Accepted Manuscript

Noncovalently functionalized carbon fiber by grafted self-assembled graphene

oxide and the synergistic effect on polymeric shape memory nanocomposites

Haibao Lu, Yongtao Yao, Wei Min Huang, David Hui

PII: S1359-8368(14)00302-3

DOI: http://dx.doi.org/10.1016/j.compositesb.2014.07.022

Reference: JCOMB 3108

To appear in: Composites: Part B

Received Date: 26 May 2014

Revised Date: 11 July 2014

Accepted Date: 20 July 2014

Please cite this article as: Lu, H., Yao, Y., Huang, W.M., Hui, D., Noncovalently functionalized carbon fiber by

grafted self-assembled graphene oxide and the synergistic effect on polymeric shape memory nanocomposites,

Composites: Part B (2014), doi: http://dx.doi.org/10.1016/j.compositesb.2014.07.022

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers

we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and

review of the resulting proof before it is published in its final form. Please note that during the production process

errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

111111111111

Noncovalently functionalized carbon fiber by grafted self-assembled graphene oxide and

the synergistic effect on polymeric shape memory nanocomposites

Haibao Lua,*, Yongtao Yaoa, Wei Min Huangb and David Huic

aScience and Technology on Advanced Composites in Special Environments Laboratory,

Harbin Institute of Technology, Harbin 150080, China

bSchool of Mechanical and Aerospace Engineering, Nanyang Technological University, 50

Nanyang Avenue, Singapore 639798

cComposite Material Research Laboratory, Department of Mechanical Engineering,

University of New Orleans, LA 70148 USA.

*Corresponding author. E-mail address: luhb@hit.edu.cn (H. Lu)

ABSTRACT: This paper presents an effective approach to significantly improve the

electrical properties and recovery performance of shape memory polymer (SMP)

nanocomposites that are able for Joule heating triggered shape recovery. Reduced graphene

oxide (GO) is self-assembled and grafted onto the carbon fibers to enhance the interfacial

bonding with the SMP matrix via van der Waals and covalent crosslink, respectively.

Experimental results verify that the electrical properties of SMP nanocomposites are

significantly improved via a synergistic effect of GO and carbon fiber. The morphology and

porous structure of GO on the carbon fiber are characterized by electron microscope and

optical microscopes, respectively. Furthermore, the behavior of electro-activated recovery and

the resultant temperature distribution within SMP nanocomposite are monitored and

characterized. We demonstrate that this simple way is able to produce electro-activated SMP

nanocomposites which are applicable for Joule heating at a lower electrical voltage.

Keywords: A: Polymer-matrix composites (PMCs), B: Electrical properties, B.

Interface/interphase, E. Assembly

222222222222

1. Introduction

Shape memory polymers (SMPs) are fascinating due to the feature of the shape memory

effect (SME) [1,2]. Most likely this feature was firstly termed in 1932 after such an effect was

discovered in an AuCd alloy [3]. After years of intensive research and development thereafter,

recently SMPs have started to attract great attention as a promising candidate for many

engineering applications in biomedicine, automobile, aerospace and textiles, etc, due to their

high elastic strain, tailorable transition temperature, ease in manufacturing and variable in

mechanical and physical properties within a wide range [4-9]. Consequently, they are now

considered as an important part in smart materials and structures [10,11]. Remarkable

progress in synthesis, characterization/analysis, modeling/simulation and structural design has

been made in a knowledge based approach [12-17]. While the SME in SMPs has been

achieved by applying a few different stimuli, via such as photoactive effect, magnetic-active

effect, water (or solvent)-active effect and electroactive effect, utilization of Joule heating to

drive the SME is desirable and convenient to apply in some practices, especially when direct

heating is not easily achievable. A variety of conductive fillers from zero dimensional

particles, one dimensional chains to two dimensional films, such as carbon nanotubes (CNTs)

[18,19], carbon nanofibers (CNFs) [20,21], carbon black [22], magnetic particles [23], short

carbon fibers [24,25], CNT alignment [26,27], CNF mat [28], nanopaper [29,30] etc have

been used as conductive fillers not only to enhance the electrical conductivity for Joule

heating, but also to improve the thermal conductivity [31-33].

Graphene, a one-atom-thick sheet of sp2-bonded carbon, has attracted enormous interest

because it has superior physical properties: an electrical conductivity of 106 S/cm, a thermal

conductivity of 5000 W/m·K, and a large specific surface area, and a high aspect ratio [34,35].

