High‑Performance Thermal Interface Materials via Co‑Modification of Epoxy Resin with Reduced Graphene Oxide Nanosheets and 3‑D Graphene Networks

Abstract

With the rapid miniaturization of microelectronic devices, efficient heat dissipation has become a critical bottleneck. Graphene‑assisted epoxy resin (ER) systems have shown promise for enhancing thermal performance, yet limitations of reduced graphene oxide (RGO) nanosheets and three‑dimensional graphene networks (3DGNs) still restrict further progress. In this study, we co‑modify ER with both RGO nanosheets and 3DGNs to exploit their complementary advantages: 3DGNs provide a continuous phonon transport network, while RGO nanosheets improve interfacial coupling between graphene and the ER matrix. By optimizing the proportion of each filler and the reduction degree of RGO, we achieve a synergistic enhancement in thermal conductivity, mechanical robustness, and thermal stability, demonstrating the potential of this composite for practical TIM applications.

Background

Thermal interface materials (TIMs) reinforced with graphene have attracted growing interest due to their exceptional thermal and mechanical properties. Kim et al. reported a 1400 % increase in thermal conductivity relative to pristine ER, and Joen’s group achieved ~2 W m⁻¹ K⁻¹ with 10 wt % graphene filler. However, these values fall far short of graphene’s intrinsic conductivity (~5000 W m⁻¹ K⁻¹). The main obstacles are the discontinuous network of RGO sheets and the high interfacial thermal resistance caused by numerous graphene–matrix interfaces and defect‑induced phonon scattering. High‑quality 3DGNs fabricated by chemical vapor deposition offer a continuous, low‑defect network, but their lack of surface functional groups limits wettability with the ER matrix. Introducing a controlled amount of surface defects in 3DGNs improves interfacial bonding, yet the CVD process is complex. Co‑functionalizing 3DGNs with RGO nanosheets presents a promising strategy to combine continuous phonon pathways with enhanced interfacial coupling.

Methods

Materials

Nickel foam (300 gm⁻², 12 mm thick) was used as a template for 3DGNs fabrication. Ethanol, HCl, FeCl₃, and poly(methyl methacrylate) (PMMA, Mₙ≈996 000, 4 % in ethyl lactate) were sourced from Beijing Chemical Reagent Plant. Ethyl lactate, natural graphite, PMMA, and acetone were obtained from Aladdin Co., Ltd. Polytetrafluoroethylene (PTFE) and sodium dodecyl benzene sulfonate were purchased from Huangjiang Co., Ltd. Epoxy resin (ER) and its curing agent were provided by Sanmu Co., Ltd. Deionized water (resistivity 18 MΩ cm) was used for all aqueous preparations.

Preparation

RGO nanosheets and 3DGNs were prepared following our previously reported protocols [12–14] (see Supplementary materials). The RGO–3DGNs–ER composite was fabricated via a two‑step procedure. First, RGO nanosheets and 3DGNs were dispersed in 50 mL deionized water, sonicated for 30 min, and then treated with 1 mg sodium dodecyl benzene sulfonate. The mixture underwent hydrothermal reaction at 80 °C for 6 h in a Teflon vessel, after which it was washed thrice with deionized water. This process anchors RGO nanosheets onto the 3DGNs surface. Second, the RGO–3DGNs composite was blended with ER and curing agent in a mold: a layer of ER was poured onto the solid surface, followed by another layer of RGO–3DGNs, and the sequence was repeated three to four times to ensure capillary infiltration of ER into the porous network. The final mixture was cured at 110 °C for 3 h.

Characterization

SEM (FEI Sirion 200, 5 kV) and TEM (JEM‑2100F, 20 kV) revealed the morphology of the composites. AFM (Nanoscope IIIa and E‑Sweep) was employed in tapping mode to assess surface topography. Raman spectra (LabRam‑1B, 532 nm) quantified defect density, while XPS (RBD PHI‑5000C) confirmed elemental composition. FTIR (IR Prestige‑21) identified functional groups. Mechanical properties were measured using a Triton DMTA (frequency 1 Hz, heating rate 5 °C min⁻¹) on 2 × 4 cm specimens. Thermal conductivity was determined by laser flash analysis, and differential scanning calorimetry provided heat capacity data.

Results and Discussion

Figure 1 shows AFM and SEM images of RGO nanosheets (average lateral size 400–600 nm), 3DGNs (continuous porous network), RGO–3DGNs, and the final RGO–3DGNs–ER composite. The composite surface appears smooth with minimal porosity, indicating good infiltration of ER.

Raman analysis (Fig. 2a) shows prominent G, 2D, and D peaks for RGO, while the D peak is negligible for 3DGNs, indicating low defect density. The I_G/I_D ratio yields an average RGO lateral size of ~500 nm, consistent with AFM data. FTIR spectra (Fig. 2b) reveal a distinct 1335 cm⁻¹ peak (O=C–OH) only in RGO, which disappears after ER integration, confirming covalent bonding at the interface and reduced thermal boundary resistance.

Figure 3 displays the thermal conductivities of the composites. The pristine ER exhibits ~0.2 W m⁻¹ K⁻¹, far below practical requirements. As filler loading increases, conductivity rises almost linearly. Co‑modified RGO–3DGNs–ER samples outperform single‑filler systems at identical mass fractions, confirming synergistic effects. The superior performance originates from the continuous 3DGNs network and improved interfacial bonding via RGO functional groups. Balandin’s model (Eq. 2) links conductivity to thermal boundary resistance (δ); calculations reveal similar δ for RGO–ER and RGO–3DGNs–ER, indicating RGO resides on 3DGNs surfaces, whereas δ for 3DGNs–ER is higher due to poor wettability.

High‑temperature stability tests (Fig. 5a) show all composites exhibit reduced conductivity with increasing temperature due to enhanced Umklapp scattering. However, RGO‑added samples maintain better stability because of more sensitive Kapitza scattering. Long‑term cycling (240 h) reveals <10 % degradation in all composites, confirming robust thermal performance.

Mechanical testing (additional data in Table S1) demonstrates that both 3DGNs–ER and RGO–3DGNs–ER maintain high ultimate strength and elongation, attributed to the continuous 3DGNs scaffold and negligible disruption from surface‑bound RGO.

Conclusions

We have successfully fabricated RGO–3DGNs–ER TIMs that leverage the continuous phonon network of 3DGNs and the interfacial bonding of RGO nanosheets. When 9 wt % 3DGNs and 1 wt % RGO are incorporated, the composite reaches ~4.6 W m⁻¹ K⁻¹—10 % and 36 % higher than 10 wt % 3DGNs or 10 wt % RGO alone—while preserving excellent mechanical properties and thermal stability (≤25 % conductivity loss at 100 °C). These results underscore the promise of co‑modified graphene composites for high‑performance TIMs in advanced electronic applications.