epoxy resin nanocomposites with high thermal conductivity

International Journal of Heat and Mass Transfer 92 (2016) 15–22

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer journal homepage: www.elsevier.com/locate/ijhmt

Functionalized graphite nanoplatelets/epoxy resin nanocomposites with high thermal conductivity Junwei Gu ⇑, Xutong Yang, Zhaoyuan Lv, Nan Li, Chaobo Liang, Qiuyu Zhang ⇑ Key Laboratory of Space Applied Physics and Chemistry, Ministry of Education, Department of Applied Chemistry, School of Science, Northwestern Polytechnical University, Xi’an, Shaan Xi 710072, PR China

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 7 April 2015 Received in revised form 3 July 2015 Accepted 25 August 2015

Graphite nanoplatelets (GNPs) are performed to fabricate GNPs/bisphenol-A epoxy resin (GNPs/E-51) nanocomposites with high thermal conductivity via casting method. And the ‘‘two-step” method of methanesulfonic acid/c-glycidoxypropyltrimethoxysilane (MSA/KH-560) is introduced to functionalize the surface of GNPs (fGNPs). The KH-560 molecules have been successfully grafted onto the surface of GNPs. The thermal conductivities of the fGNPs/E-51 nanocomposites are increased with the increasing addition of fGNPs, and the corresponding thermally conductive coefficient of the fGNPs/E-51 nanocomposites is improved to 1.698 W/mK with 30 wt% fGNPs, 8 times higher than that of original E-51 matrix. The flexural strength and impact strength of the fGNPs/E-51 nanocomposites are optimal with 0.5 wt% fGNPs. The thermal stabilities of the fGNPs/E-51 nanocomposites are also increased with the increasing addition of fGNPs. For a given GNPs loading, the surface functionalization of GNPs by MSA/KH-560 exhibits a positive effect on the thermal conductivities and mechanical properties of the nanocomposites. Ó 2015 Elsevier Ltd. All rights reserved.

Keywords: Thermally conductive nanocomposites Graphite nanoplatelets (GNPs) Bisphenol-A epoxy resin (E-51) Surface functionalization

1. Introduction Thermal interface materials (TIMs) play an important role in the electronic components area due to the continued miniaturization and light weight [1–3]. Polymers have gained wider applications in different branches of industry because of their light weight, low cost and excellent chemical resistance, etc. [4–7]. However, the intrinsic low thermal conductivities of the polymers have limited their broader applications, especially in the fields of dissipating heat and maintaining operating temperature. In our previous work, several thermally conductive polymeric composites have been successfully fabricated by adding single or hybrid thermally conductive fillers, such as silicon carbide (SiC) [8,9], aluminum nitride (AlN) [10], boron nitride (BN) [11], graphite nanoplatelets (GNPs) [12,13] and SiC whisker/SiC particle (SiCw/SiCp) [14,15]. However, the improvement of the thermal conductivities of the polymeric composites is often less than expected from previous theory design. Furthermore, to fabricate polymeric composites with highly thermal conductivity, the excessive addition of thermally conductive fillers can create a significant challenge of processing behavior and mechanical properties of the polymers [16]. ⇑ Corresponding authors. Tel./fax: +86 29 88431621. E-mail addresses: (Q. Zhang).

[email protected]

(J.

Gu),

[email protected]

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2015.08.081 0017-9310/Ó 2015 Elsevier Ltd. All rights reserved.

