Solid State Communications 152 (2012) 2092–2095

Contents lists available at SciVerse ScienceDirect

Solid State Communications

0038-10

http://d

n Corr

E-m

journal homepage: www.elsevier.com/locate/ssc

Carbon nanoscrolls by pyrolysis of a polymer

Prasad Yadav a, Sambhaji Warule b, Jyoti Jog a, Satishchandra Ogale a,n

a National Chemical Laboratory, Dr Homi Bhabha Road, Pashan, Pune 411008, Indiab Center for Material for Electronics Technology, Department of Information Technology, Govt. Of India, Panchwati, Off Pashan Road, Pune 411008, India

a r t i c l e i n f o

Article history:

Received 3 May 2012

Received in revised form

28 August 2012

Accepted 31 August 2012

by Z. Tangcarbon nanoscroll was found to be 0.34 nm and the carbon nanoscrolls exhibited a surface area of

2

Available online 19 September 2012

Keywords:

A. Carbon nanoscrolls

B. Pyrolysis

C. Polymer

D. Supercapacitor

98/$ - see front matter & 2012 Elsevier Ltd. A

x.doi.org/10.1016/j.ssc.2012.08.029

esponding author. Tel.: þ91 20 2590 2260; fa

ail address: sb.ogale@ncl.res.in (S. Ogale).

a b s t r a c t

3D network of carbon nanoscrolls was synthesized starting from pyrolysis of poly(acrylic acid-co-

maleic acid) sodium salt. It is a catalyst-free process where pyrolysis of polymer leads to formation

of carbon form and sodium carbonate. Upon water soaking of pyrolysis product, the carbon form

undergoes self-assembly to form carbon nanoscrolls. The interlayer distance between the walls of

188 m /g as measured by the BET method.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Over the years several physical and chemical routes havebeen developed to synthesize technologically important formsof carbon [1–3]. Since last two decades and with the availabilityof high resolution microscopy techniques such as TransmissionElectron Microscopy and Scanning Tunnelling Microscopy, theresearch in carbon nanomaterials is progressed. There are severalcarbon nanoforms like carbon nanotubes, Fullerenes, graphene,carbon nanofibers which find interesting properties with diverseset of applications [4–7]. Among these nanoforms, carbon nanosc-rolls have received much attention during the last few years.Carbon nanoscrolls are open ended rolled sheets of graphenewhere the interlayer distance between adjacent rolled sheetscorresponds to graphitic interlayer spacing. Bacon in 1960 firstreported the presence of carbon nanoscrolls as scroll whiskers [8].In 2003 Viculis et al. showed that exfoliation of graphite producesa carbon sheet structure which on sonication leads to carbonnanoscrolls [9]. Similarly Prokofiev et al. showed that when theoxidation product of graphite was sonicated, it leads to formationof carbon nanoscrolls [10]. Recently Chen et al. showed thatcarbon nanoscrolls can be formed by scrolling of functionalizedgraphene oxide single sheets by Langmuir–Blodgett (LB) approach[11]. Properties of carbon nanoscrolls are very interesting fromapplication point of view. Due to rolled sheet structure, carbonnanoscrolls can easily undergo volume expansion. Also various

ll rights reserved.

x: þ91 20 2590 2636.

simulation studies of carbon nanoscrolls show that the carbonnanoscrolls should possess high surface area and therefore usefulfor hydrogen storage. [12,13]. These features are potentiallyimportant for a rich variety of applications, especially in the fieldof energy storage as cathode electrode in super-capacitors and orbatteries. There are a number of methods to synthesize thesepotentially important nanoforms of carbon but the thermaldecomposition of polymers (without or with a catalyst) repre-sents an interesting synthesis strategy which can yield carbonfibres, carbon nanotubes, graphene and amorphous carbon[14–22]. Although there are many reports on the synthesis ofcarbon fibres and amorphous carbon by decomposition of poly-mers, there are relatively limited efforts towards catalyst-freethermal decomposition of suitably selected polymers for thesynthesis of novel carbon forms [17]. Interestingly, the propertiesof the carbon forms depend on the starting precursors and thesynthesis protocol. Rayon and polyacrylonitrile (PAN) are used asprecursors for most of the commercial carbon fibres, although useof other precursors such as pitch, phenolic resins, and poly(viny-lidene fluoride) (PVDF) poly(styrene sulphonate-co-maleic acid)has also been reported [15,23–26]. Other than this various carbonnanoforms like carbon nanotube, graphene can be synthesized bypyrolysis of polymers [16–19]. As polymers are the richest sourceof carbon, thermal decomposition of polymer is also an attractivestrategy to synthesize novel carbon forms on the large scale.

