Boron nitride@graphene oxide hybrids for epoxy composites with enhanced thermal conductivity

Abstract

Polymer-based materials have widely been used for electronics packaging owing to their excellent physical and chemical properties. However, polymer materials usually have low thermal conductivity, which thus may impair the performance and reliability of modern electronics. In this paper, we report an epoxy-based composite with increased thermal conductivity by using graphene oxide-encapsulated boron nitride (h-BN@GO) hybrids as fillers. The thermal conductivity of the obtained composites increased with the loading of h-BN@GO hybrids to a maximum of 2.23 W m⁻¹ K⁻¹ when the loading of h-BN@GO hybrids was 40 wt%, which is double that of composites filled with h-BN. This increase is attributed to the presence of GO, which improved the compatibility of h-BN with epoxy resin, along with the reduced interfacial thermal resistance between h-BN and epoxy resin. In addition, the effects of h-BN@GO hybrids on the thermal and dielectric properties of epoxy composites were also investigated. The prepared h-BN@GO/epoxy composites exhibit outstanding performance in dimensional stability, slightly reduced thermal stability, and enhanced dielectric properties, which make them suitable as excellent electronics packaging materials.

---

1. Introduction

Polymer-based materials that are used to support and protect modern electronics play an increasingly important role in electronics packaging applications owing to their unique properties such as light weight, easy processing, low cost, etc. With the development of modern electronics towards higher speed and frequency, smaller size and multiple functions, heat removal has become a crucial issue. Unfortunately, most polymer-based materials have an undesirable thermal conductivity of below 0.6 W m⁻¹ K⁻¹, which cannot meet the heat-dissipating requirements of modern electronics.¹

Conventionally, the introduction of thermally conductive fillers into polymers is a common method of increasing the thermal conductivity of polymer-based materials, which is due mainly to it being an easy process and having the potential for large-scale production. Various low-cost ceramic fillers, including alumina (Al₂O₃),² silicon nitride (Si₃N₄),³ aluminium nitride (AlN),^4,5 and silicon carbide (SiC),⁶ etc., have widely been used as thermally conductive fillers. However, the obtained polymer composites still have a low thermal conductivity of below 2.0 W m⁻¹ K⁻¹. In recent years, carbon nanotubes^1,7–9 and graphene^10–15 have also been applied as fillers to fabricate polymer composites with high thermal conductivity, owing to their intrinsic extremely high thermal conductivity. For example, theoretical simulations and experimental results demonstrate that the thermal conductivity of graphene is as high as 6000 W m⁻¹ K⁻¹ (ref. 16). However, the high electrical conductivity of these carbon-based materials limits their applications in the electronics packaging field, in which electrical insulation is required. Among thermally conductive fillers, hexagonal boron nitride (h-BN), which has a similar hexagonal structure to graphite, has attracted far more interest because of its unique advantages.^17–22 For example, in contrast to graphite, h-BN has a wider band gap, lower density and higher thermal conductivity compared with the aforementioned ceramic fillers. Therefore, h-BN has great potential to achieve high thermal conductivity in polymer-based materials.²³ In fact, the increase in thermal conductivity by using h-BN is usually limited, owing to high interfacial thermal resistance between h-BN and polymer matrices.²⁴ In order to minimize interfacial thermal resistance, lots of effort has been devoted to the functionalization of h-BN.^25–27 However, the increase in thermal conductivity is still insufficient, because the functionalization of h-BN usually compromises the intrinsic thermal conductivity of h-BN. Therefore, it is desirable to develop a facile and eco-friendly strategy for functionalizing h-BN.

In this work, we fabricated a novel thermally conductive h-BN@GO hybrid by an electrostatic self-assembly strategy. Epoxy-based composites using this hybrid as a filler were then fabricated. It was found that the thermal conductivity of the obtained composites increased with the loading of h-BN@GO hybrids and reached a maximum of 2.23 W m⁻¹ K⁻¹ when the loading of h-BN@GO hybrids was 40 wt%, which is double that of composites filled with h-BN. This increase can be attributed to the presence of GO, which improves the compatibility of h-BN with epoxy resin, along with the reduced interfacial thermal resistance between h-BN and epoxy resin. Furthermore, the prepared h-BN@GO/epoxy composites exhibited excellent performance in dimensional stability and thermal conductivity, together with dielectric properties, and retained thermal stability for electronics packaging applications.