Bulk quantities of graphene can be produced effectively by the thermal exfoliation of graphite

oxide of which the sheets are reduced and exfoliated simultaneously upon rapid heating [36].

In order to turn graphite oxide into graphene oxide (GO), the most common techniques are

333333333333

sonication, stirring, or a combination of both. Thermally-exfoliated graphene is normally

comprised of few-layer graphene sheets with specific surface areas approximate to 1500 m2/g

[37-39]. Besides oxygen epoxide groups (bridging oxygen atoms), other functional groups

experimentally found are: carbonyl (=CO) and hydroxyl (-OH) prepared using sulphuric acid

(e.g. Hummers method) [40,41]. It is expected that GO should be able to fulfil its roles as a

fixed structure as well as a filler for reinforcement in epoxy-based polymer due to the

potential covalent bonds. Larger specific surface area provides a stronger connection between

CF and polymer. On the other hand, the uniform distribution of GO on CF surfaces enables

good dispersion of GO in the polymer around CF. Even at low volume fractions, the vast

interfacial area created by well-dispersed GO could affect the behavior of the surrounding

polymer matrix and create a co-continuous network to fundamentally change the physical

properties of the polymer matrix [42-44]. In this study, a synergistic effect of GO and carbon

fiber is introduced. The GO is self-assembled and grafted onto the carbon fibers to enhance

the reliability in bonding between carbon fiber and SMP matrix via van der Waals and

covalent crosslink, respectively. A combination of GO and carbon fibers is consequently

utilized to significantly improve the electrical properties to achieve actuation of SMP

composites at a low electrical voltage. A schematic illustration of the working mechanism of

the self-assembled GO grafted onto carbon fiber to improve the interfacial bonding with

epoxy-based SMP is shown in Figure 1.

2. Experimental details

2.1 Raw materials

Commercially available carbon-fibers (12K, 1.80 g/cm3), with an average diameter of 7 μm,

were used as reinforcing fillers in this study. The matrix used in this study is an epoxy-based

fully formable thermoset SMP resin, which was composed of a mixture of epoxy resin and

curing agent. After mixing and curing, this polymer has the thermo-responsive SME, in

addition to high toughness and strength. Its glass transition temperature is 110oC. Natural

444444444444

graphite flakes were received with an average diameter of 10μm. Sulfuric acid and

hydrochloric acid were supplied from Tianjin Chemical Factory, China. KMnO4 was obtained

from Sinopharm Chemical reagent Co., Ltd. (Shanghai, China).

2.2 Preparation of Graphene Oxide

GO was prepared and synthesized following a modified Hummers’ method. The raw

graphite, NaNO3, and sulfuric acid were stirred together for 15 min below 5oC. A certain

amount of KMnO4 was slowly added into the mixture while stirring maintained to prevent the

mixture temperature exceeding 20oC. And then the mixture was transferred into a 25oC water

bath and stirred for 30 min to form a thick paste. Consequently, 200 mL deionized water was

added into the mixture accompanied with strong mechanical stirring. Subsequently, the

mixture was further treated in a 700 mL of deionized water and 60 mL of 30% H2O2 solution.

The solution was then filtered and subjected to cyclic washing with deionized water. Finally,

the resulting graphite oxide was dispersed in deionized water and sonicated for 1 h to obtain

graphene oxide. Finally, the mixture was centrifuged at 7000 rpm for 15 min to get rid of the

unexfoliated graphite oxide.

2.3 Fabrication of SMP nanocomposite

GO has been regarded as a hydrophilic material since it was discovered because of its

excellent water solubility. Distilled water was used as the solvent. As prepared GO of 0.02,

0.03, 0.04, 0.05 and 0.06 g was mixed with 60 mL of distilled water to form suspension. GO

suspension was then sonicated for 30 min. After that, the suspension was filtrated under a

high pressure to enabled GO for self-assemble and subsequently grafted onto the virgin

carbon fibers with the aid of a hydrophilic polycarbonate filter membrane. The suspension

and 60 mL of distilled water were then filtered through the filter. In the next step, the carbon

fibers grafted with GO were dried in an oven at 120 oC for 2 h to further remove the

remaining water. Finally, six pieces of nanopaper with different weight ratio of GO: carbon

fiber (namely, 0: 0.14, 0.02: 0.14, 0.03: 0.14, 0.04: 0.14, 0.05: 0.14 and 0.06: 0.14) were

555555555555

obtained to prepare nanocomposite samples. These six pieces of mat were individually placed

on the bottom of a metallic mold. Resin transfer molding technique was applied to make SMP

nanocomposite. The relative pressure of the resin transfer molding was kept at a constant level

of 6 bar. After the mold was filled, the mixture was cured at a ramp of approximately 1oC/min

from room temperature (22oC) to 100oC and kept for 5 h before being ramped to 120oC at

20oC per 180 min. Finally, it was ramped to 150oC at 30oC per 120 min to produce the final

SMP nanocomposites.

