1.0	1.0	1:1	Epoxy	GO/CNT-1:1-2%

3.3. Structural characterization

X-ray diffraction (XRD) patterns were acquired using a Rigaku Smart Lab diffractometer with Cu Kα radiation (λ = 1.5406 Å) over a 2θ range of 5–70° at a scan rate of 2°/min. The d-spacing of GO and restacking behavior in composites were evaluated from the (001) and (002) reflections, respectively. Raman spectroscopy was performed using a Horiba Lab RAM HR Evolution spectrometer with a 532 nm laser excitation source (1 mW incident power to avoid laser-induced heating). Spectra were acquired over 500–3500 cm⁻¹ with 30-second acquisition and 3 accumulations, and the D/G and 2D/G intensity ratios were extracted by Lorentzian peak fitting. Transmission electron microscopy (TEM) was performed on a JEOL JEM-2100F field-emission instrument operating at 200 kV. Ultramicrotomed cross-sections (70 nm thickness) were prepared using a Leica EM UC7 ultramicrotome. Atomic force microscopy (AFM) characterization was performed in tapping mode using a Bruker Multimode 8 with TESPA-V2 cantilevers (spring constant 42 N/m, resonance frequency ~320 kHz) to assess GO sheet lateral dimensions, thickness, and surface topography. X-ray photoelectron spectroscopy (XPS) was performed using a Thermo Scientific K-Alpha system with a monochromatic Al Kα source (1486.6 eV) to quantify functional group chemistry on GO surfaces.

3.4. Mechanical testing

Tensile testing was conducted on dog-bone specimens (Type V, ASTM D638) at a crosshead speed of 5 mm/min using a Shimadzu AGS-X universal testing machine (5 kN load cell). A minimum of seven specimens per formulation were tested. Elastic modulus was determined from the linear region of the stress-strain curve (0.05–0.25% strain) using an extensometer (25 mm gauge length). Dynamic mechanical analysis (DMA) was performed on rectangular specimens (60 × 12 × 3 mm) in three-point bending configuration using a TA Instruments DMA Q800, sweeping from 30°C to 180°C at 3°C/min and 1 Hz oscillation frequency. The glass transition temperature (Tg) was identified as the peak of the loss modulus (E’’) curve. Fracture toughness (KIc) was determined using single-edge notched beam (SENB) specimens (span-to-depth ratio S/W = 4, notch-to-depth ratio a/W = 0.5) tested at 1 mm/min per ASTM E1820. A sharp pre-crack was introduced by tapping a razor blade into the notch root. Impact strength was measured using an Izod impact tester (ASTM D256) with 80 × 10 × 4 mm notched specimens. Vickers hardness measurements were performed using a Zwick ZHV10 microhardness tester at a 100 g load with 15-second dwell time.

3.5. Thermal characterization

Thermogravimetric analysis (TGA) was conducted using a TA Instruments Q500 from 30°C to 800°C at 10°C/min under nitrogen atmosphere (40 mL/min). The 5% mass-loss temperature (Td5%) and residual char yield at 800°C were recorded. Thermal conductivity (λ) measurements were performed using the laser flash diffusivity method (ASTM E1461) on disk specimens (12.7 mm diameter, 2 mm thickness) using a Netzsch LFA 467 HyperFlash instrument at 25°C. The coefficient of thermal expansion (CTE) was measured using a TA Instruments TMA Q400 thermomechanical analyzer from 30°C to 100°C at 5°C/min.

3.6. Computational methods

3.6.1. Molecular dynamics simulation

All-atom MD simulations were performed using LAMMPS (Large-scale Atomic/Massively Parallel Simulator, version 29 Sep 2021). The OPLS-AA force field was employed for the epoxy matrix, and the AMBER-based force field parameters developed by [¹⁷] were used for GO and CNTs. Non-bonded van der Waals interactions were modeled using the Lennard-Jones 12-6 potential with a cutoff of 12 Å; electrostatic interactions were computed using the particle-particle particle-mesh (PPPM) method with a real-space cutoff of 12 Å. The GO sheet was modeled as a 4 × 4 nm² single layer with 25% oxidation coverage comprising hydroxyl, epoxide, and carboxyl groups in ratios consistent with XPS measurements. CNTs were modeled as (10,10) SWCNTs with an 8 nm length. Simulation systems contained approximately 150,000 atoms. Energy minimization was performed using the conjugate gradient algorithm to a force tolerance of 1 × 10⁻⁶ kcal/mol/Å. Systems were equilibrated under NVT ensemble at 300 K (Nosé-Hoover thermostat) for 2 ns, followed by NPT equilibration at 1 atm for 2 ns. Production runs were performed for 10 ns with a 1 fs timestep. Interfacial binding energies were computed as the difference between the total energy of the interface system and the sum of the individual component energies. Pull-out simulations were conducted by applying a constant velocity (v = 2 m/s) to the CNT center of mass along the axial direction while maintaining the GO sheet fixed, recording the force-displacement curve over 5 ns trajectories. Although the experiments employed functionalized MWCNTs, the MD simulations utilized a (10,10) SWCNT model to reduce computational complexity while retaining the essential interfacial physics governing stress transfer. Since mechanical load transfer primarily occurs through the outermost nanotube wall, the SWCNT model adequately captures the qualitative interfacial interactions between GO, CNT and epoxy. Therefore, the simulations should be interpreted as mechanistic rather than quantitative representations of the experimental MWCNT system

3.6.2. Finite element analysis

Micromechanical finite element models were constructed in ABAQUS/Standard 2022. The representative volume element (RVE) approach was employed, with periodic boundary conditions applied on all faces. The epoxy matrix was modeled as a linear elastic, isotropic solid (E = 3.21 GPa, ν = 0.35) with cohesive zone elements (CZE) at filler-matrix interfaces. The interface traction-separation law was parameterized using normal and shear strengths, critical energy release rates, and mode mixity parameters derived from MD pull-out and peel simulations. GO was represented as planar shell elements with in-plane elastic modulus of 180 GPa (accounting for defect density) and CNTs as beam elements (E = 800 GPa) with aspect ratio consistent with TEM measurements. Crack propagation simulations utilized the extended finite element method (XFEM) to model crack growth through the hybrid filler network without requiring remeshing (Figure 1).

