Composites: Part B 59 (2014) 191–195

Contents lists available at ScienceDirect

Composites: Part B

journal homepage: www.elsevier .com/locate /composi tesb

Self-assembled multi-layered carbon nanofiber nanopaperfor significantly improving electrical actuation of shape memorypolymer nanocomposite

1359-8368/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.compositesb.2013.12.009

⇑ Corresponding authors. Tel.: +86 45186412259 (H. Lu).E-mail addresses: luhb@hit.edu.cn (H. Lu), Jihua.Gou@ucf.edu (J. Gou).

Haibao Lu a,⇑, Fei Liang b, Yongtao Yao a, Jihua Gou b,⇑, David Hui c

a Science and Technology on Advanced Composites in Special Environments Laboratory, Harbin Institute of Technology, Harbin 150080, Chinab Composite Materials and Structures Laboratory, Department of Mechanical, Materials and Aerospace Engineering, University of Central Florida, Orlando, FL 32816, USAc Composite Material Research Laboratory, Department of Mechanical Engineering, University of New Orleans, LA 70148, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 10 September 2013Accepted 7 December 2013Available online 15 December 2013

Keywords:A. Smart materialsB. Electrical propertiesB. Interface/interphaseE. Assembly

This study presents an effective approach to significantly improve the electrical properties of shape mem-ory polymer (SMP) nanocomposites that show Joule heating triggered shape recovery. Carbon nanofibers(CNFs) were self-assembled to form multi-layered nanopaper to enhance the bonding and shape recoverybehavior of SMP, respectively. It was found that both glass transition temperature (Tg) and electricalproperties of the SMP nanocomposites have been improved by incorporating multi-layers of self-assembled nanopapers. The electrically actuated shape recovery behavior and the temperature profileduring the actuation were monitored and characterized at a voltage of 30 V.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Shape memory effect (SME) of polymers is the ability to store aprogrammed shape indefinitely and fully recover to an originalshape in response to an environmental stimulus, such as heat,light, electricity, magnetism, and water [1,2]. Unlike shape mem-ory alloys (SMAs), the SME of shape memory polymers (SMPs) ispredominantly an entropic phenomenon [3,4]. More appropriateterms for the two structural requirements for the SME are switchsegment and netpoint [5]. This dual-state system is essential to en-abling the SME in these polymers [6,7]. For amorphous networks,shape memory behavior is realized by programming around theglass transition temperature (Tg). In a typical shape memory cycle,the material is heated above Tg, deformed to the desired shape,then cooled rapidly below the Tg to fix the programed shape forindefinite storage. The SMPs are attractive for great research inter-est and a number of potential applications due to their advantages,such as their light weight, ease in manufacturing and the ability fortailored properties to precisely meet the needs of a particularapplication [8–10]. Moreover, they can store and recover largedeformation, which is desirable for deployable and morphingstructures [11]. Despite tremendous progress in synthesis, analysis,characterization, actuation methods and modeling enables us todevelop SMPs through a knowledge based approach [12–14].

Fundamental research aims to other stimuli different than heat(e.g. light, electric current or alternating magnetic fields) and en-abling more desirable requirements on demand [15–19]. Amongthese approaches, the utilization of electrically Joule resistive heat-ing to trigger the shape recovery of SMPs is desirable for practicalapplications where it would not be possible to use heat. Some pre-vious efforts used electrically conductive SMP composites with car-bon nanotubes (CNTs) [20], short carbon fibers [21], carbon black[22], carbon fiber [23], carbon nanofibers (CNFs) [24], nanopaper[25], graphene [26], etc. The nanopaper was used with some degreeof success owing to significant improvement in electrical conduc-tivity [27]. In this study, multi-layers of the nanopaper were incor-porated into the SMP to enhance the bonding between thenanopaper and the SMP matrix. The multi-layers of the nanopaperalso improved the electrical properties to achieve a fast actuationat a low electric voltage for the SMP nanocomposites.