In the light of these findings we present catalyst-free synthesisof carbon nanoscrolls by self-assembly of pyrolysed carbonform in aqueous medium. The pyrolysed carbon part of decom-position products which obtained from pyrolysis of poly(acrylicacid-co-maleic acid) sodium salt undergo self-assembly upon

P. Yadav et al. / Solid State Communications 152 (2012) 2092–2095 2093

its aqueous soaking. The structure and properties of these carbonnanoscrolls are determined using a variety of techniques. Thepossible mechanism of formation of carbon nanoscrolls is dis-cussed as well as their application in supercapacitor is explored.

2. Experimental details

The poly(acrylic acid-co-maleic acid) sodium salt was obtainedfrom Sigma Aldrich (Mol wt 50,000). For pyrolytic decomposition,the polymer was heat treated in a furnace at 500 1C for 1 h in airatmosphere. The choice of the pyrolysis temperature was made byanalyzing thermogravimetric data of poly(acrylic acid-co-maleicacid) sodium salt (see the Supporting information SI-I). Thepyrolysed sample was then characterized using X-ray diffraction(XRD). It was confirmed that the product consists of carbon formand sodium carbonate (see Supporting information SI-II). Theproduct formed by pyrolysis of polymer was then soaked in waterto remove the sodium carbonate and was dried subsequently. In atypical washing process the composite (about 2 g) was added to100 ml D.I. water, mixed by shaking thoroughly and then thesolution was kept undisturbed for some time. The carbon masssettled down after some time and a clear solution of watersaturated with sodium carbonate was removed by a pipette. Thecarbon mass was again dispersed in fresh D.I. water. This processwas repeated till the pH of carbon mass dispersed in water becameneutral. This carbon mass was then dried. The dried carbon residuewas then examined for the structural, chemical and morphologicalproperties using various techniques such as X-ray diffraction (XRD,Philips X’Pert PRO), Raman spectroscopy (Confocal micro-Ramanspectrometer LabRAM ARAMIS Horiba Jobin-Yvon apparatus withlaser excitation wavelength of 532 nm), high resolution transmis-sion electron microscopy (HRTEM, FEI Tecnai 300) and fieldemission scanning electron microscopy (FESEM, Hitachi S-4200).The surface area and porosity study were performed by thestandard nitrogen adsorption (the BET method) isotherm at 77 K.

Fig. 1. (A) SEM, (B) TEM, (C) and (D) High reso

Electrochemical measurements were performed with AutolabPGSTAT Potentiostat. Specific capacitance of carbon nanoscrollswas determined by cyclic voltammetry (CV) measurements in0.5 M H2SO4. The scan rate for CV measurement was kept 50 mV/s.

3. Results and discussion

Fig. 1 shows the electron microscope images of the dried carbonresidue. Fig. 1A shows the high resolution scanning electronmicroscopy (SEM) image, while Fig. 1B shows the transmissionelectron microscopy (TEM) image. Both these images reveal a 3Dinterlaced random wire-like network configuration. Fig. 1C showsthe high resolution TEM image of a single tubule. The averagelength of the same is several microns while the diameter is around15–20 nm. The average d-spacing between the two neighbouringcarbon walls is around 0.34 nm (line plot of d-spacing shown asinset) [27]. The TEM of the terminal end of one isolated tubule isalso shown in the Fig. 1D, which resembles a rolled sheet ofgraphene, termed as a nanoscroll [9,27,28].