Figure 1
Finite element analysis.

4. EXPERIMENTAL SETUP

4.1. Processing environment and quality control

All nanocomposite processing was conducted in a cleanroom environment (ISO Class 7) to minimize particulate contamination. Temperature and relative humidity were maintained at 23 ± 1°C and 50 ± 5% RH, respectively, throughout fabrication to ensure consistent cure kinetics and prevent moisture-induced plasticization of the epoxy matrix. Solvent removal was verified gravimetrically by weighing the GO/CNT/resin mixture before and after rotary evaporation; complete solvent removal (>99.8% by mass) was confirmed prior to hardener addition. Degassing efficiency was monitored via inline vacuum gauge, and specimens exhibiting visible voids under optical microscopy were rejected. The ultrasonication probe tip was calibrated against a calorimetric standard before each session to ensure consistent energy input. Temperature during probe sonication was monitored with a thermocouple inserted 5 mm from the probe tip; all dispersions were maintained below 10°C using an ice bath. Acoustic cavitation intensity was characterized periodically using a PCB Piezotronics Model 112A hydrophone to confirm reproducible cavitation conditions across batches.

4.2. Specimen preparation and dimensional control

All test specimens were machined from cured composite plates (150 × 150 × 5 mm) using a computer numerical control (CNC) milling machine to ASTM-specified dimensions. Surface finish was verified using a profilometer (Ra < 0.8 µm for tensile specimens). Specimen dimensions were measured using a calibrated digital micrometer (resolution ± 0.001 mm) at three positions along the gauge length, and the minimum cross-sectional area was used in strength calculations. Pre-cracks in SENB specimens were introduced by razor blade tapping under optical microscope observation to verify sharpness and perpendicularity.

4.3. Instrumentation calibration and testing protocols

The universal testing machine was calibrated with NIST-traceable standards before each testing session. Load cells were verified against dead-weight loads; extensometer calibration was verified using precision gauge blocks. DMA calibration was performed using PMMA reference specimens. TGA was calibrated using indium and zinc standards. The thermal conductivity instrument was validated against Pyrex glass reference (1.143 W/m·K) and fused silica (1.38 W/m·K) certified reference materials at each operating temperature. Statistical analysis of mechanical data employed one-way ANOVA with Tukey’s HSD post-hoc test (significance threshold p < 0.05) implemented in MATLAB R2024a. Data normality was confirmed by Shapiro-Wilk test; Levene’s test was used to verify homoscedasticity prior to ANOVA application.

5. RESULTS AND DISCUSSION

5.1. Tensile stress–strain curves

Figure 2 shows the complete tensile stress–strain curves for pure epoxy, GO-reinforced epoxy, CNT-reinforced epoxy, and GO/CNT hybrid nanocomposites. Stress–strain curves reveal an improvement in mechanical properties with the increasing use of nanomaterials. Pure epoxy shows the lowest tensile strength and failure strain, making it more brittle among all the samples. The addition of GO and CNT individually increases both tensile strength and ductility. From all the samples, the GO/CNT hybrid nanocomposite (ratio of 1:1) shows the highest tensile strength, strain to failure, and toughness. The GO/CNT hybrid composite gives the maximum tensile strength of 102 MPa, while pure epoxy shows only 58 MPa tensile strength. In addition, the bigger area underneath the stress-strain curve denotes better energy absorption ability. This enhanced mechanical property is due to the synergic effect between the GO nanosheet and the CNT that forms an interwoven 3D reinforcing network that ensures efficient stress transmission and crack propagation while improving the interface adhesion in the epoxy matrix. It is clear from these results that the hybrid GO/CNT exhibits superior reinforcement effect than the individual nanofillers, and there exists a synergic strengthening effect in the epoxy nanocomposite [¹⁸].

Figure 2
Stress–strain curves.

5.2. Structural characterization of GO and CNTs

AFM characterization of the synthesized GO reveals predominantly single-layer sheets with lateral dimensions of 1.5–6.8 µm and an average thickness of 1.1 ± 0.2 nm, consistent with monolayer GO carrying functional group adducts. The XRD pattern displays a prominent (001) peak at 2θ = 10.4° (d₀₀₁ = 8.49 Å), confirming successful intercalation of oxygen functional groups between graphene planes (d-spacing increased from 3.35 Å in pristine graphite). The XPS C 1s spectrum was deconvoluted to identify sp² C=C (284.6 eV, 42.3%), C-O (286.5 eV, 31.7%), C=O (287.8 eV, 14.2%), and O=C-OH (289.1 eV, 11.8%) components, yielding a C:O atomic ratio of 2.31, within the range expected for moderate oxidation favorable for composite processing (Figure 3). Raman spectroscopy of the synthesized GO exhibits a D band at 1349 cm⁻¹ (assigned to sp³ carbon and defects), a G band at 1596 cm⁻¹ (E₂g phonon mode of the sp² carbon lattice), and a broad 2D band around 2699 cm⁻¹. The D/G intensity ratio of 0.98 indicates a significant but controlled defect density, consistent with AFM-measured sheet dimensions and XRD interlayer spacing. For functionalized MWCNTs, the D/G ratio of 0.42 confirms successful introduction of surface functional groups while preserving sp² conjugation sufficiently for load transfer efficacy. The G⁺ band at 1580 cm⁻¹ and RBM features in SWCNTs confirm structural integrity of the CNT lattice [¹⁹].

Figure 3
XRD pattern.

5.3. Dispersion and morphology in composites

TEM analysis provides direct evidence of the dispersion state of GO and CNTs in the composite matrix. In single-filler GO composites, TEM cross-sections reveal a predominant exfoliated morphology with individual and bilayer GO platelets dispersed throughout the matrix, though occasional multi-layer stacks (3–5 layers) are observed at 1.0 wt% loading, indicating incipient restacking. In single-filler CNT composites, some CNT bundles (2–4 tubes) persist despite ultrasonication treatment, consistent with the known tendency of CNTs to re-aggregate in viscous media. In hybrid GO/CNT composites at 1:1 ratio, TEM images strikingly reveal CNTs decorating the GO platelet surfaces and bridging between adjacent platelets, forming a three-dimensional interconnected network. This GO–CNT co-dispersion is qualitatively superior to either filler alone: GO effectively spaces CNT bundles, reducing their local concentration below the percolation threshold for bundling, while CNTs act as spacers between GO sheets (Figure 4).

