Fabrication of high-k epoxy composites with low dielectric loss based on polymer shell-coated multiwalled carbon nanotubes

Abstract

A core–shell microstructured hybrid was controllably synthesized by coating cross-linked polymer shell onto multiwalled carbon nanotubes (MWCNTs) via direct in situ free-radical polymerization, and was compounded with epoxy to solve the conundrum of percolative composites, namely large dielectric loss caused by direct contact between conductive fillers near the percolation threshold. The shell thickness of the prepared polymer-coated MWCNTs could be varied from 10 to 60 nm by controlling the weight ratio of the monomers. Because the insulative polymer shells serve as electrical barriers between the MWCNTs cores to form a continuous MWCNT-barrier network, the direct contact between the conductive fillers in the composites near the percolation threshold, that is to say the high leakage currents, can be effectively relieved. When the content of the coated MWCNTs reached 10 wt%, the dielectric properties of the composites achieved a dielectric constant value as high as 173.3 and a dielectric loss value as low as 0.021 at 10⁵ Hz. The effect of the thickness of the coating layer on the dielectric properties of the composites was also investigated. This simple method of coating MWCNTs is expected to be applied to materials with a high dielectric constant and low dielectric loss in the field of microelectronics.

---

Introduction

Recent years have witnessed a steady increase in the development of electronic information, power industries, environmentally friendly energy-conversion and storage systems, such as batteries, fuel cells, electrochemical capacitors and dielectric-based capacitors. Among these energy-storage technologies, dielectric capacitors possess an intrinsic high power density due to their very fast energy uptake and delivery, and thus hold great promise for the generation of high performance power electronics used in hybrid electric vehicles, medical devices, and electrical weapon systems.^1–3 Conventional dielectric materials are ceramics or ceramic/polymer composites with large dielectric permittivity, coupled with high stiffness and excellent thermal stability.⁴ However these materials possess intrinsic shortcomings such as high filler content, low mechanical properties and poor processing characteristics.

As a matter of fact, low-cost materials with a high dielectric constant (high-k) and low dielectric loss based on electric conductors and flexible polymers have attracted worldwide attention because of the large possibility of inheriting the performance advantages of polymers compared with ceramic/polymer composites.^5–10 Within the past decade, carbon nanotubes (CNTs) have gained great attention worldwide owing to their special structure and properties. More interestingly, with the addition of an extremely small content of CNTs to a polymer, the dielectric constant will be greatly increased, suggesting that CNT/polymer composites generally have an extremely low percolation threshold.¹¹ However, high-k electric CNT/polymer composites always exhibit very high dielectric loss, which definitely hinders their application. Moreover, the combination of a large surface area and a high aspect ratio with attractive van der Waal interaction forces in CNTs make them aggregate into bundles; thus, they are often difficult to mix with polymers.^12–14

To deal with the problems illustrated above, some researchers devoted themselves to the surface coating of CNTs.^15–21 Zhang et al. prepared unique cyanate ester composites with high dielectric constants based on multi-functional CNTs coated with phosphaphenanthrene terminated hyperbranched polysiloxane. Qiang et al.²² prepared new high-k composites based on unique multi-branched polyaniline and CNT hybrids, in which the multi-branched structure, large polyaniline concentration and flexible Si–O chains help the multi-branched polyaniline coat the surfaces of CNTs through π–π interactions and the CNT hybrids have good dispersion in epoxy matrix. Yu et al.²³ synthesized hierarchical composites of polyaniline fibers on the surface of exfoliated graphite using a chemical oxidation method, which were used to improve the dielectric properties of the electroactive polymer poly(vinylidene fluoride). The dielectric constant and loss tangent of the composites were 17 and 0.06 (103 Hz) when the polyaniline/exfoliated graphite loading was 3 wt%. Other researchers promoted the dielectric performance of the composite by improving the processing method. Chang et al.¹⁰ developed pure CNT/epoxy composites by using microwave curing and Wang et al.²⁴ fabricated high-k CNTs/epoxy composites with low dielectric loss through a layer-by-layer casting technique.