2. Experimental

CNFs are available with diameters ranging from 20 to 150 nmand lengths of 5–15 lm. Distilled water was used as a solvent.The raw CNFs of 0.6 g were mixed with 600 ml of distilled waterto form a suspension. The surfactant Triton X-100 (C14H22O(C2H4-

O)n) of 2 ml was introduced to aid the dispersion of the CNFs.The CNF suspension was then sonicated with an ultrasound powerlevel of 1200 W for 20 min. After which, the suspension was fil-trated under a high pressure and a nanopaper was self-assembled

192 H. Lu et al. / Composites: Part B 59 (2014) 191–195

on the hydrophilic polycarbonate membrane. The CNF nanopaperwas further dried in an oven at 120 �C for 2 h to remove the resid-ual water and surfactant.

The polymer matrix is a polyurethane-based fully formablethermosetting SMP resin. The cured resin is engineered with a Tgof 50 �C. The resin transfer molding process was used to fabricatethe SMP nanocomposites. One, two, three and four layers of thenanopaper were placed on the bottom of the metallic mold.The SMP resin was then injected into and filled into the mold.The relative pressure of the resin transfer molding was kept atapproximately 6 bars. After filling the mold, the mixture wascured at room temperature for 24 h to produce the final SMPnanocomposites.

3. Results and discussion

3.1. Morphology and structure of multi-layered SMP nanocomposite

Scanning electron microscopy (SEM) was used to study the sur-face morphology and structure of multi-layered nanopaper en-abled SMP nanocomposite. Fig. 1(a) shows the typical surfaceview of the CNF nanopaper at an accelerating voltage of10.00 keV. The CNFs have a diameter ranging from 20 to 150 nm,and are entangled with each other. No large aggregates of nanofi-bers were observed. The self-assembled and multi-layered CNFnanopaper confirms a continuously conductive network. Fig. 1(b)illustrates macroscopic structure variations of the SMP nanocom-posite with two layers of nanopaper. In this two-layered nanopa-per enabled SMP nanocomposite, there are four interfacesbetween the nanopaper and the SMP matrix. Therefore, it isexpected to improve the interface bonding of SMP nanocompositethrough the multiple layers of the nanopaper. Fig. 1(c and d) revealthe morphology and structure of the interface bonding betweenthe multi-layered nanopaper and SMP matrix. The interfacialbonding between the matrix and multi-layered nanopaper is ex-pected to improve the mechanical properties and performance ofthe polymer composites [28,29]. Furthermore, it also will facilitate

Fig. 1. The morphology and structure of multi-layered nanopaper and the SMP nanoconanopaper. (b) Structure of multi-layered nanopaper enabled SMP nanocomposite. (cnanopaper and the SMP matrix.

in transferring the Joule heat from the nanopaper to the SMPmatrix.

3.2. Electrical resistivity measurement

The electrical resistivity was determined with a four-pointprobe apparatus (QUAD PRO-SIGNATONE, computerized four pointresistivity system). This approach is an electrical impedance mea-suring technique that uses separate pairs of current-carrying andvoltage-sensing electrodes to make more accurate measurementsthan the traditional two-terminal sensing method. In order to re-duce experimental errors from many previous measurements, itis preferable that the tested sample is symmetrical. Fig. 2(a) showsa schematic illustration of the setup. The apparatus has four probeswith an adjustable inter-probe spacing. A constant current passedthrough two outer probes and an output voltage was measuredacross the inner probes with a voltmeter. The electrical resistancesof thirteen modes are used to calculate the electrical resistivity ofthe tested samples according to the van der Pauw expression. Theelectrical resistivity of the nanopapers with a different layer num-ber was plotted in Fig. 2(b). The electrical resistivities of the nano-papers were plotted against different locations. Thirteen locationswere chosen to determine the electrical resistivity for the testednanopapers. Each data point denotes the resistivity at a particularzone. In the experiments, it was found that the average values ofelectrical resistivity are 6.41 X cm, 3.22 X cm, 2.12 X cm and1.67 X cm, as the SMP nanocomposite incorporated with one-layer, two-layer, three-layer and four-layer CNF nanopaper, respec-tively, as shown in the curve of Fig. 2(c). The electrical resistivity ofSMP nanocomposite is improved with an increase in the layernumber of nanopaper. It is expected that the more CNFs in thenanopaper, the more conductive paths are formed in the continu-ous network. Given more conductive paths, more electrons are in-volved in an electrical circuit. Therefore, the amplitude of electriccurrent and current carrying capability both increase. Additionally,more conductive paths will increase the probability of forming rel-atively shorter distances for a reduced electrical resistance to theelectric current [30].

mposite at an accelerating voltage of 10.00 keV. (a) Morphology of multi-layered) and (d) Morphology and structure of the interface between the multi-layered

Fig. 2. (a) Schematic illustration of four-point probe method for measuring electrical resistivity. (b) Electrical resistivity of one-, two, three- and four-layered CNF nanopapers,each experimental data was measured thirteen times. (c) Electrical resistivity curves of the one-, two, three- and four-layered CNF nanopapers.