2. Experimental

2.1 Materials

h-BN powders with a size of 2.0 μm were purchased from Denka, Japan. Graphene oxide nanosheets (GO) were purchased from the Shanxi Institute of Coal Chemistry, China. The coupling agent 3-aminopropyltriethoxysilane (APTES) was purchased from J&K Scientific Co., Ltd (Beijing, China). Liquid epoxy resin (bisphenol A epoxy resin, DER 330) was purchased from Dow Chemical Company. The curing agent (hexahydrophthalic anhydride) and catalyst (N,N-dimethylbenzylamine) were purchased from Sigma-Aldrich. All the materials were used as received.

2.2 Surface modification of h-BN by APTES

h-BN (5.0 g) powders were dispersed in a mixture of ethanol (95 mL) and H₂O (5 mL) by ultrasonication. Formic acid was then added to the solution to accelerate the hydrolysis of APTES, followed by the addition of APTES (0.5 g). The mixture was stirred continuously and refluxed at 90 °C for 24 h under a nitrogen atmosphere. The solution was finally filtered and then washed several times with deionized water, followed by drying in a vacuum at 50 °C for 24 h.

2.3 Preparation of h-BN@GO hybrids

GO aqueous solution (2 mg mL⁻¹, 50 mL) was dispersed in deionized water (500 mL) by ultrasonic treatment. Then, functionalized h-BN powders (0.5 g) were added to the GO solution under mild stirring at 30 °C for 90 min. Owing to the positive charge of GO and negative charge of h-BN, electrostatic interaction between GO and h-BN occurred, which formed h-BN@GO hybrids. Finally, the solution was centrifuged to collect the h-BN@GO hybrids and the h-BN@GO hybrids were washed with deionized water several times to remove the redundant GO. The wet fillers were freeze-dried in a vacuum at −50 °C for 24 h to prevent aggregation of the hybrids.

2.4 Preparation of h-BN@GO/epoxy composites

h-BN@GO/epoxy composites were fabricated by a solvent-free method. In a typical experiment, h-BN@GO hybrid fillers (0.75 g), epoxy resin (7.10 g), curing agent (7.31 g) and catalyst (0.35 g) were put into a sealed ball-milling tank. The rotational speed of the planetary mill was set at 300 rpm to generate a rolling action of the balls, which applied shearing forces to the mixture. Finally, the uniform suspensions of the composite were gently poured into a mold, followed by degassing and heating in a vacuum for further curing. The composites were cured at 120 °C for 2 h and at 150 °C for an additional 2 h. A series of h-BN@GO/epoxy-based composites were fabricated with the h-BN@GO loading increasing from 5 to 40 wt%. For comparison, composites filled with unmodified h-BN were also prepared using the same procedures.

2.5 Characterization

Dynamic light scattering (DLS) was utilized to determine the particle size in the obtained aqueous suspension by using a Mastersizer 3000 laser diffraction particle size analyzer. Fourier transform infrared spectroscopy (FTIR) spectra from 400 to 4000 cm⁻¹ of the h-BN@GO hybrid fillers were recorded using a Bruker Vertex 70 spectrometer with pure KBr as the background. The crystal structure of h-BN@GO was identified by X-ray diffraction (XRD) using a Shimadzu XRD-6000 diffractometer equipped with Cu Kα radiation (λ = 0.154 nm) at 40 kV and 30 mA. The morphology of the fillers and composites was determined using field-emission scanning electron microscopy (FE-SEM, Nova Nano SEM 450, FEI) and transmission electron microscopy (TEM, FEI Tecnai G2 F20 S-TWIN). The thermal stability of the fillers and composites was characterized by thermogravimetric analysis (TGA, Q600, TA Instruments). The measurements were conducted under an air atmosphere at a flow rate of 100 mL min⁻¹ over the temperature range of 30–1000 °C with a ramp rate of 10 °C min⁻¹. Thermal conductivity was measured using a LW-9389 TIM resistance and conductivity measurement apparatus (Long Win Science and Technology, Taiwan) designed according to the American Society for Testing and Materials (ASTM) D-5470 standard. The thermal conductivity was calculated according to eqn (1):