Epoxy resins possess high mechanical properties, excellent dimensional & thermal stabilities, low cost and easy processing [17–21]. However, the intrinsic low thermal conductivity of epoxy resins has limited their wider application in microelectronic packaging. Graphite nanoplatelets (GNPs) possess super diameter/ thickness ratio, and can contact with each other easily inner the polymeric matrix [12,13]. Moreover, the value of thermal conductivity for GNPs is reported to be as high as 3000–5000 W/mK [22,23], similar to that of graphene (theoretical value of 5000 W/ mK) [24–26]. However, the price of GNPs is about 65 dollars/kg, much cheaper than that of graphene (more than 500 dollars/kg). Therefore, it is expected that GNPs are suitable for fabricating the epoxy resins nanocomposites with more highly thermal conductivity and a relatively lower cost. In our present work, graphite nanoplatelets (GNPs) are introduced to fabricate GNPs/bisphenol-A epoxy resin (GNPs/E-51) nanocomposites with high thermal conductivity via casting method. And the ‘‘two-step” method of methanesulfonic acid/c-glycidoxypropyltrimethoxysilane (MSA/KH-560) is performed to functionalize the surface of GNPs (fGNPs). The surface performance of pristine GNPs and fGNPs are analyzed and characterized by static precipitation, X-ray photoelectron spectroscopy (XPS), Fourier transform infrared (FTIR) and thermogravimetric analyzer (TGA). In addition, the mass fraction and surface functionalization of GNPs affecting on the mechanical properties, thermal

16

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22

conductivities and thermal stabilities of the nanocomposites are also investigated.

are washed by EtOH and distilled water in sequence, and finally dried at 120 °C in a vacuum oven for 24 h.

2. Experiments

2.3. Fabrication of the nanocomposites

2.1. Materials

The mixtures of E-51 matrix, MeHHPA, DMP-30 and GNPs are stirred uniformly firstly, degassed in a vacuum vessel to remove air bubbles, and then poured into the preheated glass mold. Finally the mixtures above are cured according to the following technology: 100 °C/1 h + 120 °C/2 h + 150 °C/4 h, followed by post-curing at 190 °C for another 3 h.

Graphite nanoplatelets (GNPs), KNG-180, with diameter of 40 lm, super diameter/thickness ratio of 250, are received from Xiamen Knano Graphene Technology Co. Ltd. (Fujian, China); Bisphenol-A epoxy resin (E-51), is received from Xi’an Resin Factory (Shaanxi, China); Both methyl hexahydrophthalic anhydride (MeHHPA) and 2, 4, 6-tris (dimethylaminomethyl) phenol (DMP-30), are purchased from Xi’an Hangang Chemical Group Co., Ltd (Shaanxi, China); Methanesulfonic acid (MSA) is received from Chengdu Kelong Chemical Co. Ltd. (Sichuan, China); c-glycidoxypropyltrimethoxysilane (KH-560) is supplied by Nanjing Shuguang Chemical Group Co., Ltd. (Jiangsu, China); Acetone, ethanol (EtOH) and tetrahydrofuran (THF) are all supplied by Tianjin Fu Yu Fine Chemical Co., Ltd. (Tianjin, China). 2.2. Surface functionalization of GNPs (fGNPs) GNPs are firstly immerged in EtOH and THF for 24 h at room temperature for each step, then washed by distilled water, and finally dried at 100 °C in a vacuum oven for 24 h; The obtained GNPs are then immersed in 30 wt% MSA/distilled water for 36 h at 80 °C, and then washed by 10 wt% NaOH and distilled water in sequence; The mixtures of obtained GNPs (HO-g-GNPs) and KH560/EtOH/distilled water (1/50/50, wt/wt/wt) are reacted for 6 h at 70 °C. Finally, the MSA/KH-560 functionalized GNPs (fGNPs)

2.4. Analysis and characterization X-ray photoelectron spectroscopy (XPS) analyses of the samples are carried out by K-Alpha equipment (Thermo Electron Corporation, USA) to measure element components on the surface of GNPs and fGNPs; Differential scanning calorimetry (DSC) analyses of the samples are carried out at 10 °C/min (nitrogen atmosphere), over the whole range of temperature (40–200 °C) by DSC1 (MettlerToledo Corporation, Switzerland); Thermal gravimetric (TG) analyses of the samples are carried out at 10 °C/min (argon atmosphere), over the whole range of temperature (40–800 °C) by STA 449F3 (NETZSCH, Germany); Scanning electron microscopy (SEM) morphologies of the samples are analyzed by VEGA3-LMH (TESCAN Corporation, Czech Republic); Thermal conductive coefficients of the samples are measured using a Hot Disk instrument (AB Corporation, Sweden) by standard method (Isotropic), which is based upon a transient technique. The measurements are performed on two parallel samples by putting the sensor (3.2 mm diameter) between two slab shape samples. The sensor supplies a heat pulse

Fig. 1. Dispersion states of pristine GNPs and fGNPs in different solvents.