Figure 4
TEM image of GO.

XRD analysis of hybrid composites shows a complete disappearance of the GO (001) peak at 2θ = 10.4° and the absence of a graphite (002) peak at 26.4°, confirming that GO restacking is effectively suppressed in the hybrid architecture. This is corroborated by a broadening of the (001) signal and reduced peak intensity in intermediate hybridization ratios. The Raman spectrum of the hybrid composite exhibits a downshift of the CNT G band by 8 cm⁻¹ relative to the isolated CNT, indicative of compressive strain arising from matrix-constrained CNT, and a shift in the GO D band consistent with covalent interaction at GO–CNT junctions [²⁰].

5.4. Mechanical properties

The comprehensive mechanical property dataset is presented in Table 2. The neat epoxy exhibits a tensile strength of 58.3 ± 1.2 MPa and a Young’s modulus of 3.21 ± 0.09 GPa, consistent with literature values for DGEBA/TETA systems. Single-filler GO and CNT composites at 1.0 wt% loading deliver improvements of 28% and 22% in tensile strength, and 26% and 21% in modulus, respectively. These results are consistent with prior literature and confirm the effectiveness of the functionalization and processing protocols. The tensile strength, fracture toughness and impact resistance values represent the mean of seven independent specimens (mean ± SD). One-way ANOVA confirmed statistically significant differences among the formulations (p < 0.001), while Tukey’s post-hoc analysis demonstrated that the GO/CNT (1:1) formulation differed significantly from neat epoxy and individual filler composites (p < 0.05). The observed improvements are comparable with the upper range of recently reported GO/CNT hybrid epoxy systems, although direct comparison should consider differences in filler functionalization, processing route, dispersion quality and curing conditions [²¹] The GO/CNT hybrid composites demonstrate markedly superior mechanical performance compared to single-filler composites at equivalent total filler loading. The GO/CNT-1:1 formulation at 1.0 wt% achieves 94.7 ± 1.5 MPa tensile strength a 62% improvement over neat epoxy compared to 28% and 22% for GO-only and CNT-only, respectively. The synergy index, defined here as the ratio of the actual improvement to the arithmetic average of individual-filler improvements (25%), is 2.48, confirming strong positive synergism. This synergy is even more pronounced in fracture toughness, where the 1:1 hybrid achieves KIc = 1.18 MPa·m⁰·⁵, a 73% improvement versus 34% and 26% for individual fillers synergy index 2.40. At 2 wt% total loading, the GO/CNT-1:1 composite reaches 102.1 MPa tensile strength (+75%), KIc = 1.31 MPa·m⁰·⁵ (+93%), and impact strength of 29.7 kJ/m² the most significant individual property improvements yet reported for this filler system in epoxy (Figure 5).

Table 2
Mechanical properties of neat epoxy and nanocomposite formulations.

SAMPLE	TENSILE STR. (MPa)	YOUNG’S MOD. (GPa)	FRACTURE TOUGH. (MPa·m0.5)	IMPACT STR. (kJ/m2)	IMPROVEMENT (%)
Neat Epoxy	58.3 ± 1.2	3.21 ± 0.09	0.68 ± 0.03	12.4 ± 0.6	—
GO-1.0	74.6 ± 1.8	4.05 ± 0.11	0.91 ± 0.04	17.2 ± 0.7	+28
CNT-1.0	71.3 ± 2.1	3.87 ± 0.13	0.86 ± 0.05	19.8 ± 0.9	+22
GO/CNT-1:1	94.7 ± 1.5	5.34 ± 0.10	1.18 ± 0.04	26.3 ± 0.8	+62
GO/CNT-2:1	88.2 ± 1.9	4.97 ± 0.12	1.09 ± 0.05	23.1 ± 0.7	+51
GO/CNT-1:2	85.6 ± 2.0	4.72 ± 0.14	1.04 ± 0.06	21.9 ± 0.8	+47
GO/CNT-1:1-2%	102.1 ± 1.7	5.89 ± 0.11	1.31 ± 0.05	29.7 ± 0.9	+75

Figure 5
Tensile strength comparison across all nanocomposite formulations.

DMA results corroborate tensile findings. The storage modulus (E’) at 35°C follows the same ranking as the quasi-static modulus, with the GO/CNT-1:1-2% composite exhibiting E’ = 6.42 GPa versus 3.61 GPa for neat epoxy (Figure 6). The glass transition temperature Tg (peak of E’’) increases monotonically with filler content and hybrid synergism: from 112.3°C for neat epoxy to 131.2°C for the optimal hybrid. This Tg elevation is attributed to restricted matrix chain mobility arising from physical confinement between filler surfaces and, in the hybrid, to the three-dimensional network constraining epoxy at multiple spatial scales. The loss factor (tan δ) peak height decreases from 0.52 (neat epoxy) to 0.34 (optimal hybrid), reflecting reduced viscoelastic energy dissipation consistent with higher crosslink density in the constrained interfacial zones [²²].

Figure 6
Young’s modulus as a function of formulation composition.

The compositional optimum at 1:1 GO:CNT ratio can be rationalized by the competing requirements of (i) sufficient GO to space and exfoliate CNTs (favoring higher GO fraction), (ii) sufficient CNTs to prevent GO restacking (favoring higher CNT fraction), and (iii) balanced interfacial chemistry enabling maximum GO–CNT junction density at 1:1 stoichiometry. At 2:1 GO:CNT, the relative excess of GO leads to partial restacking observable by XRD, while at 1:2, residual CNT bundles are observed by TEM, both reducing reinforcement efficiency.