In our previous work, we have introduced a way of coating CNTs by using crosslinkable materials with different chemical reactions, in which the control and constraint of the electrical conductivity of the CNTs were achieved due to the insulating layer that is intrinsic to hybrid materials^12,25,26 such as polyhedral oligomeric silsesquioxane, organic materials^27,28 such as ionic liquids and inorganic materials^29–31 such as boron nitride, titanium dioxide and silica. Here we introduce high-k epoxy composites with low dielectric loss through a home-made hybrid nano-filler according to a controllable coating technique based on the normal free radical polymerization of divinylbenzene (DVB) on the surface of multiwalled CNTs (MWCNTs). Due to the π–π conjugated structure of the monomer, DVB is more inclined to adsorb onto the surface of MWCNTs. Suitable initiators could attack the surface of MWCNTs or DVB, forming MWCNT radicals and monomer free radicals to realize further reaction.³² Consequently, DVB cross-linked polymers were coated onto the MWCNT surface, producing a core–shell microstructure with a MWCNT centre hybrid. Afterwards, we focused on the mass production of poly(DVB)-coated MWCNTs (poly(DVB)@MWCNTs) to produce high performance dielectric MWCNTs/epoxy composites with a high dielectric constant and low dielectric loss.

Experimental

Materials

MWCNTs (average outer diameter: 40–60 nm, length: 5–15 μm, purity: >99.9%), were purchased from Shenzhen Nano Port Company, China. DVB (purity: 80% purchased from J&K company) was used as a cross-linking monomer. AIBN was used to initiate free-radical polymerization and organic solvents, such as N,N-dimethylformamide (DMF), toluene, and methanol, were purchased from Beijing Chemical Factory and rectified before use. Diglycidyl ether of bisphenol A (E-51) with an epoxide equivalent weight of 196 g mol⁻¹ was purchased from Wuxi Resin Plant, China. The curing agent polyether amine D230 was purchased from J&K (purity > 99%).

Preparation of poly(DVB)-coated MWCNTs

The poly(DVB)-coated MWCNTs (denoted as poly(DVB)@MWCNTs) hybrids were prepared by a one-step in situ free-radical polymerization in our lab.³³ In detail, 500 mg pristine MWCNTs were firstly dispersed in 600 ml solvent of DMF. The solution was poured into a three-necked round bottom reaction flask with a certain amount of DVB monomers. After ultrasonic dispersion in an ultrasonic bath at 600 W and room temperature for 1 h, a proportion of initiator (AIBN) was added. The reaction was induced with vigorous stirring under N₂ atmosphere in an oil bath at 80 °C for 12 h. The products were filtered through a 0.22 μm polytetrafluoroethylene membrane and washed with toluene three times to remove impurities and were dried in a vaccum oven at 60 °C for 24 h. The characteristic parameters of the core–shell structured poly(DVB)@MWCNTs hybrids are shown in Table 1 and the ESI.†

Table 1 The characteristic parameters of the core–shell structured poly(DVB)@MWCNTs hybrids

Sample name	Thickness of shell (nm)	Length (μm)	Weight percentage of shell (wt%)
MWCNTs	0	5–15	0
Poly(DVB)@MWCNTs-1	10 ± 2	5–15	23
Poly(DVB)@MWCNTs-2	30 ± 5	5–15	45
Poly(DVB)@MWCNTs-3	60 ± 10	5–15	68

---

Preparation of nanocomposites

In a typical process for preparing poly(DVB)@MWCNTs/epoxy composites, 10 g epoxy resin and appropriate amounts of MWCNTs were thoroughly blended at 80 °C for 40 min with vigorous stirring under sonication to form a homogeneous mixture. After that, polyetheramine D230 (3 g) was added into the mixture with stirring for 10 min to obtain MWCNTs/epoxy prepolymer, and then the obtained prepolymer was degassed under vaccum at room temperature for 30 min. Finally the prepolymer was cast into a mold for post curing and curing following the procedure of 80 °C/2 h + 100 °C/2 h + 120 °C/30 min successively, and then the cured product was demolded to obtain the poly(DVB)@MWCNTs/epoxy composites. The dielectric properties were measured using a broadband dielectric spectrometer (Agilent 4294A instrument with a 16451B fixture) at a constant temperature of 25 °C. The fracture plane samples were investigated using a field emission scanning electron microscope (SEM, HITACHI S-4700). The slice samples were investigated using a high resolution transmission electron microscope (JEM-2100).