17

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22

of 0.03 W for 20 s to the sample and the associated change in temperature is recorded. And the thermal conductivity of the individual samples is obtained [2]. Flexural strength of the samples is measured by Electron Omnipotence Experiment Machine SANSCMT5105 (Shenzhen New Sansi Corp., China) at ambient temperature according to standard ISO178-1993, the testing speed is 2 mm/min and the corresponding dimension of specimen is 80 mm
15 mm
4 mm. Impact strength of the samples is measured by ZBC-4B impact testing machine (Shenzhen New Sansi Corp., China) at ambient temperature according to standard ISO 179-1993, the impact speed is 2.9 m/s and the corresponding dimension of specimen is 80 mm
10 mm
4 mm. 3. Results and discussion 3.1. Surface functionalization of GNPs The dispersion states of pristine GNPs and fGNPs in different solvents are shown in Fig. 1. Compared with that of pristine GNPs, fGNPs can maintain stability in the acetone, EtOH, distilled water and THF for more than 12 h. The reason is that surface functionalization of GNPs (fGNPs) leads to creation of sterical barrier of stabilization, finally to provide remarkably stable suspension. Fig. 2 shows entire XPS scanning spectra of pristine GNPs and fGNPs, and the results calculated by sensitivity factor are listed in Table 1. There are carbon (C) and oxygen (O) elements on the pristine GNPs’ surface. After the surface functionalization of GNPs (fGNPs), the content of C element is increased slightly. However, O element has a slight decrease. Meanwhile, the silicon (Si) element is also appeared on the fGNPs’ surface, which confirms that KH-560 molecules have been introduced onto the GNPs’ surface. To further analyze the component changes of fGNPs surface, the deconvolution of C1s and Si2p peaks for the spectra of fGNPs is carried out using Gaussian–Lorentzian fit, and the contents of C, Si and other possible groups are calculated according to the square of vary

peaks, shown in Tables 2 and 3. Results show that the –C–Si– peak is only 10.74% of C near 283.64 eV in Cls peak of fGNPs, and the corO responding peaks of –C–C– and are about 72.69% and C O 16.57% of C respectively. Meantime, the devolution of the Si2p shows that the binding energy of 43.28% of Si increases by 2.52 eV besides existing of Si near 100.74 eV. We can deduce that the increasing binding energy of 43.28% of Si is for a chemical bonding of KH-560 to GNPs surface to form –C–O–Si– [9]. Fig. 3 shows FTIR curves of pristine GNPs and fGNPs. Strong peak at 3430 cm1 can be assigned to the hydroxyl (–OH) stretching vibration peak, and the band at 1620 cm1 is ascribed to the absorption vibration peak of C@C bond. The band at 1070 cm1 belongs to the absorption peak of CAO bond. After the surface functionalization of GNPs (fGNPs), a new characteristic vibration peak of SiAC is appeared near 650 cm1, and the band near 1050 cm1 is also appeared owing to the appearance of SiAO from KH-560 molecules. It demonstrates that the KH-560 molecules have been introduced onto the GNPs’ surface. TGA curves of pristine GNPs and fGNPs are presented in Fig. 4. The weight loss of pristine GNPs at 800 °C is less than 1.0 wt%, which is mainly ascribed to the loss of absorbed water on GNPs’ surface. After the surface functionalization of GNPs (fGNPs), the weight of fGNPs is also less than 1.0 wt% at the beginning of the experiment (40–275 °C), the moment is mostly due to rudimental

Table 2 Components of C1s spectra of fGNPs. Number of C1s sub-peaks

Peak position/eV

XAT/%

C1

283.64

10.74

C2

284.28

72.69

C3

285.66

16.57

Corresponding group

C Si C C O

C O

Table 3 Components of Si2p spectra of fGNPs.