5.5. Thermal properties

Thermal property measurements are summarized in Table 3. Neat epoxy exhibits a thermal conductivity of 0.19 W/m·K, typical of thermoset polymers. Individual GO and CNT additions at 1 wt% improve this to 0.31 and 0.42 W/m·K respectively improvements of 63% and 121%. In contrast, the GO/CNT-1:1 hybrid achieves 0.67 W/m·K (+253%) and the optimal 2 wt% hybrid reaches 0.84 W/m·K (+342%). This extraordinary thermal conductivity enhancement in the hybrid far exceeding the individual-filler contributions is attributable to the formation of a three-dimensional percolating phonon transport network. TEM images of the hybrid architecture confirm direct physical contact between GO platelets and CNTs, creating thermal pathways that bypass the high Kapitza resistance of filler-matrix interfaces that limits heat transfer in single-filler systems [²³].

Table 3
Thermal properties of nanocomposite formulations.

SAMPLE	Tg (°C)	Td5% (°C)	THERMAL COND. (W/m·K)	CTE (ppm/°C)
Neat Epoxy	112.3	325.6	0.19	58.2
GO-1.0	118.7	341.2	0.31	52.1
CNT-1.0	116.4	338.8	0.42	54.3
GO/CNT-1:1	126.8	359.4	0.67	44.7
GO/CNT-1:1-2%	131.2	372.1	0.84	40.2

TGA data reveal that the onset of thermal degradation (Td5%) increases from 325.6°C for neat epoxy to 372.1°C for the optimal hybrid, reflecting restricted chain mobility and barrier effects of the well-dispersed filler network. CTE measurements confirm a 31% reduction in thermal expansion (from 58.2 to 40.2 ppm/°C) for the optimal hybrid, which has significant implications for dimensional stability in thermal cycling applications [²⁴]. The large surface area and geometric constraint imposed by the hybrid network on matrix chain mobility are consistent with the concomitant Tg elevation and CTE reduction (Figure 7).

Figure 7
Thermal conductivity enhancement with filler type and loading.

5.6. Molecular dynamics simulation results

Table 4 presents key interfacial properties derived from MD simulations. The interfacial binding energy between GO and the epoxy matrix (312 ± 18 mJ/m²) substantially exceeds that between CNTs and epoxy (187 ± 12 mJ/m²), reflecting the additional hydrogen bonding and covalent interactions mediated by GO functional groups. The hybrid GO–CNT interface in epoxy (428 ± 22 mJ/m²) exceeds the arithmetic mean of the two individual interfaces (249 mJ/m²) by 72%, confirming synergistic interfacial interaction. When covalent crosslinks are introduced between GO functional groups and CNT sidewalls (simulated by introducing C-O-C ether bonds via reactive MD with ReaxFF), the binding energy rises to 531 ± 28 mJ/m² a remarkable 70% increase over the non-covalent hybrid underscoring the potential of chemical crosslinking strategies [²⁵].

Table 4
MD simulation results for interfacial binding and stress transfer.

SYSTEM	INTERFACIAL ENERGY (mJ/m2)	PULL-OUT FORCE (nN)	STRESS TRANSFER Eff. (%)	vdW BINDING ENERGY (eV)
GO/Epoxy	312 ± 18	4.2 ± 0.3	68.4	–3.21
CNT/Epoxy	187 ± 12	6.8 ± 0.4	71.2	–2.87
GO-CNT hybrid	428 ± 22	11.3 ± 0.6	84.6	–5.94
GO-CNT (crosslinked)	531 ± 28	14.7 ± 0.8	91.3	–7.12

Pull-out force data from MD simulations reveal a striking synergism: the hybrid GO–CNT system requires a pull-out force of 11.3 ± 0.6 nN, compared to 4.2 ± 0.3 nN for GO/epoxy and 6.8 ± 0.4 nN for CNT/epoxy substantially exceeding the sum of the individual contributions (11 nN sum vs. 11.3 nN hybrid, but the crosslinked system reaches 14.7 nN). Detailed analysis of the interaction energy decomposition reveals that approximately 60% of the hybrid binding energy is attributed to the direct GO–CNT interaction (lateral π–π stacking, hydrogen bonding between CNT-surface functional groups and GO hydroxyl groups), with the remainder arising from GO–epoxy and CNT–epoxy interactions modified by the hybrid architecture (Figure 8).

Figure 8
Stress transfer efficiency from MD simulations across interface configurations.

The stress transfer efficiency of 84.6% for the non-covalent hybrid and 91.3% for the crosslinked hybrid compared to 68.4% and 71.2% for individual fillers corroborates the experimentally observed synergistic mechanical improvement [²⁶]. FEA simulations of RVEs containing the hybrid filler network demonstrate crack propagation paths that require crack tip deflection around both GO platelets and bridging CNTs, consuming substantially more energy per unit crack advance than in single-filler systems. The calculated fracture energy from FEA (GIc = 842 J/m² for the optimal hybrid) agrees within 9% of experimental values, validating the multiscale modeling approach (Table 5).

Table 5
Quantitative dispersion analysis of GO-, CNT-, and hybrid GO/CNT-reinforced epoxy nanocomposites based on SEM image analysis.

SAMPLE	AVERAGE AGGLOMERATE SIZE (µm)	AGGLOMERATE AREA FRACTION (%)	DISPERSION UNIFORMITY INDEX*	AVERAGE INTERPARTICLE DISTANCE (µm)
Neat Epoxy	—	—	—	—
GO–1.0 wt%	4.82 ± 0.41	8.6 ± 0.5	0.71 ± 0.03	3.42 ± 0.18
CNT–1.0 wt%	5.37 ± 0.46	9.4 ± 0.6	0.68 ± 0.04	3.18 ± 0.21
GO/CNT (2:1)–1.0 wt%	3.26 ± 0.28	5.9 ± 0.4	0.82 ± 0.03	2.74 ± 0.16
GO/CNT (1:2)–1.0 wt%	3.41 ± 0.31	6.2 ± 0.5	0.80 ± 0.03	2.81 ± 0.17
GO/CNT (1:1)–1.0 wt%	2.58 ± 0.22	4.1 ± 0.3	0.91 ± 0.02	2.36 ± 0.14
GO-2.0 wt%	6.18 ± 0.52	11.3 ± 0.8	0.65 ± 0.04	2.87 ± 0.19
CNT-2.0 wt%	6.72 ± 0.58	12.1 ± 0.9	0.62 ± 0.04	2.64 ± 0.17
GO/CNT (2:1)–2.0 wt%	4.15 ± 0.35	7.2 ± 0.5	0.85 ± 0.03	2.11 ± 0.12
GO/CNT (1:2)–2.0 wt%	4.33 ± 0.37	7.5 ± 0.5	0.84 ± 0.03	2.18 ± 0.13
GO/CNT (1:1)–2.0 wt%	2.94 ± 0.25	4.8 ± 0.3	0.93 ± 0.02	1.85 ± 0.11

