Synthesis and characterization of thermotropic liquid crystallinepolyester/multi-walled carbon nanotube nanocomposites

Xiaoyu Wang, Juanjuan Wang, Wenfeng Zhao, Linwei Zhang, Xing Zhong, Rong Li *, Jiantai Ma *

College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou, 730000, People’s Republic of China

Applied Surface Science 256 (2010) 1739–1743

A R T I C L E I N F O

Article history:

Received 8 June 2009

Received in revised form 30 September 2009

Accepted 30 September 2009

Available online 6 October 2009

Keywords:

TLCP/MWNTs nanocomposites

In situ polymerization

Interactions

Thermal properties

Morphological properties

A B S T R A C T

Thermotropic liquid crystalline polyester (TLCP) was synthesized via low-temperature solution

polycondensation from 1,4-Bis(4-Hydroxybenzoyloxy)butane and terephthaloyl dichloride. Polymer

nanocomposites based on a small quantity of multi-walled carbon nanotubes (MWNTs) were prepared

by in situ polymerization method. The wide-angle X-ray diffraction (WAXD) results suggested that the

addition of MWNTs to TLCP matrix did not significantly change the crystal structure of TLCP. The

interactions between the molecules of the TLCP host phase and the carbon nanotubes were investigated

through Raman spectroscopy investigations. We detected a distinct wave number shift of the radial

breathing modes, confirming the carbon nanotubes interacted with the surrounding liquid crystal

molecules, most likely through aromatic interactions (p-stacking). The interactions between liquid

crystal host and nanotube guests were also evident from a polarizing microscopy (POM) study of the

liquid crystal–isotropic phase transition in the proximity of nanotubes. The thermal properties and the

morphological properties of the TLCP/MWNTs nanocomposites were investigated by thermogravimetric

analysis (TGA), differential scanning calorimetry (DSC) and scanning electron microscopy (SEM). TGA

data demonstrated the addition of a small amount of MWNTs into TLCP matrix could improve the

thermal stability of TLCP matrix. DSC results revealed that melt transition temperatures and isotropic

transition temperatures of the hybrids were enhanced.

� 2009 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Applied Surface Science

journa l homepage: www.e lsev ier .com/ locate /apsusc

1. Introduction

Carbon nanotubes (CNTs) [1] have attracted a great deal ofscientific interest as advanced materials for next generationbecause of its excellent electronic, mechanical, thermal, magneticand structural properties [2–11]. Fundamental research suggeststhat CNTs are regarded as promising reinforcements in thepolymer composites [12] due to the combination of their uniquelyexcellent properties with high aspect ratio [9] and the efficient loadtransfer from the host matrix to the tubes [13]. So far, much workhas been done in the field of preparing polymer/MWNTsnanocomposites [14–17].

Thermotropic liquid crystalline polyesters have been inten-sively studied because of their potential applications as highperformance materials in the last two decades [18,19]. Amongthem, the wholly aromatic thermotropic polyesters have generallyreceived a considerable interest for technological applications dueto their high use temperatures, excellent chemical resistance,relatively high glass transition temperatures, excellent processing

* Corresponding authors. Tel.: +86 931 8912311.

E-mail addresses: liyirong@lzu.edu.cn (R. Li), majiantai@lzu.edu.cn (J. Ma).

0169-4332/$ – see front matter � 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.apsusc.2009.09.105

and mechanical properties. Liquid crystals (LC) having the long-range orientational order can align carbon nanotubes using theself-organization effect of their host. The principal symmetry axisof the LC molecules (or molecule aggregates) spontaneously tendto align along a common direction defined as the director (n) whichcan be conveniently reoriented by the application of external fieldstermed the Freedericksz transition [20]. LC alignment should bepossible with the CNTs, offering a general route for controlledassembly of organized nanomaterials and devices.