Pristine GNPs fGNPs

Number of Si2p sub-peaks

Intensity

Si2s

Si2p

Peak position/ eV

XAT/%

Si1

100.74

56.72

Si2

103.26

43.28

Corresponding group

C Si C O Si

C1s

O1s

C-O C=C

1200

1000

800

600

400

200

0

Si-O

Binding Energy / eV

Si-C

Fig. 2. XPS scanning spectra of pristine GNPs and fGNPs.

Table 1 Elements content on the surface of pristine GNPs and fGNPs. Samples

GNPs fGNPs

Pristine GNPs fGNPs

-OH

Content of elements/%

3500

C

O

Si

86.34 87.95

13.66 9.22

– 2.83

3000

2500

2000

1500

1000

-1

Wavenumber/cm

Fig. 3. FTIR curves of pristine GNPs and fGNPs.

500

18

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22

100 95

100.0 99.5

90 Weight of loss / %

Weight of loss / %

solvent and other volatilization vaporizing on the fGNPs’ surface. The weight loss of the fGNPs reaches 1.4 wt% over the range of 275–425 °C, the moment can be contributed that the KH-560 molecule begins to decompose largely. And the weight loss of the fGNPs gets to about 3.1 wt% in the later stages of the experiment (425–800 °C), the moment KH-560 molecule chars and further decomposes, till all organic compounds volatilizes.

o

275 C

85 80 75

99.0 98.5 98.0

97.0 96.5

70

96.0

65

Pristine GNPs fGNPs

60

100

200

3.2. Thermal conductivities of the nanocomposites

97.5

300

Pristine GNPs fGNPs 300 350 400 450 500 550 600 650 700 750 800 o

Temperature / C

400

500

600

700

800

o

Temperature / C

Thermal conductivity / (W/mK)

Fig. 4. TGA curves of the pristine GNPs and fGNPs.

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

Pritine GNPs/E-51 fGNPs/E-51 0

5

10

15

20

25

30

Mass fraction of GNPs / % Fig. 5. Mass fraction and surface functionalization of GNPs influencing on the thermal conductivities of the nanocomposites.

The mass fraction and surface functionalization of GNPs influencing on the thermal conductivities of the nanocomposites are shown in Fig. 5. The thermal conductivities of the GNPs/E-51 nanocomposites are increased with the increasing addition of GNPs. For a given GNPs loading, the surface functionalization of GNPs by MSA/KH-560 can further improve the thermal conductivities of the nanocomposites. The thermally conductive coefficient of the fGNPs/E-51 nanocomposite is greatly improved to 1.698 W/mK with 30 wt% fGNPs, 8 times higher than that of pristine E-51 matrix (0.201 W/mK). With a small amount addition of GNPs, there is smaller increment for the thermal conductivities of the GNPs/E-51 nanocomposites, which is ascribed to hardly contacting of GNPs–GNPs. With the increasing addition of GNPs, the corresponding thermally conductive channels of GNPs–GNPs are easily formed (Shown in Fig. 6 and Table 4), thus the thermal conductivities of the GNPs/ E-51 nanocomposites are obviously improved. Herein, the EDS analysis can confirm that the white substances (in Fig. 6) on the fracture surface of the fGNPs/E-51 nanocomposites are fGNPs, which could provide the corresponding proof of the GNPs contacts to some extent. For a given GNPs loading, fGNPs possess better interfacial compatibility and lower interfacial thermal resistance with E-51 matrix, which are in favor of the phonon transport, finally to further increase the thermal conductivities of the fGNPs/E-51 nanocomposites. 3.3. Mechanical properties of the nanocomposites Fig. 7 shows the mass fraction and surface functionalization of GNPs influencing on the mechanical properties of the nanocomposites.

Fig. 6. Schematic diagram of thermally conductive channels formation for fGNPs/E-51 nanocomposites with 30 wt% addition of fGNPs.