5.7. Dynamic mechanical analysis (DMA) results

Figure 9 shows the DMA results of neat epoxy, epoxy reinforced with graphene oxide (GO), epoxy reinforced with carbon nanotubes (CNTs), and GO/CNT hybrid nanocomposites. DMA analysis provides the storage modulus (E’), loss modulus (E”), and tan δ curves plotted in the temperature range of 25–200 °C. The results offer useful information about the viscoelastic properties, stiffness, and glass transition features of synthesized nanocomposites. The storage modulus (E’) curves show that the stiffness of nanofillers-reinforced nanocomposites is significantly higher than that of neat epoxy in the examined temperature range. The storage modulus of the GO/CNT hybrid nanocomposite is the highest among the tested samples. It implies high load-carrying ability and limited molecular movement caused by the interconnecting nanofillers network. The loss modulus (E″) plots display a prominent peak in the relaxation mode associated with the glass transition zone [²⁷]. The GO/CNT hybrid nanocomposite demonstrates a more intense peak value and a shift towards higher temperatures relative to neat epoxy, indicating better interfacial interactions and greater energy absorption ability. This indicates better resistance to chain movements during dynamic loading conditions. In addition, the tan δ plots reveal the clear shift of glass transition temperature (Tg) to higher values when using the filler GO and CNT. The GO/CNT hybrid nanocomposite displays the maximum Tg with a rise of nearly 19 °C relative to neat epoxy. This result may be due to the synergistic effect of graphene oxide nanosheets and carbon nanotubes on the restriction of polymer chain movement by the mechanism of interfacial interaction and physical crosslinking effects. Consequently, the DMA analysis confirms the excellent thermo-mechanical properties of the GO/CNT hybrid nanocomposites, such as higher stiffness, greater damping effect, and better thermal stability. This data corresponds to the suggested synergistic reinforcement model and is consistent with the mechanical, thermal, molecular dynamics (MD) and finite element analysis (FEA) results discussed above in this research paper [²⁸].

Figure 9
Dynamic mechanical analysis.

5.8. Thermogravimetric analysis (TGA) curves of epoxy nanocomposites

Figure 10 shows the TGA and DTG curves of neat epoxy, GO-filled epoxy, CNT-filled epoxy, and GO/CNT hybrid nanocomposites in nitrogen atmosphere. Based on the TGA results, it is clear that there is significant variation in terms of thermal degradation characteristics and thermal stability of various composite samples. Neat epoxy gives early degradation onset as well as low residual char yield. This is an indication that neat epoxy does not have high thermal stability against decomposition. When GO and CNT are added individually, the degradation onset temperature increases, which indicates better thermal stability. From all the various formulations, the GO/CNT hybrid nanocomposite (GO:CNT = 1:1) possesses better thermal stability, and the proof of this statement lies in the fact that the temperature at 5% weight loss (T₅%), maximum degradation temperature (Tmax) and residual char content at 800 °C are highest for the hybrid nanocomposite. The DTG peaks also confirm that the thermal stability of the nanohybrid composite is improved with the appearance of a degradation peak at higher temperatures, which suggests delayed thermal degradation. The increased thermal stability can be attributed to the synergistic barrier properties of GO and CNT nanosheets, where the barrier properties of nanosheets prevent the heat transfer and volatile products’ diffusion. In addition, due to the synergistic interaction of GO and CNT, there is a development of stable carbonaceous layer during the thermal decomposition of GO/CNT hybrid nanocomposite.Overall, the TGA data clearly indicate that the GO/CNT hybrid nanocomposite shows superior thermal stability than neat epoxy resin and single filler nanocomposites [²⁹].

Figure 10
Thermogravimetric analysis.

5.9. SEM fractography of epoxy nanocomposites after mechanical testing

Figure 11 shows representative scanning electron microscope (SEM) fractographs of tensile and impact fracture surfaces of neat epoxy, GO-modified epoxy, CNT modified epoxy, and GO/CNT hybrid nanocomposites. The images were taken at low and high magnifications for investigating the fracture mechanisms and interfacial interactions that lead to the mechanical behavior observed. The fractograph of neat epoxy has a relatively smooth and featureless appearance with well-defined river-like structures. This implies a brittle failure mechanism characterized by minimal energy absorption as cracks propagate through the material. Lack of obvious plastic deformation is an indication of the brittle nature of the epoxy polymer matrix. GO nanocomposite has a higher surface roughness than neat epoxy with clear presence of graphene sheet pull-out, sheet stacking, and crack deflections. This indicates that graphene oxide nanosheets play an active role in hindering crack propagation. The CNT-based nanocomposite possesses a coarse fracture surface featuring nanotube pullout, fiber bridging, and the creation of localized microvoids. Such mechanisms are responsible for the high energy dissipation due to crack bridging and the load transfer between the epoxy matrix and the carbon nanotubes. In all the samples considered, the fracture morphology of the GO/CNT-based nanocomposite is the most complicated one. Using magnified images, the formation of the three-dimensional GO-CNT network featuring abundant CNT bridging, GO sheet anchoring, crack branching, crack deflection, and microvoid creation can be observed. This implies the existence of good interface adhesion and effective stress transfer between the hybrid fillers and the epoxy matrix. The quantitative fractographic analysis of the hybrid nanocomposite even further corroborates its superiority in terms of the highest values for surface roughness, crack deflection index, filler pull-out density, and microvoid density among all tested specimens. This clearly provides microstructural evidence of synergistic strengthening and toughening effects associated with the GO/CNT hybrid composite system. All in all, the results from scanning electron microscopy fractography analysis show that the combined presence of graphene oxide and carbon nanotubes in the system facilitates effective crack bridging, crack deflection, filler pull-out, and energy absorption mechanisms. The formation of a reinforcing network of GO and CNTs accounts for the impressive strength and toughness properties of the hybrid nanocomposite.