In order to control and optimize the aligning effect of the liquidcrystalline matrix it is important to investigate the mechanismbehind it. On the one hand the degree of orientational order of thehost phase is obviously a most relevant parameter, a highlyordered liquid crystal phase being potentially more successfulprovided that the nanotubes are well incorporated in the phase[21–23]. On the other, the interactions across the liquid crystal–carbon nanotube interface are expected to play a crucial role. Evena highly ordered host phase will have no strong aligning effect ifthe nanotubes in it do not experience an energy penalty if theirorientation differs from the preferred orientation of the liquidcrystal [24]. The stronger the interactions between nanotubes andliquid crystal molecules, the greater this energy penalty should be.Hence, strong interactions between the guest and the host

X. Wang et al. / Applied Surface Science 256 (2010) 1739–17431740

molecules should ensure an efficient transfer of order from the LCto the nanotubes. TLCP is a typical aromatic polymer and containsbenzene ring on its backbone. Therefore, a strong interactionbetween TLCP and MWNTs is expected due to p–p stacking, whichprovides feasibility to explore TLCP/MWNTs composites.

Previous work in our laboratory [25] has demonstrated thefeasibility of adding MWNTs to monotropic nematic LC (MPPB) toinvestigate the interactions between MWNTs and host matrix andstudy the electrical conductivity and thermal stability of thenanocomposite. But, no improvement in thermal stability was seenat the initial state of degradation. In this paper, we prepared TLCP/MWNTs nanocomposites by in situ polymerization method andfurther investigated the interactions between MWNTs and hostmatrix. Especially, we detected the thermal stability of nanocom-posites as a function of the MWNTs content in the matrix polymer.We examined that adding carbon nanotubes into TLCP had effecton the properties of the TLCP, such as melt transition temperature(Tm), isotropic transition temperatures (Ti) and thermal stability.We expected that this study will be helpful in providing theunderstanding on the physical properties of the TLCP nanocom-posites reinforced with a small quantity of MWNTs.

2. Experimental

2.1. Materials

MWNTs (diameters: 20–40 nm, purity: 95–98%) prepared bythe catalytic decomposition of CH4 were provided by ShengzhenNanotech Port Ltd. Co. (China). In a typical experiment, MWNTswere added into concentrated HNO3 and this mixture was refluxedat 85 8C for 8 h. Then the mixture was diluted with excessdeionized water and then vacuum-filtered through millipore PTFEmembranes until the pH of the filtrate reached approximately 7.The filtrate solid was dried in vacuo at 100 8C for 24 h.

1,4-Bis(4-Hydroxybenzoyloxy)butane and terephthaloyldichloride were used as received. Other reagents includingtriethylamine, 1,2-dichloroethane and acetone were AR gradeand were used as supplied. The TLCP and TLCP/MWNTs nano-composite were synthesized as follows.

2.2. Polymer preparation [26–28]

The TLCP was synthesized as follows, a solution of 1,4-Bis(4-Hydroxybenzoyloxy)butane 1.65 g (0.005 mol) and triethylamine(0.697 g, 0.012 mmol) in 1,2-dichloroethane (30 mL) was cooled to0 8C. Then, terephthaloyl dichloride 1.02 g (0.005 mol) was addedin one portion. The mixture was stirred at 0 8C for 1 h and at roomtemperature for 4 h under a nitrogen atmosphere, and then pouredinto acetone to precipitate the polymer. The white, powderypolymer was filtered off, washed carefully with water and dried at80 8C in a vacuum oven. The yield was 94.0%. FT-IR (KBr): 2957(CH2), 1716 (C55O), 1603 and 1502 (aromatic), 1266–1113 (C–O) cm�1. Elemental microanalysis (content %): N (0.00), C (66.74),H (4.57). Tm: 279 8C, Ti: 408 8C. The structure of thermotropic liquidcrystalline polyester was showed in Scheme 1.

2.3. Preparation of TLCP/MWNTs nanocomposites

The TLCP/MWNTs nanocomposites were synthesized by thefollowing steps, the MWNTs were added into the solution of 1,4-

Scheme 1. Structure of thermotropic liquid crystalline polyester (TLCP).