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22 Table 4 Elements and contents of white substances on the fracture surface of the fGNPs/E-51 nanocomposites. Element

Weight/%

C O Si

93.82 4.15 2.03

Total

100.00

Pristine GNPs/E-51 fGNPs/E-51

2

Impact strength / (kJ/m )

21

18

15

Furthermore, for a given GNPs loading, the surface functionalization of GNPs (fGNPs) can further increase the mechanical properties of the nanocomposites. The reason is that the interfacial compatibility of fGNPs and E-51 matrix is improved. Meanwhile, epoxy group of KH-560 molecules on fGNPs’ surface can react with E-51 matrix, further to enhance the interface bonding strength between fGNPs and E-51 matrix, finally to improve the mechanical properties of the nanocomposites. Fig. 8 shows SEM morphologies of impact fracture for pristine fGNPs/E-51 nanocomposites and fGNPs/E-51 nanocomposites. For a given GNPs loading, the inner defects in the fGNPs/E-51 nanocomposites are decreased obviously (Fig. 8b’ and c’). It reveals that, compared with that of GNPs, fGNPs have relatively better interfacial compatibility with E-51 matrix, which is benefit for decreasing the inner defects, finally to increase the mechanical properties of the nanocomposites. It is also consistent with the results for the mechanical properties of the nanocomposites.

12

3.4. Thermal properties of the nanocomposites 9

6

(a) 0

5

10

15

20

25

30

Mass fraction of GNPs / %

Pristine GNPs/E-51 fGNPs/E-51

135

Flexural strength / MPa

19

120

105

90

75

(b) 60 0

5

10

15

20

25

30

Mass fraction of GNPs /% Fig. 7. Mass fraction and surface functionalization of GNPs influencing on the mechanical properties of the nanocomposites.

Both the flexural strength and impact strength of the nanocomposites are increased up to 0.5 wt% incorporation, but decreased with excessive addition of GNPs. Compared with those of pristine E-51 (flexural strength for 106.2 MPa and impact strength for 13.6 kJ/m2), the maximum flexural strength and impact strength of the fGNPs/E-51 nanocomposite with 0.5 wt% fGNPs is improved to 133.7 MPa and 20.8 kJ/m2, increased by 25.8% and 52.9%, respectively. Appropriate addition of GNPs (0.5 wt%) can effectively transfer stress, cause shear yield and prevent the crack propagation, finally to improve the mechanical properties of the GNPs/E-51 nanocomposites. However, with the excessive addition of GNPs, more interfacial defects and stress concentration points are easily introduced into the E-51 matrix, finally to decrease the mechanical properties of the nanocomposites.

Fig. 9 shows the DSC curves of pristine E-51 and the nanocomposites. The Tg values of the GNPs/E-51 nanocomposites are increased firstly, but decreased with the excessive addition of GNPs. The reason is that, one hand, the addition of GNPs can effectively limit the movement of molecular chains, to increase the Tg values of the GNPs/E-51 nanocomposites. On the other hand, GNPs play a role of effective physical cross-link point inner E-51 matrix [27]. And the proportion of such physical cross-link section in whole E-51 matrix is increased with the increasing addition of GNPs. However, the strength from physical cross-link role is lower than that of chemical bond from chemical cross-link role, therefore, the corresponding Tg values of the GNPs/E-51 nanocomposites are decreased. When the mass fraction of GNPs is less than 0.5 wt%, the influence of the former is superior to that of the latter. On the contrary, the influence of the latter is superior to that of the former when the mass fraction of GNPs is more than 0.5 wt%. Meantime, for a given GNPs loading, compared with those of pristine GNPs/E-51 nanocomposites, the fGNPs/E-51 nanocomposites possess higher Tg values. The reason is that the epoxy group of KH-560 molecules on fGNPs’ surface can react with E-51, further to enhance the interface bonding strength between fGNPs and E-51 matrix, finally to increase the Tg values of the fGNPs/E-51nanocomposites. TGA curves of pristine E-51 and the fGNPs/E-51 nanocomposites are presented in Fig. 10. And the corresponding characteristic thermal data of pristine E-51 and the fGNPs/E-51 nanocomposites are listed in the Table 5. The corresponding weight loss temperatures of the fGNPs/E-51 nanocomposites are all increased with the increasing addition of fGNPs at the same weight loss stage (5 wt% and 30 wt%). The corresponding heat-resisting index is 192 °C (pristine E-51), 195 °C (0.5 wt% fGNPs), 199 °C (10 wt% fGNPs), 200 °C (20 wt% fGNPs) and 203 °C (30 wt% fGNPs), respectively. It indicates that the thermal stabilities of the fGNPs/ E-51 nanocomposites are gradually improved with the increasing addition of fGNPs. The reason is that, compared with that of pristine E-51, fGNPs has more excellent thermal conductivity, which can be easier to absorb external thermal energy. Moreover, the better interfacial compatibility between fGNPs and E-51 matrix can represent a good mix of fGNPs and E-51 matrix, which can also enhance the thermal stabilities of the fGNPs/E-51 nanocomposites. Meanwhile, the corresponding residual mass x values of pristine E-51 and the fGNPs/E-51 nanocomposites are 6.3%, 7.2%, 12.5%, 19.0% and 23.1% respectively, which also reveals the physical role of fGNPs to E-51 matrix.