K = −QL/AΔT (1)

where K is the thermal conductivity (W m⁻¹ K⁻¹), Q is the heat flux (W), L and A are the thickness (m) and area (m²) of a specimen, respectively, and ΔT is the temperature difference between temperature sensors of the hot meter bar. The dielectric properties, including the dielectric constant (ε_eff) and dielectric loss (tan δ), were measured using an impedance analyzer (Agilent 4294A) in the frequency range of 1000 Hz to 10 MHz. The coefficient of thermal expansion (CTE) was measured using a thermomechanical analyzer (TMA, NETZSCH Instruments model 402FI) at a heating rate of 5 °C min⁻¹ from 25 °C to 250 °C. The sample length and width were both 2.5 cm, and all values presented in this paper are the average values of at least three measurements of each sample.

3. Results and discussion

3.1 Characterization of h-BN@GO hybrids

h-BN is a two-dimensional material and its density is 2.27 g cm⁻³. The size distribution of h-BN is provided in Fig. 1. DLS analysis shows that the mean diameter of h-BN is 2.0 μm. Fig. 2 shows an AFM image of GO and its corresponding height profile. The height of this film is ∼1 nm between the silicon substrate and GO. It is possible to observe folded or overlapped nanosheets in other regions of the image.

Fig. 1 Results of DLS for aqueous suspension of h-BN.

Fig. 2 AFM image of GO (a) and the corresponding height profile (b).

The h-BN@GO hybrids were prepared via electrostatic self-assembly between h-BN and GO, and the mechanism is illustrated in Fig. 3. It is well known that GO can disperse well in water owing to the presence of hydrophilic groups such as carboxyl, carbonyl, and hydroxyl groups. As for hydrophobic h-BN, functionalization by APTES also allows it to disperse well in water, which is due to the introduction of hydrophilic amino and hydroxyl groups on its surface, as shown in Fig. 3. When the solution of GO and h-BN was mixed, all the fillers quickly settled to the bottom forming brown h-BN@GO hybrids, as shown in Fig. 3. This is a clear indication of self-assembly between GO and h-BN particles. Functionalization of h-BN particles leads to a change in their surface charge from neutral to positive, and thus h-BN has the ability to attract negatively charged GO sheets in water.^24,28 In addition, the whole assembly process is so easy and quick that the h-BN@GO hybrids can be fabricated on a large scale.

Fig. 3 Schematic of the preparation of h-BN@GO hybrids.

Fig. 4 shows the XRD patterns of h-BN, GO and h-BN@GO hybrids. Pristine h-BN exhibits a well-crystallized structure. The distinct characteristic peaks at 26.6°, 41.6°, 43.8°, 50.1°, 55.0°, 75.9°, and 82.1° correspond to the (002), (100), (101), (102), (004), (110) and (112) lattice planes, respectively. Besides the characteristic peaks of h-BN, the h-BN@GO hybrids exhibit one distinct peak at 11.4°, which is similar to that of GO and demonstrates the successful preparation of h-BN@GO hybrids.

Fig. 4 XRD patterns of h-BN, GO and h-BN@GO hybrids.

TGA curves provide further evidence of the successful self-assembly of h-BN with GO, and the GO content in the h-BN@GO hybrids can be determined. Fig. 5 shows the TGA curves of GO, h-BN and h-BN@GO hybrids in air. Pristine h-BN exhibited high thermal stability up to 900 °C without any decomposition, which is consistent with a previous report.²⁹ In contrast, pristine GO displayed poor thermal stability and thermally decomposed completely at 600 °C. As for h-BN@GO hybrids, there was approximately zero mass loss at temperatures below 260 °C and only 5 wt% mass loss when the temperature reached 900 °C, which suggests that the GO content in h-BN@GO is approximately 5 wt%.

Fig. 5 TGA curves for h-BN, GO and h-BN@GO hybrids.