Bis(4-Hydroxybenzoyloxy)butane and triethylamine in 1,2-dichlor-oethane, followed by supersonic treatment for 2 h. The followingprocedure was the same as above by in situ polymerization method.In order to study whether the different content of MWNTs haveeffect on the properties of TLCP/MWNTs nanocomposites, weprepared a series of nanocomposites with different content. Forsimplicity, the composites were referred to as 0, 0.01, 0.1, 1% TLCP/MWNTs and so on, in which TLCP and MWNTs represented thecomponents used to prepare the composites and the numberdenoted the MWNTs weight percent in the composites.

2.4. Characterization

The Fourier transform infrared spectrum (FT-IR) of TLCP wasrecorded from 400 to 4000 cm�1 to identify the structure of thesamples using a Nicolet Nexus 670 FT-IR from the Thermo NicoletInc, USA. Wide-angle X-ray diffraction measurements wereperformed at room temperature on a Rigaku (D/Max-IIIB) X-rayDiffractometer using Ni-filter Cu-Ka radiation. The scanning was 88/min over a range of 2–608 at room temperature. The scanning typewas continuous scan. Raman scattering spectrum was measured atroom temperature using an excitation wavelength of 532 nm (JY-HR800) under the backscattering geometry. A polarized opticalmicroscope (ECLIPSE 80i, Nikon) equipped with a THMSE 600(LINKAM) hot stage was used to examine the liquid crystallinebehavior, at a heating rate of 10 8C/min. The morphology of thesurfaces of the samples was investigated using a JSM-6701 (Japan)Scanning electron microscopy (SEM). The thermogravimetricanalysis (TGA) and differential scanning calorimetry (DSC) ofTLCP/MWNTs was carried out under N2 atmosphere on NetzschSTA 449C equipment. The samples were heated at 10 8C/min.

3. Results and discussion

The WAXD measurement plots are shown in Fig. 1 and someparameters are summarized in Table 1. It can be seen that bothTLCP and TLCP/MWNTs nanocomposites had one group of lowintensity diffraction peaks located at the range of 15–288. Theseindicated that there were some crystal parts in these three kinds ofproducts except the amorphous parts. Besides, the position ofdiffraction peaks and interplanar spacing (d) values of TLCP andTLCP/MWNTs nanocomposites were similar, illustrating that thethree kinds of products attribute to one crystal systems. The d

values of TLCP/MWNTs nanocomposites were bigger than thesevalues of TLCP due to the crystal degree decrease little because of

Fig. 1. WAXD of pure TLCP and TLCP/MWNTs nanocomposites.

Table 1WAXD results of pure TLCP and TLCP/MWNTs nanocomposites.

Sample (wt.%) Peak A Peak B Peak C Peak D Peak E

2u (8)d (A)

2u (8)d (A)

2u (8)d (A)

2u (8)d (A)

2u (8)d (A)

0 15.98

5.54

19.32

4.59

22.68

3.92

23.28

3.82

27.12

3.29

5 15.92

5.56

19.22

4.61

21.74

4.08

22.92

3.88

27.10

3.29

10 15.92

5.56

19.28

4.60

21.60

4.11

23.12

3.84

27.05

3.30

Fig. 2. Raman spectra of MWNTs and TLCP/MWNTs nanocomposites.

X. Wang et al. / Applied Surface Science 256 (2010) 1739–1743 1741

the introduction of MWNTs. These results suggested that theaddition of MWNTs to TLCP did not significantly change the crystalstructure of TLCP.

Fig. 2 displays the Raman spectra of MWNTs and TLCP/MWNTsnanocomposites. Two characteristic peaks observed at approxi-mately 1375 and 1600 cm�1 from MWNTs curve can be termed asD-band and G-band, respectively. The G-band was related to thestructural intensity of the sp2-hybridized carbon atoms of the CNT.The D-band reflected the disorder-induced carbon atoms, resultingfrom the defects in the nanotube and their ends [29]. It can be seenthat the roman shift of G-band of 5% TLCP/MWNTs nanocompo-sites was clearly increased and the roman shift of D-band was