The possible mechanism for formation of carbon nanoscrollscan be given as below. On pyrolysis of polymer there is formationof carbon sheet with sodium carbonate. Formation of carbonsheet on pyrolysis of polymer can be considered as a bottom-upprocess wherein the 1D polymer chain is converted to 2D carbonsheet. We suggest a possible mechanism with reference to Fig. 2.When the poly(acrylic acid-co-maleic acid) sodium salt is heated,adjacent polymer chains undergo a condensation reaction withloss of sodium carbonate at every linkage (Fig. 2A). There is anadditional loss of a CO molecule leading to the formation of 2Dstable six-membered ring structure. Due to the high temperaturetreatment, the carbon six membered ring structures undergoaromatization. When the condensation followed by cyclizationreaction occurs only between the maleate units on the adjacentpolymer chains, aromatic hexagonal ring systems are formedleading to the formation of an extended graphitic network.

lution TEM images of carbon nanoscrolls.

Fig. 2. Schematic of the suggested mechanism for the formation of graphitic

carbon sheets during pyrolysis of poly(acrylic acid-co-maleic-acid) sodium salt.

Fig. 3. (A) XRD (B) Raman (C) FTIR and (D) UV–vis–NIR diffuse reflectance

spectrum of carbon nanoscrolls curve(a) MWCNT curve( b), chemically converted

graphene ‘curve ( c)’ and GO. ‘curve (d)’ Inset of Fig. 4D shows the photolumines-

cence spectrum of carbon nanoscroll.

Fig. 4. Cyclic voltammetric behaviour of as synthesized carbon nanoscroll (Curve

A) and argon annealed carbon nanoscroll (Curve B).

P. Yadav et al. / Solid State Communications 152 (2012) 2092–20952094

However there are several possibilities of similar cyclizationbetween the maleate units and acrylate units of different polymerchains resulting in the formation of pentagonal and heptagonaldefects. The overall structure thus formed by this process is agraphene-like sheet with defects. The by-products of the pyrolysisprocess are CO and Na2CO3. Since the pentagonal/heptagonaldefects exert a strain in the graphene-like sheet of carbonobtained by this process, there is a strong tendency of this sheetto roll around like a carbon nanotube. But the sodium carbonateformed in the process of pyrolysis, obstruct the rolling ofnanosheets into carbon nanoscrolls. Once the pyrolized productis soaked in D.I. water, the sodium carbonate dissolves in waterand these as formed carbon sheets favour the formation of carbonnanoscrolls as shown in Fig. 2B. In this context it is important tomention that Chen et al. also discussed the scrolling of grapheneoxide (GO) sheets on sonication in aqueous medium to carbonnanoscrolls [11]. In our case during the process of water soakingthe sodium carbonate dissolves in water, and rolling ofnanosheets lead to formation of 3D network of carbon nanosc-rolls. This is due to favourable van der Walls interaction withinthe carbon sheet which tends to scroll around itself in order tohave minimum energy state.

Fig. 3A shows the XRD spectrum of the as synthesized carbonnanoscrolls. The broad peak at around 261 is a characteristic ofgraphitic carbon which has a d-spacing of �0.34 nm. This value ofinterlayer distance consistent with that obtained from the HRTEMimage (Fig. 1c inset). This broad peak at 2y¼261 is absent in theas-pyrolysed polymer (see supporting information SI-II). Thissuggests that the rolling of carbon sheet takes place duringaqueous soaking of pyrolysed polymer and it leads to formationof carbon nanoscrolls.

Fig. 3B shows the Raman spectrum of as synthesized carbonnanoscrolls. The 1590 cm�1 signature peak corresponds to the Gband. Generally this peak is present at 1580 cm�1 in the case ofgraphite, CNTs and defect-free single layer graphene and carbonnanoscrolls [29–32]. As the extent of oxidation in graphiticsystem increases this peak shifts to higher wave numbers[33,34]. In our case, oxygen functional groups which arises duringpyrolysis process increase the defect sites on carbon nanoscrollssurface, thereby shifting the G band peak from 1580 to1590 cm�1. The D band at about 1320 cm�1 is also present incarbon nanoscrolls, which corresponds to disordered nature(defects) of the graphite like system. This peak is usually presentin MWCNTs, carbon nanoscrolls and is absent in defect-freegraphite and defect-free single layer graphene [29–34]. In carbonnanoscrolls or carbon nanotubes the D band in Raman spectrumarises due to defect sites which are inherent due to their rolledstructure. Fig. 3C shows the FTIR spectrum for the as synthesizedcarbon nanoscrolls. The characteristic band at 1600 cm�1 incarbon nanoscrolls is due to CQO asymmetric stretching ofcarboxylate anion that is reconfirmed by another peak at