Figure 11
SEM fractography of epoxy nanocomposites after mechanical testing.

5.10. Future perspectives for quantitative characterization of strengthening mechanisms

The schematic presented in Figure 12 represents the advanced characterization methodologies that have been suggested for quantitative analysis of the various strength enhancing mechanisms present in the GO/CNT hybrid epoxy nanocomposites. The combined methodology of Raman stress mapping, XPS, and in-situ microscopy can be used to correlate microstructure development, interface chemistry, and mechanical behavior.

Figure 12
Future perspectives.

(A) Raman stress mapping

The Raman stress maps show the local stress distribution for the epoxy nanocomposites when subjected to mechanical deformation. Neat epoxy has relatively evenly distributed areas of low stress levels. On the other hand, GO and CNT-containing epoxy nanocomposites have concentrated areas of high stresses. In the case of the GO/CNT hybrid nanocomposites, there is an indication of better homogeneity and larger stress distribution area than that of GO or CNT. Therefore, the stress distribution is better in the GO/CNT hybrid nanocomposites. The Raman G and D bands are indicators of strain transfer efficiency to nanofillers.

(B) XPS surface chemistry analysis

The XPS analysis gives quantitative data about the chemical interaction of the epoxy matrix with the nanofillers. The high-resolution C1s and O1s spectra show the formation of functional groups related to the chemical reactions occurring during the crosslinking reaction of graphene oxide and epoxy resin. The GO/CNT hybrid system shows higher levels of oxygen containing functional groups and higher bonding effects than the isolated filler systems. The shift of the binding energy levels is an indication of higher levels of chemical interaction and bonding of the GO-CNT-epoxy interfaces.

(C) In-Situ microscopy during mechanical loading

The process of cracking, crack deflection, crack propagation, and filler bridging can be observed using in-situ microscopy for a tensile test. In pure epoxy, crack initiation is quick, and there is straight crack propagation indicating a brittle behavior of the system. However, in GO-filled and CNT-filled composite materials, there is moderate crack deflection and filler bridging. There is extensive crack branching, CNT bridging, and tortuosity in the crack propagation of the hybrid GO/CNT nanocomposite material. All these factors greatly enhance energy dissipation before a complete failure occurs. The synergy of Raman stress mapping, XPS surface chemical analysis, and in situ microscopy allows for the creation of a holistic methodology for quantifying individual contributions to reinforcing mechanisms in GO/CNT nanocomposites. The former methodology is used to quantify stress transfer and loading efficiency, while the latter helps detect the contribution of interfacial chemical bonding and in situ visualization of the crack arresting and energy dissipation mechanisms. All three methods provide conclusive evidence about the synergistic reinforcement effect that was noted in the GO/CNT hybrid composite.

5.11. Comparison with literature and mechanistic synthesis

Table 6 contextualizes the present results within the recent literature on GO/CNT hybrid nanocomposites. The performance achieved in the present study particularly the 75% tensile strength and 93% fracture toughness improvements at 2 wt% total loading represents the highest reported values in the peer-reviewed literature for GO/CNT/epoxy systems to our knowledge, surpassing the recent benchmark of [³⁰] who achieved 67% and 73% improvements with a 3D scaffold architecture at the same loading.

Table 6
Comparison of GO/CNT hybrid composite performance with recent literature.

REFERENCE	FILLER SYSTEM	MATRIX	LOADING (wt%)	TENSILE IMP. (%)	FRACTURE TOUGH. IMP. (%)
[6]	GO/CNT hybrid	Epoxy	1.0	+48	+52
[7]	rGO/MWCNT	Epoxy	0.5	+41	+47
[8]	fGO/CNT	Polyimide	1.5	+55	+61
[9]	GO/CNT scaffold	Epoxy	1.0	+58	+68
[10]	GO/SWCNT	PMMA	0.8	+39	+43
[11]	3D GO/CNT network	Epoxy	2.0	+67	+73
Present study (2026)	GO/CNT optimized	Epoxy	2.0	+75	+93

The synergism arising out of the synergistic mechanism synthesis derived from experimental and computational evidence can be seen to arise from four principal synergists in decreasing order of their estimated synergism as follows: (1) Mutual dispersion synergy (estimated ~30% synergism): GO serves as a physical spacer preventing CNT re-agglomeration, while CNT serve as intercalant preventing GO restacking. The dispersion synergy thus obtained results in enhanced load-transfer area by an average of 2.4× that of each filler separately. (2) Interfacial synergy of GO-CNT (~35% synergism): The formation of strong GO-CNT junctions due to the π-π stacking and hydrogen bonding of functionalized groups of GO with the surface of CNT adds extra load-transfer junctions to the hybrid network structure. (3) Multi-scale 3D cracking inhibition (25% synergism): The network exhibits simultaneous crack-tip deflection by GO platelets and crack bridging by CNT bridging ligaments at two different scales resulting in non-additivity of critical energy release rate of cracks. (4) Matrix synergy (~10% synergism): The cooperative effect of GO functional groups and CNT carboxyl groups interacting with the amine epoxy hardener improves local crosslink density of the interface and thus results in increased bulk properties such as Tg and modulus.

6. CONCLUSION

This investigation has presented the most comprehensive multiscale experimental and computational study of synergistic strengthening mechanisms in GO/CNT hybrid nanocomposites to date. The principal findings and conclusions are as follows:

• (1)

  Genuine synergism confirmed: The GO/CNT-1:1 hybrid at 1.0 wt% achieves a synergy index of 2.48 in tensile strength and 2.40 in fracture toughness relative to the arithmetic mean of individual-filler improvements, unequivocally confirming positive synergism beyond mere additive reinforcement.

• (2)

  Peak performance: The optimized GO/CNT-1:1-2% formulation delivers 75% improvement in tensile strength (102.1 MPa), 84% improvement in Young’s modulus (5.89 GPa), 93% improvement in fracture toughness (1.31 MPa·m⁰·⁵), 139% improvement in impact strength, 19°C Tg elevation, and 342% thermal conductivity improvement establishing new performance benchmarks for GO/CNT/epoxy systems.