H. Lu et al. / Composites: Part B 59 (2014) 191–195 193

3.3. Thermomechanical property analysis

Differential scanning calorimetry (DSC) experiments were per-formed with DSC 204F1, Netzsch, Germany. All experiments wereconducted with a constant heating and cooling rate of 10 �C min�1.The samples were investigated in the temperature range from 25to 120 �C. They were heated from 25 to 150 �C, then cooled downto 25 �C. Whenever a maximum or minimum temperature in thetesting program was reached, this temperature was kept constantfor 2 min. In the polyurethane SMP, the Tg is responsible for theSME. Exceeding it activates the polymer segments and therebyallowing the material to regain its permanent (or original) form.Therefore, the Tg is a critical parameter and is necessary for char-acterizing the shape recovery performance of SMPs. The changein Tg as function of the temperature for the SMP nanocompositesincorporated with different number of layers is presented inFig. 3(a and b). Tg is determined by the heating run as 50.02,51.68, 50.69, and 52.92 �C for the SMP nanocomposite with one-,two-, three- and four-layered nanopaper, respectively. While theTg is determined by the cooling run as 60.99, 61.24, 62.23, and62.45 �C for the SMP nanocomposite with one-, two-, three- andfour-layered nanopaper, respectively. It can be seen that the glasstransition occurs within a temperature range from 50 to 65 �C. Theexperimental results could be attained from the multi-layerednanopaper that enhances the bonding strength between carbonfiber and the SMP matrix, resulting in improved thermomechanicalproperties. However, with an increase in weight content above acertain value, the thermomechanical properties of the SMP matrix

would be decreased due to an increase in separation distanceamong the macromolecules.

3.4. Electrically triggered shape recovery and temperature distribution

The tested SMP nanocomposite was incorporated with a four-layered nanopaper and originally had a flat shape. After the samplewas heated above its Tg, it could be deformed into a desirableshape upon application of an external force. Cooling back to roomtemperature, the SMP nanocomposite with a temporary (de-formed) shape was subsequently fixed. The electrical actuationwas studied on a ‘‘P’’ shaped SMP nanocomposites sample. The flatsample with a dimension of 80 � 10 � 4 mm3 was bent into a ‘‘U’’shape at 80 �C. Images were taken with a digital camera at a con-stant frame rate of 30 Hz, and with an appropriate visual rangeto detect the sample’s curvature. Snapshots of the shape recoverysequence of the SMP nanocomposite sample is shown in Fig. 4. Aconstant 30 V DC voltage was applied to the SMP nanocomposite.It took 80 s to complete the shape recovery. It showed very littlerecovery ratio during the first 10 s, but then exhibited a fasterrecovery behavior until 60 s. Finally, the SMP nanocomposite sam-ple regained its original (permanent) shape. Therefore, the functionand effectiveness of the four-layered nanopaper in the actuation ofSMP nanocomposite by resistive Joule heating were experimen-tally demonstrated and characterized.

Simultaneously, an infrared video camera (FLIR Infrared Cam-era) was used to record and monitor the shape recovery behaviorand temperature distribution. Nine snap-shots of the tested SMP

Fig. 3. The Tg determined by the DSC measurement for the SMP nanocomposites incorporated with one-, two-, three- and four-layered nanopaper, respectively.

Fig. 4. Electrically resistive heating-induced shape memory effect of the SMPnanocomposites incorporated with four-layered CNF nanopaper.

Fig. 5. Snap shots of shape recovery and temperature distribution of SMPnanocomposite incorporated with four-layered nanopaper.

Fig. 6. Temperature distribution curves of the SMP nanocomposite driven byloading an electric voltage of 30 V at various heating times.

Fig. 7. A comparison in electric power of the SMP nanocomposites incorporatedwith one-, two-, three- and four-layered nanopaper at an electric voltage of 30 V.