Fig. 3. Polarizing optical microscopy images of pure TLCP and 1 wt.% TLCP/MWNTs. (a) P

phase, and (c) the transition of the 1 wt.% TLCP/MWNTs from isotropic to the nematic

slightly increased as compared to pure MWNTs. From the figure wecan see that even though the relative intensity of G-band and D-band of 1% TLCP/MWNTs has reduced, the roman shifts of themwere still increased. Besides, the roman shifts of D-band and G-band of 1% TLCP/MWNTs was bigger than the 5%TLCP/MWNTs’s,which proved the interactions between TLCP and MWNTsdecreased when employing high MWNTs loadings. It should bepointed out that the nanotubes embedded in a different liquidcrystal host, formed by molecules having a similar molecularstructure but without phenyl rings, did not show any shift in theradial breathing mode band. The clear effect with TLCP can beassociated with p-stacking interactions at the carbon nanotubesurface, involving the aromatic core structure of the TLCPmolecules, affecting the vibrational modes of carbon nanotubes.In particular, the radial breathing mode band, related to vibrationsperpendicular to the long axis of the tube, can be affected byforeign molecules adsorbing onto the nanotube surface. This bandis thus a sensitive probe for studying the interactions betweennanotubes and the TLCP host molecules. Recently the existence ofstrong binding interaction between CNT and LC molecules havebeen carefully studied by Lebovka et al. using Fourier transforminfrared spectroscopy [30].

In order to further corroborate this interpretation we havetaken help of the long-range intermolecular interactions of liquidcrystals to make the interaction with the MWNTs macroscopicallyobservable. Fig. 3 shows the representative POM micrographs ofTLCP and TLCP/MWNTs nanocomposites. The TLCP has a transitiontemperature from the nematic liquid crystal phase to the ordinaryisotropic liquid phase at 279 8C. It was seen from Fig. 3(a) that thenematic phase of pure TLCP exhibited a characteristic brighttexture due to its birefringence. When the phase became isotropic,losing its order and consequently the birefringence, the materialappeared uniformly black independent of sample rotation. Wehave performed a careful temperature scan, studying the behaviorin the close vicinity of the phase transition. We observed that thebirefringent texture was still stable in Fig. 3(b), indicating thepersistence of the nematic phase of the nanocomposites. A furtherincrease in temperature took the sample into the isotropic phase, ifwe decreased the temperature, we noted that the first signs ofbirefringence appeared around the nanotube bundles, see Fig. 3(c).In other words, the CNT bundles act as nucleation centres for liquidcrystalline organization. These observations confirm that the TLCPmolecules interact with the MWNT surface, leading to an increaseddegree of order in the vicinity of MWNTs. Since the higher degree oforder of the molecules at the CNT surface is transferred to

ure TLCP at nematic phase, (b) and (d) 1 wt.% and 5 wt.% TLCP/MWNTs at nematic

phase.

Fig. 5. TGA thermograms of pure TLCP and TLCP/MWNTs nanocomposites.

X. Wang et al. / Applied Surface Science 256 (2010) 1739–17431742

neighbour molecules, a region with liquid crystalline order existsaround the bundles even at temperatures where TLCP normally isisotropic, rendering the surface interactions macroscopicallyobservable. It was seen that Fig. 3(a) is brighter than Fig. 3(b)and (d). The TLCP/MWNTs nanocomposite had low birefringence.The reason may be that the sample was too black to enablesufficient light transmission through it. It was also probably relatedto the polarizability of the LC molecule influenced by MWNTs,which brought about the different refractive index from the LCmedium [31]. Besides, the nematic phase of TLCP/MWNTscomposites was still observed when MWNTs content was up to5%, but it was not obvious (Fig. 3(d)). This may be due to CNTssegregation and reagglomeration when melting the composite, orto poor MWNTs dispersion in the monomer when employing highMWNTs loadings [32].

The representative SEM images of the compact powder surfacesof nanocomposites are shown in Fig. 4. It was seen that nanotubeswere aligned along the LC director and formed extended rod TLCP/

Fig. 4. (a) SEM of 1 wt.% TLCP/MWNTs, (b) SEM of 5 wt.% TLCP/MWNTs and (c) SEM

of 10 wt.% TLCP/MWNTs.