The surface morphologies of h-BN@GO hybrids together with that of pristine h-BN were investigated by SEM and TEM. Fig. 6a shows that pristine h-BN microplatelets with an average lateral size of about 2 μm exhibit a highly flaked structure and smooth surface. However, after electrostatic self-assembly h-BN was decorated with ultrathin GO nanosheets, which show clear creases and rough textures on the h-BN surface (Fig. 6b). The wrinkles of GO on the h-BN surface are further confirmed by the TEM images, in which flexible, ultrathin GO sheets have successfully stuck to the h-BN particles. The tighter wrapping morphology of GO on the h-BN surface allows the removal of absorbed water and air from h-BN, which will improve the interaction with h-BN.

Fig. 6 SEM images of (a) raw h-BN and (b) h-BN@GO hybrids; (c) and (d) TEM images of h-BN@GO hybrids.

The surface chemical composition of h-BN@GO was further characterized by FTIR spectroscopy, as shown in Fig. 7. h-BN exhibits a strong characteristic in-plane B–N stretching vibration at 1367 cm⁻¹ and an out-of-plane bending vibration at 810 cm⁻¹ respectively.^30–32 After the amination reaction of h-BN, several minor bands at around 2982 cm⁻¹, which correspond to C–H stretching vibrations of the hydrocarbon chains of the grafting APTES, are observed.³² In the spectrum of GO, the peaks located at 3422 and 1631 cm⁻¹ correspond to –OH and –C=C groups, respectively. As for the h-BN@GO spectrum, the characteristic peaks of h-BN and GO are all observed, which reveals that GO has stuck to the surface of h-BN. Note that the presence of abundant hydroxyl and carboxyl groups on h-BN@GO will facilitate the chemical interactions of h-BN@GO with polymer matrices.

Fig. 7 FTIR spectra of h-BN, GO, functionalized h-BN and h-BN@GO hybrids.

3.2 Properties of the h-BN@GO/epoxy composites

Fig. 8 shows the thermal conductivity of GO/epoxy, h-BN/epoxy and h-BN@GO/epoxy composites with different loadings of filler. As can be seen, pure epoxy resin had a low thermal conductivity of 0.25 W m⁻¹ K⁻¹, which is in agreement with a previous report.³¹ After the addition of filler, GO, h-BN or h-BN@GO, the thermal conductivity increased with an increase in filler loading. For GO/epoxy composites, the thermal conductivity rose slowly and reached a maximum value of 0.39 W m⁻¹ K⁻¹ at a loading of 2 wt%. This can be attributed to the aggregation of GO in composites, which made it unavailable for the formation of a thermally conductive network. For h-BN/epoxy composites, the increase in thermal conductivity was also limited. For example, the thermal conductivity only increased from 0.25 W m⁻¹ K⁻¹ for pure epoxy resin to 0.8 W m⁻¹ K⁻¹ for the h-BN/epoxy composite with 40 wt% h-BN. In contrast, h-BN@GO was more effective in increasing thermal conductivity than pristine h-BN or GO. For example, at a filler loading of 40 wt%, the thermal conductivity of the h-BN@GO/epoxy composite reached 2.24 W m⁻¹ K⁻¹, which is double and ten times higher than that of h-BN/epoxy composites and pure epoxy resin, respectively. In addition, at a filler loading of below 20 wt%, the effect of h-BN@GO on thermal conductivity was limited. This is due to the fact that at a low filler loading (<20 wt%) most h-BN@GO clusters were separated from each other in the matrix, which made thermal transport inefficient. At a h-BN@GO content of above 20 wt%, thermally conductive networks were gradually formed, which led to a greater increase in thermal conductivity compared with the composites filled with h-BN. There are two possible factors that can explain the greater increase in thermal conductivity for h-BN@GO hybrids compared with h-BN. One factor is that wrapping by GO can bridge the h-BN particles in the epoxy matrix, which promotes the formation of thermally conductive networks. On the other hand, as confirmed by FTIR, the presence of abundant hydroxyl and carboxyl groups on h-BN@GO facilitate the chemical interactions of h-BN@GO with the epoxy matrix and thus reduce the interfacial thermal resistance between h-BN@GO and the epoxy matrix, finally leading to higher thermal conductivity for h-BN@GO/epoxy composites.