3. Results and discussion

3.1 Morphology of CNT assemblies onto carbon fibers

The surface morphologies and structure of GO functionalized carbon fibers were observed

by field emission scanning electron microscopy (FESEM) at two scales of 2 µm and 10 µm,

respectively, as shown in Figure 2. Figures 2(a), (b) and (c) reveal the surface morphology of

self-assembled GO in which many wide and deep grooves are visible along the axial direction

of carbon fiber. It is shown that different sized GO sheets are attached to the surface and stick

to the carbon fibers along the outer edges of the grooves with different angles due to the high

specific surface area of GO, forming a new hierarchical structure. The GO was grafted onto

and uniformly deposited on the carbon fiber via the membrane filtration process. With an

increase in the weight fraction of GO, GO sheets homogeneously cover the entire surface of

carbon fibers (Figures 2(b) and (c)). Both images reveal that GO sheets can be easily

introduced into the interfacial regions of carbon fiber.

On the other hand, optical microscope observation of self-assembled GO sheets was

performed on using a three-dimensional optical-resolution photoacoustic microscopy (VHX-

900) and Olympus stereomicroscope (OLS3100) equipped with a CCD camera. The

maximum magnifications of the ocular and objective are 10× and 100×, respectively, and thus

the overall magnification is 1000×. Optical microscope provides an effective approach to

characterize and provide three-dimensional visualization of the internal morphology and

666666666666

structure of GO sheets. The GO assembly structure has pores with an average size of around

50 µm, as shown in Figures 3(a) and (b). Therefore the polymer resin is able to penetrate into

this porous structure to closely attach to the GO. It is expected and has been verified that the

GO can significantly improve the interfacial shear strength and interlaminar shear strength of

carbon fiber reinforced polymer composite [45]. A stereomicroscopic three-dimension

observation was conducted on the self-assembled GO to further confirm the internal

morphology and structure. In the testing, the thickness readings were taken from the back of

the mat to determine the thickness of the GO grafted onto carbon fiber mat. The thickness

distribution of the GO is shown in Figure 3(c). The three-dimensional graph reveals the

thickness distribution of the GO on the carbon fiber mat within an area of 480 μm × 640 μm.

It is concluded that the thickness is in a range of 0 to 178.3 μm due to the random layered

structure. At the same time, it is shown that the self-assembled GO is uniformly dispersed on

the carbon fiber.

3.2 Raman spectra characterization

Raman spectroscopy (Horiba HR 800 spectrometer) was used to characterize disorder and

defects in GO. 632 nm radiation from a 20 mW air-cooled Argon ion laser was used as the

exciting source. The graphene and GO could manifest themselves in Raman spectra by means

of changing in relative intensity of two main peaks: D and G bands. Figure 4 shows the

Raman spectra of graphene and GO. The D peak of graphene located at 1348 cm-1 and at 1346

cm-1 for GO streams from a defect-induced breathing mode of sp2 rings. Here, Raman

spectroscopy is shown to provide a powerful tool to differentiate sp2 carbon nanostructures of

graphene and GO which have many properties in common and others that differ. The intensity

of D band is related to the size of the in-plane sp2 domains [46,47]. The decrease of the D

peak intensity indicates the formation of less sp2 domains. The G bonding of C=C shifts from

1591 cm−1 to 1593 cm−1. All sp2 carbon lattice arises from the stretching of C-C bond. The G

peaks are at approximate 1593 cm−1 for GO and at 1591 cm−1 for graphene. The relative

777777777777

intensity ratio of both peaks (ID/IG) is a measure of the degree of disorder and is inversely

proportional to the average size of the sp2 clusters [48]. The D/G intensity ratio for GO is of

0.95 that is larger than that of the graphene of 0.85. This suggest that more graphitic domains

are formed and the sp2 cluster number increases.