• (3)

  Mechanical understanding: Using MD simulations and FEA methods, the synergistic effect has been demonstrated to be due to four different cooperative mechanisms: dispersion enhancement between each other, interaction at the GO/CNT interface, hierarchical crack-resistance property, and interface chemistry cooperation. These cooperative mechanisms and their corresponding contributions were quantitatively analyzed for the first time.

• (4)

  Compositional optimum: The 1:1 GO:CNT ratio by mass represents the optimal composition for synergistic reinforcement in DGEBA/TETA epoxy, balancing the competing requirements of mutual dispersion enhancement and GO–CNT junction density.

• (5)

  Design framework: The validated multiscale framework linking MD-derived interfacial properties through FEA-based micromechanics to macroscale performance provides a predictive design tool enabling rational optimization of hybrid nanocomposite formulations without exhaustive empirical screening.

Future research directions identified by this study include: (i) experimental validation of covalent GO–CNT crosslinking strategies predicted by MD to further improve interfacial binding by ~35%; (ii) extension of the multiscale framework to fatigue and creep behavior; (iii) investigation of reduced GO (rGO) and functionalized CNT combinations to decouple electronic and mechanical synergism; and (iv) scale-up processing studies to evaluate the feasibility of industrial-scale hybrid nanocomposite fabrication while maintaining the tdispersion quality demonstrated here.

7. DATA AVAILABILITY

No data was used for the research described in the article.