Characterization

Fig. 1 shows the dependence of dielectric constants and dielectric losses on frequency for MWCNTs/epoxy with various filler content. It can be seen that the dielectric constant of the composite closely relates to the frequency and this phenomenon is especially prominent when the MWCNT content exceeds 3 wt% (Fig. 1a). On one hand, the dielectric constant obviously decreases with an increase in the frequency. On the other, the dielectric constant of the composite increases with an increase in the proportion of the MWCNTs. Especially when the content of MWCNTs reaches 10 wt%, the value of the dielectric constant soars to 226 at 10⁶ Hz, showing the typical features of percolative polymer composites. The same phenomenon is also reflected on the AC conductivity of the composite (Fig. 2a). The conductivity value of the composite with 10 wt% MWCNTs is about 10⁻² S cm⁻¹ at 10³ Hz and has no obvious frequency dependence compared with the composites whose content of MWCNTs is less than 3 wt%, which indicates the formation of a conductive path at a high filler content. However, the same principle doesn’t apply to the poly(DVB)@MWCNTs/epoxy composites (Fig. 3 and 4).

Fig. 1 Dielectric constants (a) and dielectric losses (b and c) of MWCNTs/epoxy resin composites.

Fig. 2 Frequency dependence of the AC conductivity of MWCNTs/epoxy (a), poly(DVB)@MWCNTs-3/epoxy (b), and poly(DVB)@MWCNTs-1/epoxy composites (c).

Fig. 3 Dielectric constants (a) and dielectric losses (b) of poly(DVB)@MWCNTs-3/epoxy resin composites.

Fig. 4 Dielectric constants (a) and dielectric losses (b and c) of poly(DVB)@MWCNTs-1/epoxy resin composites.

The dielectric constant of the poly(DVB)@MWCNTs/epoxy composite is less sensitive to the change of frequency since the CNT surfaces are coated with an insulative polymer layer, where the conductive path cannot form at the same proportion of fillers compared with pure MWCNTs. At different coating layer thicknesses, the dielectric properties show diverse regulation. The dielectric constant of the poly(DVB)@MWCNTs-3/epoxy composite increases slightly with the increase in filler content (Fig. 3a), as does the AC conductivity (Fig. 2b), which means that the whole composite still exhibits properties of the pure epoxy resin and the percolative phenomenon doesn’t come into being. Thus, the mechanism for the increase in the dielectric constant is mainly due to the Debye polarization²⁴ and the contribution of the MWCNTs does not dominate.

However, with the coating layer becoming thinner, the composite shows a fantastic and contrary performance. For example, the dielectric constant value of the poly(DVB)@MWCNTs-1/epoxy composite (Fig. 4a) is higher compared to the poly(DVB)@MWCNTs-3/epoxy (Fig. 3a) composite with the same filler content at the same frequency. When the content reaches 10 wt%, the dielectric constant achieves a high value of 173.3 at 10⁵ Hz. It is worth noting that the dielectric constant of the poly(DVB)@MWCNTs-1/epoxy composite decreases sharply at a frequency of 10⁶ Hz (Fig. 4a). This phenomenon is attributed to the space charge polarization between the interfaces of the epoxy and coating layer and the MWCNTs, which causes tunneling effects, thus the bound charge can not exist in the interface stably.^34–36 This situation can also be reflected in the AC conductivity data (Fig. 2c), which show one magnitude change compared with the other composites.

The variation trend of the dielectric loss is also different with a change in the frequency. As the filler content of MWCNTs and poly(DVB)@MWCNTs increases, the dielectric loss of the composite shows different trends. The dielectric loss of pure MWCNTs/epoxy rapidly increases with increasing filler concentration, and the dielectric loss is as high as 4.9 at 10⁵ Hz when the MWCNT content reaches 7 wt% (Fig. 1c). The dielectric loss of the poly(DVB)@MWCNTs/epoxy composite could be controlled to levels lower than that of the pure MWCNTs/epoxy composite. As for 10 wt% MWCNTs filled poly(DVB)@MWCNT/epoxy composite, the dielectric loss of the resultant composites is as low as 0.021 (poly(DVB)@MWCNTs-1/epoxy) and 0.027 (poly(DVB)@MWCNTs-3/epoxy) at 10⁵ Hz, respectively (Fig. 4c and 3b).