Fig. 8 Thermal conductivity of (a) GO/epoxy composites and (b) h-BN@GO/epoxy and h-BN/epoxy composites with different mass fractions.

To confirm the aforementioned deduction, the interfacial thermal resistance between h-BN@GO and the epoxy matrix was estimated according to previous theoretical models.³³ The Maxwell-Garnett effective medium approximation (EMA) has commonly been used to calculate interfacial thermal resistance.³⁴ Unfortunately, it is not able to fit our data, because it is based on the assumption that the fillers are completely surrounded by the matrix. Therefore, the nonlinear model proposed by Foygel et al. should be suitable for our h-BN@GO/epoxy composites. The Foygel model can be described as:³³

K = k₀〈V_f − V_C(β)^t(β)〉 (2)

where K is the thermal conductivity of the composites, V_f is the volume fraction, k₀ is a pre-exponential factor, which depends on the thermal conductivity of the contacting fillers, and V_C is the critical volume fraction of fillers, which is defined as:³⁵

V_C = 0.62/β (3)

where β is the aspect ratio of the fillers and t(β) is a thermal conductivity exponent, which is dependent on the aspect ratio of the fillers. Based on k₀ and t(β), the interfacial thermal resistance can be estimated according to the following formulation:

R₀ = (k₀LV_C^t(β))⁻¹ (4)

Based on the experimental results, we can obtain a value of V_C of 0.02. By fitting the experimental values of thermal conductivity using eqn (2), we can obtain values of the parameters k₀ and t(β), as shown in Fig. 9. The best fit gives values of k₀ for h-BN/epoxy composites and h-BN@GO/epoxy composites of 8 W m⁻¹ K⁻¹ and 30 W m⁻¹ K⁻¹, respectively. We also note that the exponent t(β) for h-BN and h-BN@GO was found to be 1.5 and 1.7, respectively, which is the signature of a three-dimensional (3D) transport process. According to eqn (2), k₀ is the expected contribution of the thermally conductive filler network alone. Obviously, the value of k₀ of h-BN@GO/epoxy composites is much higher than that of h-BN/epoxy composites and is comparable to that of a single-walled carbon nanotube network in poly(methyl methacrylate) (PMMA) composites. It is 100 times higher than the thermal conductivity of epoxy resin. However, it is still much lower than the value of 260 W m⁻¹ K⁻¹ of pure h-BN, which indicates that the thermally conductive network constructed from h-BN@GO has deteriorated due to interfacial thermal resistance. Using the obtained values of the parameters k₀ and t(β), the values of R₀ for h-BN/epoxy and h-BN@GO/epoxy resin composites are 4.10 × 10⁸ K W⁻¹ and 2.37 × 10⁸ K W⁻¹, respectively. Obviously, the interfacial thermal resistance of h-BN@GO/epoxy composites is lower than that of h-BN/epoxy composites, which confirms our deduction that GO can not only be used as a bridge between h-BN but also improves the compatibility between h-BN and epoxy resins.

Fig. 9 Comparison between the simulated thermal conductivity based on the Foygel model and the experimental values.

A thermal percolation phenomenon has not been explicitly explained. Some studies have demonstrated thermal percolation behavior in carbon-based composites,³⁶ whereas others have shown continuous linear dependence.³⁷ In our work, we believe that the loading of the matrix with h-BN@GO, which increased the thermal conductivity, is characterized by a percolation threshold. The conductivity of a polymer composite in the vicinity of the percolation threshold is described by eqn (2). Using Foygel's results and the diameter and length of our h-BN@GO particles, we estimate the value of V_C = 0.02. The best fit to eqn (2) with k₀ and t(β) as fitting parameters provides values of k₀ = 30 W m⁻¹ K⁻¹ and t(β) = 1.7 for h-BN@GO/epoxy composites. The fit follows all of the experimental data points, which suggests that the thermal conductivity relies on a percolation network of h-BN@GO, rather than isolated filler particles. Below the percolation threshold (V_f < 2 vol%), h-BN@GO particles are not sufficiently in contact, which causes phonon scattering, and the thermal conductivity is below 0.62 W m⁻¹ K⁻¹. Above the thermal conductivity percolation threshold (V_f > 2 vol%), the rise in the thermal conductivity indicates that a conducting network with an increased number of direct h-BN@GO–h-BN@GO contacts and a decreased number of polymer-mediated boundaries has been formed. Therefore, heat is effectively transported through the percolation network.