3.3 Electrical resistivity measurement

The electrical resistivity of the resultant composites was determined based on the measured

data using a van der Pauw four-point probe apparatus. The apparatus has four probes with an

adjustable inter-probe spacing. A constant current passed through two outer probes and an

output voltage is measured across the inner probes with a voltmeter. In order to reduce

experimental errors, a symmetrical tested sample is preferable. The specific resistivity and

Hall effect of an electrically conductive material is measured by cutting a sample into a

symmetrical shape [49]. Current contact points A and B and voltage contacts points C and D

as shown in Figure 5(a). The electrical resistivity of the carbon fiber mats grafted with

different GO contents (in weight) plotted in Figure 5(b). Figure 5(c) reveals the discrepancy

of measured results. It was found that the average value of electrical resistivity decreases from

1.46, 1.28, 1.23 to 1.16 Ω·cm, as the weight content of GO increases from 0.03, 0.04, 0.05 to

0.06 g, while the weight of the carbon fiber is kept at a constant of 0.14 g. And the

experimental error in the testing is limited to 4%. While the electrical resistivity of 0.05g

reduced graphene oxide being reduced as measurement times increase is resulted from the

more nonuniform dispersion of GOs. The electrical resistivity of tested samples is improved

and lowered with an increase in the weight concentration of GO. It is expected that the more

GO being assembled and invloved, both the amplitude of electric current and current carrying

capability increase, resulted by a lower electrical resistivity. And it should be noted that the

improved electrical properties are resulted by the synergitc effect of GO and carbon fiber.

888888888888

3.4 Electrically triggered shape recovery and temperature distribution

As experimentally demonstrated above, the electrical property of SMP nanocomposite is

improved. Therefore, it is expected that these nanocomposites could show the electrically

induced SME. The tested SMP nanocomposite had a permanent flat shape (60×6×1.5 mm3).

After the nanocomposite was heated above 120oC, it was deformed into a “U” shape upon

application of an external force. After cooling back to room temperature, the freestanding

nanocomposite largely maintained the deformed shape. The synergistic effect of carbon fiber

(0.14 g in weight) and self-assembled GO (0.03 g in weight) on the electrical actuation was

investigated on SMP nanocomposites (1.2 g in weight). A constant 5.5 V DC voltage was

applied to the SMP nanocomposite. While the applied electric current was 0.45 A. The

electro-responsive shape recovery was recorded by a video camera. Snapshots of the shape

recovery sequence are shown in Figure 6. Experimental results reveal that it took 80 s to

complete the shape recovery. There was not much recovery during the first 20 s, but

remarkable recovery was recorded at 60 s and virtually no more further recovery after that till

80 s. Finally, the SMP nanocomposite regained its permanent shape and the recovery ratio is

close to 95%. Simultaneously, an infrared video camera was also used to record and monitor

the shape recovery behavior and temperature distribution in the recovery process. Snap-shot

of the tested SMP nanocomposite is presented in Figure 7. As we can see, higher temperature

is observed where internal strain is higher during the shape recovery process, which is

attributed to the higher local resistive heating loss. When an electrical current is passed

through Joule heating effect results in temperature increased, and thus the shape recovery of

SMP nanocomposite.

4. Conclusion

A series of experiments were conducted to study the synergistic effect of carbon fiber and

GO on SMP nanocomposites, of which the actuation was triggered by means of Joule heating.

The GO was grafted onto the carbon fiber to significantly enhance the reliability in bonding,

999999999999

and further helped to transfer the resistive Joule heating between the carbon fiber and the

SMP. As demonstrated, the electrically driven shape recovery was carried out by means of

applying an electric voltage of 10 V and a power of 2.5 W. Furthermore, temperature

distribution of the SMP nancomposite was monitored during the recovery process to present

the synergistic effect of carbon fiber and rGO on the shape recovery induced by Joule heating.

We demonstrated a simple way to produce electro-activated SMP nanocomposites by

application of deposition of rGO onto carbon fibers, and thus Joule heating was possible at a

lower electrical voltage.

Notes

The authors declare no competing financial interest.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (NSFC)

(Grant No. 51103032) and Fundamental Research Funds for the Central Universities (Grant

No. HIT.BRETIV.201304).

101010101010101010101010

References

[1] Lendlein A, Jiang H, Jünger O, Langer R. Nature (London) 2005;434:879-82.

[2] Tan Q, Liu LW, Liu YJ, Leng JS. Compos Part B-Eng 2013;53:218-25.

[3] Sun L, Huang WM, Ding Z, Zhao Y, Wang CC, Purnawali H, Tang C. Mater Design 2012;

33:577-640.