20

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22

Fig. 8. SEM morphologies of impact fracture for pristine GNPs/E-51 nanocomposites and the fGNPs/E-51 nanocomposites.

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22

30wt% fGNPs/E-51 30wt% GNPs/E-51

Heat flow / (w/g)

20wt% fGNPs/E-51 20wt% GNPs/E-51 10wt% fGNPs/E-51 10wt% GNPs/E-51 0.5wt% fGNPs/E-51 0.5wt% GNPs/E-51 Pristine E-51

Exo 50

75

100

125

150

175

o

Temperature / C Fig. 9. DSC curves of the pristine E-51 and the fGNPs/E-51 nanocomposites.

100

Pristine E-51 0.5 wt% fGNPs/E-51 10 wt% fGNPs/E-51 20 wt% fGNPs/E-51 30 wt% fGNPs/E-51

Weight of loss / %

90 80 70

21

and mechanical properties. The thermal conductivities of the fGNPs/E-51 nanocomposites are increased with the increasing addition of fGNPs, and the corresponding thermally conductive coefficient of the fGNPs/E-51 nanocomposites is improved to 1.698 W/mK (higher than that of GNPs/E-51 nanocomposite for 1.350 W/mK) with 30 wt% fGNPs, 8 times higher than that of pristine E-51 matrix (0.201 W/mK). Both the flexural strength and impact strength of the fGNPs/E-51 nanocomposites are optimal with 0.5 wt% fGNPs. Compared with that of pristine E-51 matrix, the maximum flexural strength and impact strength of the fGNPs/E-51 nanocomposite with 0.5 wt% fGNPs is increased by 25.8% and 52.9%, respectively. Meanwhile, with the increasing addition of fGNPs, the Tg values of the fGNPs/E-51 nanocomposites are increased firstly, but decreased with the excessive addition of GNPs. Furthermore, the thermal stabilities of the fGNPs/E-51 nanocomposites are also improved gradually with the increasing addition of fGNPs. The surface functionalization of GNPs (fGNPs) is helpful for further improving the thermal conductivities and mechanical properties of the fGNPs/E-51 nanocomposites by minimizing interfacial thermal resistance and improving interfacial compatibility between fGNPs and E-51 matrix.

Conflict of interest None declared.

Acknowledgments

60 50 40 30 20 10 0 100

200

300

400

500

600

700

800

The authors are grateful for the support and funding from the Foundation of National Natural Science Foundation of China (Nos. 51403175 and 81400765); Shaanxi Natural Science Foundation of Shaanxi Province (Nos. 2015JM5153 and 2014JQ6203); Space Supporting Fund from China Aerospace Science and Industry Corporation (No. 2014-HT-XGD); Aerospace Science and Technology Innovation Fund from China Aerospace Science and Technology Corporation, and the Fundamental Research Funds for the Central Universities (No. 3102015ZY066).

o

Temperature / C Fig. 10. TGA curves of pristine E-51 and the fGNPs/E-51 nanocomposites.