An interesting fact should be noted that for the poly (DVB)@MWCNTs-3/epoxy composite, the dielectric loss slightly increases with an increase in the proportion of the poly(DVB)@MWCNTs and is not sensitive to the change in frequency (Fig. 3a). However, for poly(DVB)@MWCNTs-1/epoxy, the dielectric loss first increased with the rise in frequency and then dropped. After 10⁷ Hz, the dielectric loss increases again and the value is proportional to the filler content (Fig. 4b). This regulation is consistent with the principle when the fillers concentration are below the percolation threshold.⁴

The dielectric loss of the 10 wt% poly(DVB)@MWCNTs-1/epoxy composite increases sharply at a frequency of 10⁶ Hz (Fig. 4c). This is the comprehensive effect of the space charge polarization and the tunneling effects^35,36 and this regulation is in accordance with the rules below the percolation threshhold.³⁷ Consequently, the poly(DVB)@MWCNTs-1/epoxy composite can possess better dielectric properties with a high dielectric constant (173.3 at 10⁵ Hz) and low dielectric loss (0.021 at 10⁵ Hz) within a certain range of frequency. Furthermore, the method of controlling the coating layer thickness could be used to change the dielectric properties of the material. The dielectric constant and dielectric loss of epoxy composites with 10 wt% varied CNT-based fillers at certain frequencies are compared and shown in Fig. 5.

Fig. 5 The comparison of dielectric properties for epoxy-based composites containing 10 wt% MWCNTs, poly(DVB)@MWCNTs-1 and poly(DVB)@MWCNTs-3.

SEM images of the fractured surfaces of the composites containing 7 wt% MWCNTs are shown in Fig. 6a and b. HRTEM images with the same content of MWCNTs are shown in Fig. 6c and d. For the MWCNTs/epoxy composite, extracted and naked MWCNTs on the outside can be clearly seen in Fig. 6a. Due to the aggregation of MWCNTs, the fractured surfaces of the material shows typical stress concentration morphology. In contrast, Fig. 6b shows that poly(DVB)-coated MWCNTs have good compatibility with the epoxy matrix and nearly no naked tubes could be observed, which demonstrate the interaction between the poly(DVB) coating layers and epoxy matrix. TEM images of the two kinds of composites show the same results. For MWCNTs/epoxy composite, we can see from Fig. 6c that a large amount of pristine MWCNTs form serious twists and aggregates. As for the poly(DVB)@MWCNTs-1/epoxy composite, the amount of tubes in the field of view decreases sharply and the length of the tubes decrease to some extent as well. This phenomenon indicates that the interaction between the MWCNTs and epoxy is strengthened, which once again demonstrates that poly(DVB) coating layers have interactions with epoxy resin.

Fig. 6 SEM and TEM images of MWCNTs/epoxy (a and c) and poly(DVB)@MWCNTs-1/epoxy (b and d) composites containing 7 wt% MWCNTs.

Conclusions

In conclusion, a novel percolative epoxy matrix composite possessing a giant dielectric constant and low dielectric loss was prepared by using polymer-coated MWCNT hybrids as conducting components. When the content of the coated MWCNTs reached 10 wt%, the dielectric properties of the composites achieved a dielectric constant value as high as 173.3 and a dielectric loss value as low as 0.021 at 10⁵ Hz. Above the percolation threshold, the dielectric properties and conductivity of the composite containing coated MWCNTs did not follow the usual insulator–conductor transition power law. This unusual phenomenon could be explained by assuming that DVB shells serve as electrical barriers between the MWCNT cores to form a continuous interparticle-barrier-layer network and retain a stable high-k and low loss. Polymer composite dielectrics combine the properties of the percolative capacitor and barrier-layer capacitor. The excellent compatibility of the polymer shells with the epoxy matrix also ensures the dispersion of fillers in the matrix. The high-k and low loss, easy processability and low cost of these flexible composites make them suited for applications in flexible high-k components.

Acknowledgements

The authors gratefully acknowledge financial support for this work coming from National Natural Science Foundation of China (NSFC) (No. 51573010).