Thermal stability is one of the most important properties of the composites, because it influences the processing and service life performance of the composites. The TGA curves of pure epoxy resin and h-BN@GO/epoxy composites with different loadings of h-BN@GO are shown in Fig. 10. The composites remained stable up to 300 °C and then completely decomposed at above 600 °C, leaving different residues for different h-BN@GO loadings. It was found that T_d5%, which is an indication of thermal stability, increased with the addition of h-BN@GO. In detail, T_d5% gradually increased from 298 °C for pure epoxy resin to 335 °C for the composite containing 40 wt% h-BN@GO. This is mainly attributed to stabilization by h-BN, which almost does not decompose below 1000 °C.³⁸

Fig. 10 TGA curves for epoxy resin and composites with h-BN@GO hybrids.

Fig. 11 shows the dependence on frequency of the dielectric permittivity and loss for pure epoxy resin and h-BN@GO/epoxy composites. The dielectric permittivity of the composites was found to decrease as the frequency changed from 1 kHz to 10 MHz (Fig. 11a), owing to the orientation of the dipoles in the polymer chains. Owing to the comparatively high dielectric permittivity of h-BN and GO, the addition of the h-BN@GO composites resulted in a modest increase in dielectric permittivity. For example, the dielectric permittivity increased from 6.8 for pure epoxy resin to 12.5 for the 20 wt% h-BN@GO/epoxy composite. The dielectric loss of the h-BN@GO/epoxy composites with different fillers also decreased with an increase in frequency in the range of 1 kHz to 10 MHz (Fig. 11b). Furthermore, the addition of h-BN@GO led to an increase in dielectric loss at low frequency from 10³ Hz to 10⁵ Hz, and the dielectric loss was independent of the h-BN@GO loading. At high frequency, the effect of h-BN@GO on the dielectric loss was negligible, which is attributed to the presence of GO. As proved by a previous study,³⁹ the addition of GO to polymer composites will lead to an increase in dielectric loss owing to interfacial polarization at frequencies of below 10⁴ Hz. Despite the increase in dielectric loss, it still remained lower than 0.05 at frequencies of above 10⁶ Hz. Therefore, the h-BN@GO/epoxy composites with high thermal conductivity and an increased dielectric constant are suitable for applications in semiconductor devices packaging, in which both high thermal conductivity and an increased dielectric constant are required.

Fig. 11 Dependence of (a) the dielectric permittivity and (b) the dielectric loss of h-BN@GO/epoxy composites on the frequency.

Dimensional stability is a key property of polymer composites, because a mismatch in dimensional stability between two different materials will lead to deformation of the products. The coefficient of thermal expansion is an indication of the dimensional stability of the composites and can be determined by thermomechanical analysis (TMA) within the temperature range from 25 to 200 °C, as shown in Fig. 12a. The values of CTE in the glassy region and rubbery region,^40,41 which are marked CTE1 and CTE2, were determined by the slope of a linear regression plot in the temperature intervals of 40–100 °C and 180–300 °C, respectively, based on eqn (5):

where L₀ is the initial length of the sample and L is the length at temperature T. Fig. 12b shows the values of CTE1 and CTE2 of the h-BN@GO/epoxy composites as a function of different loading contents. Both CTE1 and CTE2 gradually decreased as the filler loading increased. For h-BN@GO/epoxy composites with a loading of 40 wt% h-BN@GO, CTE1 and CTE2 were found to be 43.1 and 167.2 ppm °C⁻¹, respectively, which decreased by about 23.1 and 28.9 ppm °C⁻¹ compared with neat epoxy resin. This reduction is attributed to suppression of the change in volume in the alternative layered structure between h-BN@GO and epoxy resin in the interstitial space.