Table 5 TGA characteristic thermal data of the pristine E-51 and the fGNPs/E-51 nanocomposites. Samples

Pristine E-51 0.5 wt% fGNPs/E-51 10 wt% fGNPs/E-51 20 wt% fGNPs/E-51 30 wt% fGNPs/E-51

Temperatures for weight loss/°C 5 wt%

30 wt%

367 369 372 375 376

403 405 415 418 422

Heat-resistance index*/°C

x/%

192 195 199 200 203

6.3 7.2 12.5 19.0 23.1

T Heat-resistance index ¼ 0:49  ½T 5 þ 0:6  ðT 30  T 5 Þ (1) [13], where T5 and T30 is corresponding decomposition temperature of 5% and 30% weight loss, respectively. * The sample’s heat-resistance index was calculated by Eq. (1).

4. Conclusions Static precipitation, XPS, FTIR and TGA analyses reveal that KH-560 molecules have been grafted on the GNPs’ surface. Compared with those of GNPs/E-51 nanocomposites, all the fGNPs/E-51 nanocomposites possess better thermal conductivities

References [1] B. Tang, G.X. Hu, H.Y. Gao, L.Y. Hai, Application of graphene as filler to improve thermal transport property of epoxy resin for thermal interface materials, Int. J. Heat Mass Transfer 85 (2015) 420–429. [2] H.M. Tu, L. Ye, Thermal conductive PS/graphite composites, Polym. Adv. Technol. 20 (2009) 21–27. [3] X. Fang, Q. Ding, L.W. Fan, H. Lu, Z.-T. Yu, Effects of inclusion size on thermal conductivity and rheological behavior of ethylene-glycol-based suspensions containing silver nanowires with various specific surface areas, Int J Heat Mass Transfer 81 (2015) 554–562. [4] V. Datsyuk, S. Trotsenko, S. Reich, Carbon-nanotube-polymer nanofibers with high thermal conductivity, Carbon 52 (2013) 605–608. [5] Y.C. Zhang, K. Dai, J.H. Tang, X. Ji, Z.M. Li, Anisotropically conductive polymer composites with a selective distribution of carbon black in an in situ microfibrillar reinforced blend, Mater. Lett. 64 (2010) 1430–1432. [6] W. Dai, J.H. Yu, Y. Wang, Y.Z. Song, F.E. Alam, K. Nishimura, C.-T. Lin, N. Jiang, Enhanced thermal conductivity for polyimide composites with a threedimensional silicon carbide nanowire@graphene sheets filler, J. Mater. Chem. A 3 (2015) 4884–4891. [7] P.K. Samantray, P. Karthikeyan, K.S. Reddy, Estimating effective thermal conductivity of two-phase materials, Int. J. Heat Mass Transfer 49 (2006) 4209–4219. [8] J.W. Gu, Q.Y. Zhang, J. Dang, J.P. Zhang, S.J. Chen, Preparation and mechanical properties researches of silane coupling reagent modified b-silicon carbide filled epoxy composites, Polym. Bull. 62 (2009) 689–697. [9] J.W. Gu, Q.Y. Zhang, Y.S. Tang, J.P. Zhang, J. Kong, J. Dang, H.P. Zhang, X.Q. Wang, Studies on the preparation and effect of the mechanical properties of titanate coupling reagent modified b-SiC whisker filled celluloid nanocomposites, Surf. Coat. Technol. 202 (2008) 2891–2896. [10] J.W. Gu, Q.Y. Zhang, J. Dang, J.P. Zhang, Z.Y. Yang, Thermal conductivity and mechanical properties of aluminum nitride filled linear low-density polyethylene composites, Polym. Eng. Sci. 49 (2009) 1030–1034.