194 H. Lu et al. / Composites: Part B 59 (2014) 191–195

nanocomposite sample were presented in Fig. 5. Furthermore, thetemperature distribution along the tested sample (100 selectionpoints were set) was plotted in Fig. 6. High temperatures werefound where internal strains were higher than the local resistivity.With electricity being applied, the resistive Joule heating resultedin the gradually increased in temperature. At 80 s, the maximumtemperature of the sample reached approximately 102.5 �C, andthe external electricity was turned off to avoid thermal degrada-tion of the polymer. Consequently, the temperature on the testedsample lowered to room temperature. In this manner, the

temperature distribution and shape recovery process were well re-corded and monitored for the electrical actuation of the SMPnanocomposite.

To further characterize the effect of multi-layered nanopaper onthe electrical actuation of SMP nanocomposites, a comparison ofthe electric power of the SMP nanocomposites incorporated withone-, two-, three- and four-layered nanopaper was carried out to

H. Lu et al. / Composites: Part B 59 (2014) 191–195 195

identify the improvement in the performance at the electric volt-age of 30 V, as shown in Fig. 7. The experimental results indicatethat the SMP nanocomposite incorporated with a four-layerednanopaper had the highest electric power of 15 W due to its lowestelectrical resistivity, while that of the SMP nanocomposite incorpo-rated with one-, two- and three-layered nanopaper is 3.9 W, 6.9 Wand 11.7 W, respectively. Therefore, it can be seen that the multi-layered nanopaper improve the electrical actuation of the SMPs.

4. Conclusion

A series of experiments were conducted to study the multi-lay-ered nanopaper enabled SMP nanocomposites, of which the shaperecovery was achieved by the electrically resistive Joule heating.The multi-layered nanopaper was to improve the interfacial bond-ing and enhance the reliability in bonding, and helped to transferthe resistive Joule heating from the nanopaper to the SMP matrix.Both Tg and electrical property of the nanocomposite wereimproved by the multi-layered nanopaper. The electrically drivenrecovery behavior was characterized at an electric voltage of30 V. Furthermore, the temperature distribution of the SMP nan-composite incorporated with four-layered nanopaper wasrecorded and monitored in the recovery process by electricity.We demonstrated a simple way to produce the electro-activatedSMP nanocomposites by using multi-layered nanopaper in whichJoule heating was possible at a low electrical voltage.

Acknowledgments

This research was supported by the National Natural ScienceFoundation of China (NSFC) (Grant No. 51103032) and Fundamen-tal Research Funds for the Central Universities (Grant No.HIT.BRETIV.201304).

References

[1] Hu JL, Zhu Y, Huang HH, Lu J. Recent advances in shape-memory polymers:structure, mechanism, functionality, modeling and applications. Prog PolymSci 2012;37:1720–63.

[2] Zhao Q, Behl M, Lendlein A. Shape-memory polymers with multipletransitions: complex actively moving polymers. Soft Matter 2013;9:1744–55.

[3] Rousseau IA. Challenges of shape memory polymers: a review of the progresstoward overcoming SMP’s limitations. Polym Eng Sci 2008;48:2075–89.

[4] Xie T. Recent advances in polymer shape memory. Polymer2011;52:4985–5000.

[5] Huang WM, Zhao Y, Wang CC, Ding Z, Purnawali H, Tang C, et al. Thermo/chemo-responsive shape memory effect in polymers: a sketch of workingmechanisms, fundamentals and optimization. J Polym Res 2012;19:9952–86.

[6] Hu JL, Meng HP, Li GQ, Ibekwe SI. A review of stimuli-responsive polymers forsmart textile applications. Smart Mater Struct 2012;21:053001.

[7] Nguyen TD, Qi HJ, Castro F, Long KN. A thermoviscoelastic model foramorphous shape memory polymers: incorporating structural and stressrelaxation. J Mech Phys Solids 2008;56:2792–814.

[8] Liu C, Qin H, Mather PT. Review of progress in shape-memory polymers. JMater Chem 2007;17:1543–58.

[9] Xie T, Rousseau IA. Facile tailoring of thermal transition temperatures of epoxyshape memory polymers. Polymer 2009;50:1852–6.

[10] Hu J. Advances in shape memory polymers. Woodhead Publishing Limited;2013.