MWNTs nanocomposites. The uniformly dispersed and nettedstructures in these images were attributed to the MWNTs. It wasapparent from Fig. 4(b) that there were more interstices and rods inthe composites with higher MWNTs loading. There was no nakedCNT on the surface of the nanocomposites, demonstrating that theMWNTs were well dispersed in the TLCP matrix. But, it was morepronounced in Fig. 4(c) of 10 wt.% TLCP/MWNTs, except intersticesand rods, there exist apparent massive phase. This may be due topoor MWNTs dispersion in the TLCP matrix when employing highMWNTs loadings. It was known that the pure MWNTs exhibitedhighly curved and random coiled features, which may beattributed to hydrogen bonding and van der Waals attractiveinteractions between carbon nanotubes [33]. The aromatic ringstructure of TLCP interacted strongly with the graphene sheet ofthe nanotube surface through intermolecular overlap of p-orbitals(p-stacking). So, the better dispersed of MWNTs in the TLCPmatrix, the stronger of the p-stacking interactions they have.

A comparative thermogravimetric analysis of pure TLCP andnanocomposites with 0.1–10 wt.% MWNTs were represented inFig. 5. It can be seen that the onset temperature of degradation(Tonset) and the residual weight were influenced by MWNTsloading into TICP. As shown in Fig. 5, the Tonset of pure TCLP was378.5 8C, when the content of MWNTs increased to 0.1% and 1%,the Tonset increased to 379.7 and 381.4 8C, respectively. Thedegradation of pure TLCP was faster than its nanocomposites,indicating that the presence of MWNTs in TLCP promoted thethermal stability of the materials. Besides, the residual weight ofTLCP/MWNTs nanocomposites was larger than that of TLCP.However, when the content of MWNTs increased to 10%, theinitial degradation temperature decreased sharply. This may bedue to poor MWNTs dispersion in the TLCP matrix whenemploying high MWNTs loadings. The interactions betweenTLCP and MWNTs did not strong enough to induce MWNTs actingas effective physical barriers against the thermal decompositionin the TLCP nanocomposites. The Raman spectra, POM and SEMpictures above can also prove this point. From these experimentresults, it can be proved that the TLCP nanocomposites reinforcedwith a very small quantity of MWNTs performed better thermalstability, lower degradation rate and more residual weightcompared with pure TLCP.

The DSC traces of TLCP/MWNTs nanocomposites shows similartendency in Fig. 6 and some parameters are summarized in Table 2.The endothermic peak which was corresponding to the melttransition temperature (Tm) of the pure TLCP appeared at 279 8C.Maximum Tm of the TLCP/MWNTs nanocomposites containingdifferent MWNTs contents in the DSC thermograms slightlyincreased to 281 8C. The isotropic transition temperatures (Ti) ofthe TLCP/MWNTs nanocomposites also mildly increased from 408

Fig. 6. DSC thermograms of pure TLCP and TLCP/MWNTs nanocomposites.

Table 2DSC results of pure TLCP and TLCP/MWNTs nanocomposites.

Sample (wt.%) Tm (8C) Ti (8C) DHm (J/g) DHi (J/g)

0 279 408 26.4 293.9

0.1 280 409 19.3 140.9

1 280 411 18.0 124.3

10 281 414 15.6 60.9

X. Wang et al. / Applied Surface Science 256 (2010) 1739–1743 1743

to 414 8C, compared with the pure TLCP. The melting andisotropization associated enthalpies (DHm, DHi) accordinglydecreased.

4. Conclusion

In summary, TLCP nanocomposites reinforced with a smallquantity of MWNTs were prepared by in-situ polymerizationmethod. The TLCP director field was influenced by the MWNTaggregates which probably acted as the nucleation centres. TheMWNTs were well dispersed in host matrix due to the aromaticinteraction. The sample presented a highly performance in thermal

stability, expressed as the MWNTs acting as effective physicalbarriers against the thermal decomposition in the TLCP nano-composites. Their thermal effect combining good interfacialinteraction and uniform dispersion was more effective at lowerMWNTs content than at higher MWNTs content. This workexpands the research field on LC nanocomposites and can beconsidered as the complementarity of theoretical efforts.