22

J. Gu et al. / International Journal of Heat and Mass Transfer 92 (2016) 15–22

[11] J.W. Gu, Q.Y. Zhang, J. Dang, C. Xie, Thermal conductivity epoxy resin composites filled with boron nitride, Polym. Adv. Technol. 23 (2012) 1025– 1028. [12] J.W. Gu, N. Li, L.D. Tian, Z.Y. Lv, Q.Y. Zhang, Highly thermal conductive graphite nanoplatelets/UHMWPE nanocomposites with a low thermal percolation threshold, RSC Adv. 5 (2015) 36334–36339. [13] J.W. Gu, J.J. Du, J. Dang, W.C. Geng, S.H. Hu, Q.Y. Zhang, Thermal conductivities, mechanical and thermal properties of graphite nanoplatelets/polyphenylene sulfide composites, RSC Adv. 4 (2014) 22101–22105. [14] J.W. Gu, Q.Y. Zhang, J. Dang, C.J. Yin, S.J. Chen, Preparation and properties of polystyrene/SiCw/SiCp thermal conductivity composites, J. Appl. Polym. Sci. 124 (2012) 132–137. [15] J.W. Gu, Q.Y. Zhang, J.P. Zhang, W.W. Wang, Studies on the preparation of polystyrene thermal conductivity composites, Polym. Plast. Technol. Eng. 49 (2010) 1385–1389. [16] Z. Han, A. Fina, Thermal conductivity of carbon nanotubes and their polymer nanocomposites: a review, Prog. Polym. Sci. 36 (2011) 914–944. [17] X.F. Zhang, H.Y. Sun, C. Yang, K. Zhang, M.M.F. Yuen, S.H. Yang, Highly conductive polymer composites from room-temperature ionic liquid cured epoxy resin: effect of interphase layer on percolation conductance, RSC Adv. 3 (2013) 1916–1921. [18] J.W. Gu, Q.Y. Zhang, H.C. Li, Y.S. Tang, J. Kong, J. Dang, Study on preparation of SiO2/epoxy resin hybrid materials by means of sol-gel, Polym. Plast. Technol. Eng. 46 (2007) 1129–1134. [19] B.C. Luo, X.H. Wang, Q.C. Zhao, L.T. Li, Synthesis, characterization and dielectric properties of surface functionalized ferroelectric ceramic/epoxy resin composites with high dielectric permittivity, Compos. Sci. Technol. 112 (2015) 1–7.

[20] J. Kong, Y.S. Tang, X.J. Zhang, J.W. Gu, Synergic effect of acrylate liquid rubber and bisphenol A on toughness of epoxy resins, Polym. Bull. 60 (2008) 229–236. [21] J. Jiao, X. Sun, T.J. Pinnavaia, Reinforcement of a rubbery epoxy polymer by mesostructured silica and organosilica with wormhole framework structures, Adv. Funct. Mater. 18 (2008) 1067–1074. [22] M. Mehrali, S.T. Latibari, M. Mehrali, T.M.I. Mahlia, H.S.C. Metselaar, M.S. Naghavi, E. Sadeghinezhad, A.R. Akhiani, Preparation and characterization of palmitic acid/graphene nanoplatelets composite with remarkable thermal conductivity as a novel shape-stabilized phase change material, Appl. Therm. Eng. 61 (2013) 633–640. [23] M. Mehrali, E. Sadeghinezhad, S.T. Latibari, S.N. Kazi, M. Mehrali, M.N.B.M. Zubir, H.S.C. Metselaar, Investigation of thermal conductivity and rheological properties of nanofluids containing graphene nanoplatelets, Nanoscale Res. Lett. 9 (2014) 15. [24] R.R. Nair, P. Blake, A.N. Grigorenko, K.S. Novoselov, T.J. Booth, T. Stauber, N.M.R. Peres, A.K. Geim, Fine structure constant defines visual transparency of graphene, Science 320 (2008) 1308. [25] L.H. Liu, M.D. Yan, Functionalization of pristine graphene with perfluorophenyl azides, J. Mater. Chem. 21 (2011) 3273–3276. [26] S.H. Song, K.H. Park, B.H. Kim, Y.W. Choi, G.H. Jun, D.J. Lee, B.-S. Kong, K.-W. Paik, S. Jeon, Enhanced thermal conductivity of epoxy-graphene composites by using non-oxidized graphene flakes with non-covalent functionalization, Adv. Mater. 25 (2013) 732–737. [27] G. Liu, D.J. Zhang, H. Zhang, H. Zhang, X.F. An, Z. Zhang, X.S. Yi, Study on glass transition temperature of epoxy matrix modified by nano-Al2O3, Thermoset. Resin 24 (2009) 6–9.