[4] Jiang HY, Kelch S, Lendlein A. Adv Mater 2006;18:1471-5.

[5] Mather PT, Luo XF, Rousseau IA. Annu Rev Mater Res 2009;39:445-71.

[6] Lantada AD, Morgado PL, Sanz JLM, García JM, Munoz-Guijosa JM, Otero JE. Smart

Mater Struct 2010;19:055022.

[7] Xie T. Polymer 2011;52:4985-5000.

[8] Hu JL, Zhu Y, Huang HH, Lu J. Prog Polym Sci 2012;37:1720-63.

[9] Lu HB, Huang WM, Leng JS. Compos Part B-Eng 2014;62:1-4.

[10]Lu HB, Yao YT, Lin L. Pigm Resin Technol 2013;43:26-34.

[11]Lu HB, Yao YT, Huang WM, Leng JS, Hui D. Compos Part B-Eng 2014;62:256-61.

[12]Nguyen TD, Qi HJ, Castro F, Long KN. J Mech Phys Solids 2008;56:2792-814.

[13]Xie T, Rousseau IA. Polymer 2009;50:1852-6.

[14]Wolfgang W, Karl K, Matthias H, Lendlein A. Adv Polym Sci 2010;226:97-145.

[15]Zhang L, Du HY, Liu LW, Liu YJ, Leng JS. Compos Part B-Eng 2014;59:230-7.

[16]Lu HB, Huang WM. Smart Mater Struct 2013;22:105021.

[17]Lu HB, Du SY. Polym Chem 2014;5:1155-62.

[18]Lu HB, Liang F, Yao YT, Gou JH, Hui D. Compos Part B-Eng 2014;59:191-5.

[19]Lee HF, Yu HH. Soft Matter 2011;7:3801-7.

[20]Lu HB, Gou JH, Leng JS, Du SY. Smart Mater Struct 2011;20:035017.

[21]Tang Z, Sun D, Yang D, Guo B, Zhang L, Jia D. Compos Sci Technol 2013;75:15-21.

[22]Le HH, Zulfiqar A, Mathias U, Ilisch S, Radusch H. J Appl Polym Sci 2011;120:2138-45.

[23]Li H, Zhong J, Meng J, Xian GJ. Compos Part B-Eng 2013;44:508-16.

111111111111111111111111

[24]Leng JS, Lv HB, Liu YJ, Du SY. J Appl Phys 2008;104:104917.

[25]Lu HB, Yu K, Sun SH, Liu YJ, Leng JS. Polym Int 2010;59: 766-71.

[26]Lu HB, Gou JH, Leng JS, Du SY. Appl Phys Lett 2011;98:174105.

[27]Bai PP, Yin WL, Liang F. Nanosci Nanotech Lett 2013;5:732-6.

[28]Luo XF, Mather PT. Soft Matter 2010;6:2146-9.

[29]Lu HB, Liu YJ, Gou JH, Leng JS, Du SY. Compos Sci Technol 2011;71:1427-34.

[30]Lu HB, Huang WM. Appl Phys Lett 2013;102:231910.

[31]Cho JW, Kim JW, Jung YC, Goo NS. Macromol Rapid Commun 2005;26:412-6.

[32]Zhang CS, Ni QQ, Fu SY, Kurashiki K. Compos Sci Technol 2007;67:2973-80.

[33]Lu HB, Liang F, Gou JH. Soft Matter 2011;7:7416-26.

[34]Xing G, Guo H, Zhang X, Sum TC. Huan CHA. Opt Express 2010;18:4564-73.

[35]Dreyer DR, Ruoff RS, Bielawski CW. Angew Chem Int Ed 2010;49:9336-44.

[36]McAllister MJ, Li JL, Adamson DH, Schniepp HC, Abdala AA, Liu J, Herrera-Alonso M,

Milius DL, Car R, Prud’homme RK. Chem Mater 2007;19:4396-404.

[37]Kim H, Macosko CW. Polymer 2009;50:3797-809.

[38]Steurer P, Wissert R, Thomann R, Mülhaupt R. Macromol Rapid Commun 2009;30:316-

27.

[39]Schniepp HC, Li JL, McAllister MJ, Sai H, Herrera-Alonso M, Adamson DH,

Prud’homme RK, Car R, Saville DA, Aksay IA. J Phys Chem B 2006;110:8535-9.

[40]Pandey D, Reifenberger R, Piner R. Surf Sci 2008;602:1607-13.