Notes and references

1. Z.-M. Dang, J.-K. Yuan, S.-H. Yao and R.-J. Liao, Adv. Mater., 2013, 25, 6334–6365.
2. Z.-M. Dang, J.-K. Yuan, J.-W. Zha, T. Zhou, S.-T. Li and G.-H. Hu, Prog. Mater. Sci., 2012, 57, 660–723.
3. Z. Zhang, L. Yuan, Z. Qiang, G. Liang and A. Gu, Ind. Eng. Chem. Res., 2015, 54, 938–948.
4. C.-W. Nan, Prog. Mater. Sci., 1993, 37, 1–116.
5. C. Han, A. Gu, G. Liang and L. Yuan, Composites, Part A, 2010, 41, 1321–1328.
6. C. W. Nan, Y. Shen and J. Ma, Annu. Rev. Mater. Res., 2010, 40, 131–151.
7. L. Xie, X. Huang, C. Wu and P. Jiang, J. Mater. Chem., 2011, 21, 5897–5906.
8. X. Zhao, A. A. Koos, B. T. T. Chu, C. Johnston, N. Grobert and P. S. Grant, Carbon, 2009, 47, 561–569.
9. H. Pang, C. Chen, Y.-C. Zhang, P.-G. Ren, D.-X. Yan and Z.-M. Li, Carbon, 2011, 49, 1980–1988.
10. J. Chang, G. Liang, A. Gu, S. Cai and L. Yuan, Carbon, 2012, 50, 689–698.
11. C. Yang, Y. Lin and C. W. Nan, Carbon, 2009, 47, 1096–1101.
12. D. Sun, Z. Zhou, G.-X. Chen and Q. Li, ACS Appl. Mater. Interfaces, 2014, 6, 18635–18643.
13. B. P. Grady, Macromol. Rapid Commun., 2010, 31, 247–257.
14. B. Zhang, Y. Chen, J. Wang, W. J. Blau, X. Zhuang and N. He, Carbon, 2010, 48, 1738–1742.
15. G. Viswanathan, N. Chakrapani, H. Yang, B. Wei, H. Chung, K. Cho, C. Y. Ryu and P. M. Ajayan, J. Am. Chem. Soc., 2003, 125, 9258–9259.
16. J. Zhu, J. Kim, H. Peng, J. L. Margrave, V. N. Khabashesku and E. V. Barrera, Nano Lett., 2003, 3, 1107–1113.
17. G.-X. Chen, H.-S. Kim, B. H. Park and J.-S. Yoon, J. Phys. Chem. B, 2005, 109, 22237–22243.
18. R. R. Nayak, A. M. Shanmugharaj and S. H. Ryu, Macromol. Chem. Phys., 2008, 209, 1137–1144.
19. X. Gong, J. Liu, S. Baskaran, R. D. Voise and J. S. Young, Chem. Mater., 2000, 12, 1049–1052.
20. H. Peng, P. Reverdy, V. N. Khabashesku and J. L. Margrave, Chem. Commun., 2003, 362–363.
21. Y. Tang, J. Gou and Y. Hu, Polym. Eng. Sci., 2013, 53, 1021–1030.
22. Z. Qiang, G. Liang, A. Gu and L. Yuan, Composites, Part A, 2014, 64, 1–10.
23. L. Yu, Y. Zhang, W. Tong, J. Shang, F. Lv, P. K. Chu and W. Guo, Composites, Part A, 2012, 43, 2039–2045.
24. B. Wang, L. Liu, L. Huang, L. Chi, G. Liang, L. Yuan and A. Gu, Carbon, 2015, 85, 28–37.
25. W. Zhang, Z. Zhou, Q. Li and G.-X. Chen, Ind. Eng. Chem. Res., 2014, 53, 6699–6707.
26. D. Sun, Q. Li and G.-X. Chen, Mater. Lett., 2014, 120, 90–93.
27. Q. Li, P. Peng, G. X. Chen and S. W. Yoon, J. Mater. Chem. C, 2014, 2, 8216–8221.
28. G.-X. Chen, S. Zhang, Z. Zhou and Q. Li, Polym. Compos., 2014, 36, 94–101.
29. X. Li, Q. Li and G.-X. Chen, Mater. Lett., 2014, 134, 38–41.
30. G. X. Chen, M. Miyauchi and H. Shimizu, Appl. Phys. Lett., 2008, 92, 203113.
31. G. X. Chen, H. S. Kim, B. H. Park and J. S. Yoon, Carbon, 2006, 44, 3373–3375.
32. G. Sakellariou, D. Priftis and D. Baskaran, Chem. Soc. Rev., 2013, 42, 677–704.
33. Y. Su, Y. Ren, G.-X. Chen and Q. Li, Polymer, 2016 , submitted.
34. B. Lestriez, A. Maazouz, J. Gerard, H. Sautereau, G. Boiteux, G. Seytre and D. E. Kranbuehl, Polymer, 1998, 39, 6733–6742.
35. B. Wang, G. Liang, Y. Jiao, A. Gu, L. Liu, L. Yuan and W. Zhang, Carbon, 2013, 54, 224–233.
36. X. Zhao, J. Zhao, J.-P. Cao, X. Wang, M. Chen and Z.-M. Dang, J. Phys. Chem. B, 2013, 117, 2505–2515.
37. Y. Shen, Y. H. Lin and C. W. Nan, Adv. Funct. Mater., 2007, 17, 2405–2410.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra07945h

---