[11] Kuder IK, Arrieta AF, Raither WE, Ermanni P. Variable stiffness material andstructural concepts for morphing applications. Progr Aerospace Sci 2013.doi.org/10.1016/j.paerosci.2013.07.001.

[12] Mather PT, Luo XF, Rousseau IA. Shape memory polymer research. Annu RevMater Res 2009;39:445–71.

[13] Wolfgang W, Karl K, Matthias H, Lendlein A. Shape-memory polymers andshape-changing polymers. Adv Polym Sci 2010;226:97–145.

[14] Serrano MC, Ameer GA. Recent insights into the biomedical applications ofshape-memory polymers. Macromol Biosci 2012;12:1156–71.

[15] Koerner H, Price G, Pearce NA, Alexander M, Vaia RA. Remotely actuatedpolymer nanocomposites—stress-recovery of carbon-nanotube-filledthermoplastic elastomers. Nat Mater 2004;3:115–20.

[16] Huang WM, Yang B, An L, Li C, Chan YS. Water-driven programmablepolyurethane shape memory polymer: demonstration and mechanism. ApplPhys Lett 2005;86:114105.

[17] Lu HB, Liang F, Gou JH. Nanopaper enabled shape-memory nanocompositewith vertically aligned nickel nanostrand: controlled synthesis and electricalactuation. Soft Matter 2011;7:7416–23.

[18] Mohr R, Kratz K, Weigel T, Lucka-Gabor M, Moneke M, Lendlein A. Initiation ofshape-memory effect by inductive heating of magnetic nanoparticles inthermoplastic polymers. Proc Nat Acad Sci USA 2006;103:3540–5.

[19] Li G, Fei G, Xia H, Han J, Zhao Y. Spatial and temporal control of shape memorypolymers and simultaneous drug release using high intensity focusedultrasound. J Mater Chem 2012;22:7692–6.

[20] Cho JW, Kim JW, Jung YC, Goo NS. Electroactive shape-memory polyurethanecomposites incorporating carbon nanotubes. Macromol Rapid Commun2005;26:412–6.

[21] Lu HB, Huang WM. Synergistic effect of self-assembled carboxylic acid-functionalized carbon nanotubes and carbon fiber for improved electro-activated polymeric shape-memory nanocomposite. Appl Phys Lett2013;102:231910.

[22] Le HH, Zulfiqar A, Mathias U, Ilisch S, Radusch HJ. Effect of the cross-linkingprocess on the electrical resistivity and shape-memory behavior of cross-linked carbon black filled ethylene–octene copolymer. J Appl Polym Sci2010;120:2138–45.

[23] Lu HB, Huang WM, Yao YT. Review of chemo-responsive shape change/memory polymers. Pigm Resin Technol 2013;42:237–46.

[24] Tang Z, Sun D, Yang D, Guo B, Zhang L, Jia D. Vapor grown carbon nanofiberreinforced bio-based polyester for electroactive shape memory performance.Compos Sci Technol 2013;75:15–21.

[25] Lu HB, Bai PP, Yin WL, Liang F. Magnetically aligned carbon nanotubes innanopaper for electro-activated shape-memory nanocomposites. NanosciNanotechnolgy Lett 2013;5:732–6.

[26] Lu HB, Gou J. Study on 3-D high conductive graphene buckypaper for electricalactuation of shape memory polymer. Nanosci Nanotechnol Lett2012;4:1155–9.

[27] Gou J, Tang Y, Liang F, Zhao Z, Firsich D, Fielding J. Carbon nanofiber paper forlightning strike protection of composite materials. Compos B Eng2010;41:192–8.

[28] Andrews R, Weisenberger MC. Carbon nanotube polymer composites. CurrOpin Solid State Mater Sci 2004;8:31–7.

[29] Gojnya FH, Wichmanna MHG, Köpkeb U, Fiedlera B, Schultea K. Carbonnanotube-reinforced epoxy-composites: enhanced stiffness and fracturetoughness at low nanotube content. Compos Sci Technol 2004;64:2363–71.

[30] Lantada AD, Morgado PL, Sanz JLM, García JM, Munoz-Guijosa JM, Otero JE.Intelligent structures based on the improved activation of shape memorypolymers using Peltier cells. Smart Mater Struct 2010;19:055022.