[41]Eigler S, Dotzer C, Hof F, Bauer W, Hirsch A. Chem-Eur J 2013;19: 9490-6.

[42]Ramanathan T, Abdala AA, Stankovich S, Dikin DA, Herrera-Alonso M, Piner RD,

Adamson DH, Schniepp HC, Chen X, Ruoff RS, Nguyen ST, Aksay IA, Prud'Homme RK,

Brinson LC. Nat Nanotechnol 2008;3:327-31.

[43]Rafiee MA, Rafiee J, Wang Z, Song H, Yu ZZ, Koratkar N. ACS Nano 2009;3:3884-90.

121212121212121212121212

[44]Rafiee MA, Rafiee J, Srivastava I, Wang Z, Song H, Yu ZZ, Koratkar N. Small

2010;6:179-83.

[45]Zhang XQ, Fan XY, Yan C, Li HZ, Zhu YD, Li XT, Yu LP. ACS Appl Mater Interfaces

2012;4:1543-52.

[46]Paredes JI, Villar-Rodil S, Martinez-Alonso A, Tascon JMD. Langmuir 2008;24: 10560-4.

[47]Pan YZ, Bao HQ, Li L. ACS Appl Mater Interfaces 2011;3:4819-30.

[48]Stankovich S, Dikin DA, Piner RD, Kohlhaas KA, Kleinhammes A, Jia Y, Wu Y,

Nguyen ST, Ruoff RS. Carbon 2007;45:1558-65.

[49]van der Pauw LJ. Philips Res Repts 1958;13:1-9.

131313131313131313131313

Figure captions

Fig. 1. (a) Schematic illustration of the graphite stack is oxidized to separate the individual

layers of GO. (b) The role of GO in the interfacial bonding between carbon fiber and epoxy-

based SMP matrix via Van der Waals bonding and covalent crosslink, respectively.

Fig. 2. The typical surface profile of the GO grafted onto the carbon fibers at a scale of (a) 2

µm, (b) 10 µm and (c) 20 µm, respectively.

Fig. 3. (a) and (b) internal morphology and structure of self-assembled GO on the carbon fiber

mat revealed by optical microscopies. (c) 3D graph of thickness distribution of GO.

Fig. 4. Raman spectra of graphene and GO were recorded with 632 nm laser and at a laser

power of 20 mW.

Fig. 5. Electrical resistivity of carbon fiber mat with different weights of GO. (a) Schematic

measurement mode. (b) Electrical resistivity as a function of masurement time for the carbon

fiber mat with 0.03 g, 0.04 g, 0.05 g and 0.06 g of GO. (c) Deviation of measured results.

Fig. 6. Joule heating-induced shape recovery of SMP nanocomposite incorporated with 0.14 g

of carbon fiber and 0.03 g of GO.

Fig. 7. Snapshot of Joule heating-induced shape recovery in SMP nanocomposite recorded by

infrared video camera (VarioCAM HiRessl, JENOPTIK Infra Tec.).

141414141414141414141414

Fig. 1. (a) Schematic illustration of the graphite stack is oxidized to separate the individual

layers of GO. (b) The role of GO in the interfacial bonding between carbon fiber and epoxy-

based SMP matrix via Van der Waals bonding and covalent crosslink, respectively.

151515151515151515151515

Fig. 2. The typical surface profile of the GO grafted onto the carbon fibers at a scale of (a) 2

µm, (b) 10 µm and (c) 20 µm, respectively.

161616161616161616161616

Fig. 3. (a) and (b) internal morphology and structure of self-assembled GO on the carbon fiber

mat revealed by optical microscopies. (c) 3D graph of thickness distribution of GO.

171717171717171717171717

Fig. 4. Raman spectra of graphene and GO were recorded with 632 nm laser and at a laser

power of 20 mW.

181818181818181818181818

Fig. 5. Electrical resistivity of carbon fiber mat with different weights of GO. (a) Schematic

measurement mode. (b) Electrical resistivity as a function of masurement time for the carbon

fiber mat with 0.03 g, 0.04 g, 0.05 g and 0.06 g of GO. (c) Deviation of measured results.

191919191919191919191919

Fig. 6. Joule heating-induced shape recovery of SMP nanocomposite incorporated with 0.14 g

of carbon fiber and 0.03 g of GO.

202020202020202020202020

Fig. 7. Snapshot of Joule heating-induced shape recovery in SMP nanocomposite recorded by

infrared video camera (VarioCAM HiRessl, JENOPTIK Infra Tec.).