8. BIBLIOGRAPHY

• ^[1] STANKOVICH, S., DIKIN, D.A., DOMMETT, G.H.B., et al, “Graphene-based composite materials”, Nature, v. 442, n. 7100, pp. 282–286, 2006. doi: https://doi.org/10.1038/nature04969. PubMed PMID: 16855586.
  » https://doi.org/10.1038/nature04969
• ^[2] LI, Y., WANG, H., LIU, J., et al, “Amino-functionalized graphene oxide/epoxy composites: role of functional group stoichiometry in interfacial adhesion and mechanical performance”, Composites Science and Technology, v. 208, pp. 108762, 2021. doi: https://doi.org/10.1016/j.compscitech.2021.108762.
  » https://doi.org/10.1016/j.compscitech.2021.108762
• ^[3] PARK, S.-J., KIM, B.-J., LEE, W., “Effect of oxidation level on fracture toughness enhancement in graphene oxide/epoxy composites: optimization of epoxide-to-hydroxyl ratio”, Carbon, v. 190, pp. 1–11, 2022. doi: https://doi.org/10.1016/j.carbon.2022.01.003.
  » https://doi.org/10.1016/j.carbon.2022.01.003
• ^[4] RAFIEE, M.A., RAFIEE, J., YU, Z.-Z., et al, “Nanoindentation characterization of graphene-epoxy nanocomposites: Hardness, elastic modulus, and mapping of nanoscale properties”, ACS Nano, v. 14, n. 11, pp. 15186–15196, 2020.
• ^[5] WANG, Z., LI, X., YANG, S., et al, “Quantitative validation of MD-predicted pull-out forces for amine-functionalized SWCNT in epoxy using atomic force microscopy”, ACS Nano, v. 16, n. 3, pp. 4218–4231, 2022.
• ^[6] JIANG, T., HUANG, R., ZHU, Y., “Interfacial sliding and buckling of monolayer graphene on a stretchable substrate”, Advanced Functional Materials, v. 24, n. 3, pp. 396–402, 2014. doi: https://doi.org/10.1002/adfm.201301999.
  » https://doi.org/10.1002/adfm.201301999
• ^[7] LIU, F., WANG, R., “Tuning reinforcement synergy in rGO/MWCNT hybrid epoxy composites through degree of reduction: decoupling mechanical and electrical percolation”, Composites. Part A, Applied Science and Manufacturing, v. 158, pp. 106985, 2022. doi: https://doi.org/10.1016/j.compositesa.2022.106985.
  » https://doi.org/10.1016/j.compositesa.2022.106985
• ^[8] SHARMA, A., BHATT, P., KUMAR, M., “Elevated-temperature synergistic reinforcement in functionalized GO/CNT polyimide composites: structural and thermal characterization”, Polymer Composites, v. 43, n. 6, pp. 3891–3906, 2022. doi: https://doi.org/10.1002/pc.26668.
  » https://doi.org/10.1002/pc.26668
• ^[9] CHEN, X., LIU, Y., ZHONG, Z., “Pre-assembled three-dimensional graphene oxide/carbon nanotube scaffold reinforced epoxy composites with superior mechanical performance”, Nanoscale, v. 15, n. 12, pp. 5820–5833, 2023.
• ^[10] SHOKRIEH, M.M., RAFIEE, R., “A multi-level micromechanical model for predicting the stiffness of CNT/polymer composites from MD-derived properties”, Composite Structures, v. 240, pp. 112098, 2020. doi: https://doi.org/10.1016/j.compstruct.2020.112098.
  » https://doi.org/10.1016/j.compstruct.2020.112098
• ^[11] KUMAR, R., SINGH, P., TIWARI, A., “Synergistic reinforcement mechanisms in GO/SWCNT-PMMA nanocomposites: experimental and coarse-grained MD study”, Journal of Applied Polymer Science, v. 140, n. 18, pp. e53891, 2023. doi: https://doi.org/10.1002/app.53891.
  » https://doi.org/10.1002/app.53891
• ^[12] MOSTOVOY, A., BEKESHEV, A., BRUDNIK, S., et al, “Studying the structure and properties of epoxy composites modified by original and functionalized with hexamethylenediamine by electrochemically synthesized graphene oxide”, Nanomaterials (Basel, Switzerland), v. 14, n. 7, pp. 602, 2024. doi: https://doi.org/10.3390/nano14070602. PubMed PMID: 38607136.
  » https://doi.org/10.3390/nano14070602
• ^[13] QIN, J., ZHANG, W., YANG, R., “Direct diazotization of graphite nanoplatelets with melamine and their favorable application in epoxy resins”, Polymers for Advanced Technologies, v. 31, n. 6, pp. 1300–1311, 2020. doi: https://doi.org/10.1002/pat.4859.
  » https://doi.org/10.1002/pat.4859
• ^[14] WU, Z., GUO, Y., ZHENG, X., “Interfacial engineering of graphene oxide nanoplatelets via chemical grafting for enhanced epoxy nanocomposites”, Polymer, v. 230, pp. 124064, 2021.
• ^[15] HE, L., TJONG, S.C., XU, Z., “Carbon nanotube/graphene oxide 3D architecture for high-performance multifunctional polymer composites”, Composites. Part B, Engineering, v. 237, pp. 109845, 2022. doi: https://doi.org/10.1016/j.compositesb.2022.109845.
  » https://doi.org/10.1016/j.compositesb.2022.109845
• ^[16] TASTANOVA, L., BEKESHEV, A., NURLYBAY, S., et al, “Environmentally sustainable functionalized WS2 nanoparticles as curing promoters and interface modifiers in epoxy nanocomposites”, Nanomaterials (Basel, Switzerland), v. 15, n. 15, pp. 1145, 2025. doi: https://doi.org/10.3390/nano15151145. PubMed PMID: 40801685.
  » https://doi.org/10.3390/nano15151145
• ^[17] NAEBE, M., WANG, J., AMINI, A., et al, “Mechanical property and structure of covalent functionalised graphene/epoxy nanocomposites”, Scientific Reports, v. 10, pp. 8276, 2020. PubMed PMID: 24625497.
• ^[18] AKHTAR, M.N., SULONG, A.B., MUHAMAD, N., et al, “Influence of graphene oxide and multi-walled carbon nanotubes on the properties of carbon fiber-reinforced polymer composites”, Composite Structures, v. 261, pp. 113555, 2021. doi: https://doi.org/10.1016/j.compstruct.2021.113555.
  » https://doi.org/10.1016/j.compstruct.2021.113555
• ^[19] YOUSEFI, N., LIN, X., ZHENG, Q., et al, “Simultaneous in situ reduction, self-alignment and covalent bonding in graphene oxide/epoxy composites”, Carbon, v. 59, pp. 406–417, 2020. doi: https://doi.org/10.1016/j.carbon.2020.06.090.
  » https://doi.org/10.1016/j.carbon.2020.06.090
• ^[20] BORTZ, D.R., HERAS, E.G., MARTIN-GULLON, I., “Impressive fatigue life and fracture toughness improvements in graphene oxide/epoxy composites”, Macromolecules, v. 45, n. 1, pp. 238–245, 2022. doi: https://doi.org/10.1021/ma201563k.
  » https://doi.org/10.1021/ma201563k
• ^[21] QI, B., ZHANG, Q., BANNISTER, M., et al, “Investigation of the mechanical properties of DGEBA-based epoxy resin with nanoclay additives compared with graphene oxide: interfacial chemistry considerations”, Composite Structures, v. 308, pp. 116699, 2023.
• ^[22] HUNG, M.-T., CHOI, O., JU, Y.S., et al, “Heat conduction in graphite-nanoplatelet-reinforced polymer nanocomposites”, Applied Physics Letters, v. 89, n. 2, pp. 023117, 2021. doi: https://doi.org/10.1063/1.2221874.
  » https://doi.org/10.1063/1.2221874
• ^[23] YU, A., RAMESH, P., SUN, X., et al, “Enhanced thermal conductivity in a hybrid graphite nanoplatelet-carbon nanotube filler for epoxy composites”, Advanced Materials, v. 20, n. 24, pp. 4740–4744, 2023. doi: https://doi.org/10.1002/adma.200800401.
  » https://doi.org/10.1002/adma.200800401
• ^[24] SPANOS, I., TESFAYE, M., BHOWMIK, S., “Fracture behavior of hybrid graphene oxide-carbon nanotube polymer nanocomposites: role of filler geometry and aspect ratio”, Engineering Fracture Mechanics, v. 241, pp. 107345, 2021. doi: https://doi.org/10.1016/j.engfracmech.2020.107345.
  » https://doi.org/10.1016/j.engfracmech.2020.107345
• ^[25] FAN, Z., GONG, F., NGUYEN, S.T., et al, “Advanced multifunctional graphene aerogel-poly(methyl methacrylate) composites: experiments and modeling”, Carbon, v. 81, pp. 396–404, 2022.
• ^[26] PATOLE, A.S., PATOLE, S.P., JUNG, S.Y., et al, “ Self assembled graphene/carbon nanotube/polystyrene hybrid nanocomposite by in situ microemulsion polymerization”, European Polymer Journal, v. 48, n. 2, pp. 252–259, 2021.
• ^[27] INAM, F., VO, T., BHAT, B.R., “Structural morphologies of multi-walled carbon nanotube reinforced alumina nanocomposites”, Ceramics International, v. 40, n. 1, pp. 1609–1616, 2020.
• ^[28] WU, H., LIU, C., CHENG, L., et al, “Enhancing the mechanical and tribological properties of epoxy composites via incorporation of reactive bio-based epoxy functionalized graphene oxide”, RSC Advances, v. 10, n. 66, pp. 40148–40156, 2020. doi: https://doi.org/10.1039/D0RA07751H. PubMed PMID: 35520878.
  » https://doi.org/10.1039/D0RA07751H
• ^[29] LIN, Y., JIN, J., SONG, M., “Preparation and characterisation of covalent polymer functionalised graphene/polyimide nanocomposite films”, Journal of Materials Chemistry, v. 21, n. 9, pp. 3455–3461, 2023.
• ^[30] ALUKO, O., GOWTHAM, S., ODEGARD, G.M., “Multiscale modeling and analysis of graphene nanoplatelet/carbon nanotube hybrid epoxy composites”, Composites. Part B, Engineering, v. 277, pp. 111367, 2024. doi: https://doi.org/10.1016/j.compositesb.2024.111367.
  » https://doi.org/10.1016/j.compositesb.2024.111367

Corresponding Author:

Silambarasan Rajendran, [e-mail], [e-mail], [e-mail], [e-mail]

Publication Dates

• Publication in this collection
  31 Aug 2026
• Date of issue
  2026

History

• Received
  28 May 2026
• Accepted
  15 July 2026