Fig. 12 (a) Dynamic mechanical curves of neat epoxy resin and h-BN@GO/epoxy composites with different loadings of filler. (b) CTE values of h-BN@GO/epoxy composites in the glassy region and rubbery region.

4. Conclusions

h-BN@GO hybrids were prepared by the electrostatic self-assembly of h-BN and GO, and their epoxy composites were fabricated by a solution-free curing process. The effects of h-BN@GO hybrids on the thermal and dielectric properties of epoxy composites were investigated. The thermal conductivity of the h-BN@GO composites reached 2.23 W m⁻¹ K⁻¹ at a h-BN@GO content of 40 wt%, which is approximately double that of composites filled with h-BN at the same loading. The increase is attributed to the presence of GO, which improved the compatibility of h-BN with epoxy resin, along with the reduction in interfacial thermal resistance between h-BN and epoxy resins. Theoretical analysis based on the nonlinear Foygel model demonstrates that the interfacial thermal resistance of h-BN@GO/epoxy composites (2.37 × 10⁸ K W⁻¹) is lower than that of h-BN/epoxy composites (4.10 × 10⁸ K W⁻¹). Furthermore, the prepared h-BN@GO/epoxy composites exhibited outstanding performance in dimensional stability, slightly reduced thermal stability, and enhanced dielectric properties. The high performance of the h-BN@GO/epoxy composites make them have potential applications in microelectronics packaging and as structural capacitors for aerospace devices.

Acknowledgements

The authors are grateful for the financial support from Guangdong and Shenzhen Innovative Research Team Program (No. 2011D052 and KYPT20121228160843692).