References

[1] S. Iijima, Nature 354 (1991) 56.[2] S.K. Srivastava, V.D. Vankar, V. Kumar, V.N. Singh, Nanoscale Res. Lett. 3 (2008)

205.[3] A. Leela Mohana Reddy, S. Ramaprabhu, Nanoscale Res. Lett. 3 (2008) 76.[4] J. Sandler, M. Shaffer, T. Prasse, W. Bauhofer, K. Schulte, A.H. Windle, Polymer 40

(1999) 5967.[5] H. Geng, R. Rosen, B. Zheng, H. Shimoda, L. Fleming, J.Z. Liu, Adv. Mater. 14 (2002)

1387.[6] H. Dai, Acc. Chem. Res. 35 (2002) 1035.[7] S. Subramoney, Adv. Mater. 10 (1998) 1157.[8] D. Tasis, N. Tagmatarchis, A. Bianco, M. Prato, Chem. Rev. 106 (2006) 1105.[9] E.W. Wong, P.E. Sheehan, C.M. Lieber, Science 277 (1997) 1971.

[10] L.S. Schadler, S.C. Giannaris, P.M. Ajayan, Appl. Phys. Lett. 73 (1998) 3842.[11] S.A. Manafi, M.H. Amin, M.R. Rahimipour, E. Salahi, A. Kazemzadeh, Nanoscale

Res. Lett. 4 (2009) 296.[12] P.D. Calvert, Nature 399 (1999) 210.[13] R. Andrews, M.C. Weisenberger, Curr. Opin. Solid State Mater. Sci. 8 (2004) 31.[14] B. Safadi, R. Andrews, E.A. Grulke, J. Appl. Polym. Sci. 84 (2002) 2660.[15] E.C. Chen, T.M. Wu, Polym. Degrad. Stab. 92 (2007) 1009.[16] J.Y. Kim, S.I. Han, S.H. Kim, Polym. Eng. Sci. 47 (2007) 1715.[17] J.Y. Kim, H.S. Park, S.H. Kim, Polymer 47 (2006) 1379.[18] G. Kiss, Polym. Eng. Sci. 27 (1987) 410.[19] D. Dutta, H. Fruitwala, A. Kohli, R.A. Weiss, Polym. Eng. Sci. 30 (1990) 1005.[20] V. Freedericksz, V. Tsvetkov, Phys. Z 6 (1934) 490.[21] D. Michael, L. Lynch, David, Patrick, Nano Lett. 2 (2002) 1197.[22] I. Dierking, G. Scalia, P. Morales, D. LeClere, Adv. Mater. 16 (2004) 865.[23] I. Dierking, G. Scalia, P. Morales, J. Appl. Phys. 97 (2005) 044309.[24] G. Scalia, J.P.F. Lagerwall, M. Haluska, U. Dettlaff-Weglikowska, F. Giesselmann, S.

Roth, Phys. Stat. Sol. (b) 243 (2006) 3238.[25] W.F. Zhao, J.J. Wang, J.G. He, L.W. Zhang, X.Y. Wang, R. Li, Appl. Surf. Sci. 255

(2009) 6589.[26] M. Bagheri, R.Z. Rad, React. Funct. Polym. 68 (2008) 613.[27] P.W. Morgan, J. Polym. Sci. A 2 (1964) 437.[28] D.J. Liaw, J. Appl. Polym. Sci. 65 (1997) 59.[29] S.L. Ruan, P. Gao, X.G. Yang, T.X. Yu, Polymer 44 (2003) 5643.[30] N. Lebovka, T. Dadakova, L. Lysetskiy, O. Melezhyk, G. Puchkovska, T. Gavrilko, J.

Baran, M. Drozd, J. Mol. Struct. 887 (2008) 135.[31] H. Nomura, S. Ando, T. Matsuoka, S. Koda, J. Mol. Liq. 110 (2004) 57.[32] E. Lafuente, M. Pinol, L. Oriol, E. Munoz, A.M. Benito, W.K. Maser, A.B. Dalton, J.L.

Serrano, M.T. Martınez, Mater. Sci. Eng. C 26 (2006) 1198.[33] S.N. Li, Z.M. Li, M.B. Yang, Z.Q. Hu, X.B. Xu, R. Huang, Mater. Lett. 58 (2004)

3967.