Notes and references

1. M. Bozlar, D. He, J. Bai, Y. Chalopin, N. Mingo and S. Volz, Adv. Mater., 2010, 22, 1654–1658.
2. X. Wu, P. Jiang, Y. Zhou, J. Yu, F. Zhang, L. Dong and Y. Yin, J. Appl. Polym. Sci., 2014, 131, 40528.
3. T. Kusunose, T. Yagi, S. H. Firoz and T. Sekino, J. Mater. Chem. A, 2013, 1, 3440–3445.
4. Z. Shi, M. Radwan, S. Kirihara, Y. Miyamoto and Z. Jin, Appl. Phys. Lett., 2009, 95, 224104.
5. Y. Zhou, H. Wang, L. Wang, K. Yu, Z. Lin, L. He and Y. Bai, J. Mater. Sci. Eng. B, 2012, 177, 892–896.
6. J. P. Cao, X. Zhao, J. Zhao, J. W. Zha, G. H. Hu and Z. M. Dang, ACS Appl. Mater. Interfaces, 2013, 5, 6915–6924.
7. H. Chen, M. Chen, J. Di, G. Xu, H. Li and Q. Li, J. Phys. Chem. C, 2012, 116, 3903–3909.
8. W. Cui, F. Du, J. Zhao, W. Zhang, Y. Yang, X. Xie and Y.-W. Mai, Carbon, 2011, 49, 495–500.
9. M. Wang, H. Chen, W. Lin, Z. Li, Q. Li, M. Chen, F. Meng, Y. Xing, Y. Yao, C. P. Wong and Q. Li, ACS Appl. Mater. Interfaces, 2014, 6, 539–544.
10. S. Araby, L. Zhang, H.-C. Kuan, J.-B. Dai, P. Majewski and J. Ma, Polymer, 2013, 54, 3663–3670.
11. X. Huang, C. Zhi and P. Jiang, J. Phys. Chem. C, 2012, 116, 23812–23820.
12. Q.-Q. Kong, Z. Liu, J.-G. Gao, C.-M. Chen, Q. Zhang, G. Zhou, Z.-C. Tao, X.-H. Zhang, M.-Z. Wang, F. Li and R. Cai, Adv. Funct. Mater., 2014, 24, 4222–4228.
13. B. Shen, W. Zhai and W. Zheng, Adv. Funct. Mater., 2014, 24, 4542–4548.
14. S. H. Song, K. H. Park, B. H. Kim, Y. W. Choi, G. H. Jun, D. J. Lee, B. S. Kong, K. W. Paik and S. Jeon, Adv. Mater., 2013, 25, 732–737.
15. C.-C. Teng, C.-C. M. Ma, C.-H. Lu, S.-Y. Yang, S.-H. Lee, M.-C. Hsiao, M.-Y. Yen, K.-C. Chiou and T.-M. Lee, Carbon, 2011, 49, 5107–5116.
16. Y. Zhu, S. Murali, W. Cai, X. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Adv. Mater., 2010, 22, 3906–3924.
17. Y. Lin and J. W. Connell, Nanoscale, 2012, 4, 6908–6939.
18. P. Miro, M. Audiffred and T. Heine, Chem. Soc. Rev., 2014, 43, 6537–6554.
19. A. Pakdel, Y. Bando and D. Golberg, Chem. Soc. Rev., 2014, 43, 934–959.
20. L. H. Li, J. Cervenka, K. Watanabe, T. Taniguchi and Y. Chen, ACS Nano, 2014, 8, 1457–1462.
21. L. H. Li, E. J. Santos, T. Xing, E. Cappelluti, R. Roldan, Y. Chen, K. Watanabe and T. Taniguchi, Nano Lett., 2015, 15, 218–223.
22. L. H. Li, T. Xing, Y. Chen and R. Jones, Adv. Mater. Interfaces, 2014, 1, 1300132.
23. S. Y. Pak, H. M. Kim, S. Y. Kim and J. R. Youn, Carbon, 2012, 50, 4830–4838.
24. J. Zhang, Y. Hong and Y. Yue, J. Appl. Phys., 2015, 117, 134307.
25. S. Shahid, K. Nijmeijer, S. Nehache, I. Vankelecom, A. Deratani and D. Quemener, J. Membr. Sci., 2015, 492, 21–31.
26. M. Wang, Z. Wang, N. Li, J. Liao, S. Zhao, J. Wang and S. Wang, J. Membr. Sci., 2015, 495, 252–268.
27. N. Tajaddod, K. Song, E. C. Green, Y. Zhang and M. L. Minus, Macromol. Mater. Eng., 2016, 301, 315–327.
28. T. Zhou, S. Nagao, T. Sugahara, H. Koga, M. Nogi, K. Suganuma, T. T. Nge and Y. Nishina, RSC Adv., 2015, 5, 20376–20385.
29. J. Yu, X. Huang, C. Wu, X. Wu, G. Wang and P. Jiang, Polymer, 2012, 53, 471–480.
30. T. H. Chiang and T. E. Hsieh, J. Inorg. Organomet. Polym. Mater., 2006, 16, 175–183.
31. M. Donnay, S. Tzavalas and E. Logakis, Compos. Sci. Technol., 2015, 110, 152–158.
32. T. Fujihara, H.-B. Cho, T. Nakayama, T. Suzuki, W. Jiang, H. Suematsu, H. D. Kim, K. Niihara and J. Blendell, J. Am. Ceram. Soc., 2012, 95, 369–373.
33. R. J. Warzoha and A. S. Fleischer, Nano Energy, 2014, 6, 137–158.
34. C.-W. Nan, G. Liu, Y. Lin and M. Li, Appl. Phys. Lett., 2004, 85, 3549–3551.
35. M. Foygel, R. D. Morris, D. Anez, S. French and V. L. Sobolev, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 71, 104201.
36. J. Xiang and L. T. Drzal, Sol. Energy Mater. Sol. Cells, 2011, 95, 1811–1818.
37. K. M. Shahil and A. A. Balandin, Nano Lett., 2012, 12, 861–867.
38. W. Zhou, S. Qi, H. Li and S. Shao, Thermochim. Acta, 2007, 452, 36–42.
39. J. Y. Kim, W. H. Lee, J. W. Suk, J. R. Potts, H. Chou, I. N. Kholmanov, R. D. Piner, J. Lee, D. Akinwande and R. S. Ruoff, Adv. Mater., 2013, 25, 2308–2313.
40. K. C. Yung and H. Liem, J. Appl. Polym. Sci., 2007, 106, 3587–3591.
41. C. Zhi, Y. Bando, T. Terao, C. Tang, H. Kuwahara and D. Golberg, Adv. Funct. Mater., 2009, 19, 1857–1862.

---