1373 cm�1, which is related to the symmetric stretching modeof the carboxylate groups [35]. The broad peak at around3200 cm�1 is a characteristic signature of the hydroxyl group.Also it is interesting to compare the UV–vis–NIR diffuse reflec-tance spectra of carbon nanoscrolls with MWCNT, GO, CCG, as inFig. 4D. Graphene, MWCNT etc. do not possess a band gap; on theother hand GO possesses band gap which depends on the extentof oxidation [36]. Since the pyrolysis process is carried out in air,the resulting carbon nanoscrolls have oxygen functional groups. Itis clear from the optical data shown in Fig. 3D, that the trend forour carbon nanoscroll sample is intermediate between that forCCG/MWCNT and GO (increasing nature towards longer wave-lengths). Clearly there is some contribution from oxidized carbonin our nanoscrolls. The Inset of Fig. 2D shows the PL spectrum ofthe carbon nanoscrolls sample. The excitation wavelength was320 nm. The PL emission occurs at 360 nm. Carbon nanoscrollshave carboxyl functional groups which increase the defect siteson carbon nanoscrolls. These trap states give the photolumines-cence near 360 nm [37].

Table1BET surface area and porosity data for the carbon

nanoscrolls.

Carbon nanoscrolls

BET surface area (m2/g) 188.4

Pore volume (cm3/g) 0.266

Pore Radius Dv(r) (nm) 1.48

P. Yadav et al. / Solid State Communications 152 (2012) 2092–2095 2095

Table 1 shows the comparison of Brunauer Emmett Teller(BET) surface area and porosity data for the carbon nanoscrolls.The surface area of the carbon nanoscrolls is �188 m2/g. Due tohigh surface area of carbon nanoscrolls, these were exploited forthe supercapacitor application.

Fig. 4 ‘Curve a’ shows the CV of carbon nanoscrolls. The CVprofiles exhibit a nearly symmetrical or slightly skewed rectan-gular shape. The value of specific capacitance was calculated byintegrating area of CV curves. The specific capacitance value forcarbon nanoscrolls is around 2 F/g. The low value of specificcapacitance is due to defect sites present in carbon nanoscrollswhich limits the electron transport properties of carbon nanosc-rolls. This is due to air pyrolysis of polymer which is one of thesteps in the formation of carbon nanoscrolls. These as synthesizedcarbon nanoscrolls were further annealed in argon atmosphere at1000 1C for 2 h. Curve b in Fig. 3 shows the CV of argon annealedcarbon nanoscroll which shows specific capacitance value ofaround 30 F/g. The specific capacitance value of argon annealedcarbon nanoscrolls is comparable to that of carbon nanotubeswhich is reported to be about 40–50 F/g [38,39].

Before we conclude it is useful to point out that the uniquemorphology of carbon nanoscrolls may be amenable to interest-ing magnetic effects. Hence, we carried out Electron Spin Reso-nance (ESR) and SQUID measurements on the sample. Thecorresponding results are presented in supplementary informa-tion SI-IV. The ESR spectrum of carbon nanoscrolls suggestspresence of strong paramagnetic centres with g value �2.0015.There are reports of paramagnetic centre in graphene, carbonnanotubes etc. with g value of 2.00–2.005 [40-42]. Also thepresence of single ESR peak suggests no interaction of hydrogen(proton) with the unpaired electrons. The SQUID magnetizationdata suggests a weak coupling of paramagnetic centres to givemild ferromagnetism. We will pursue the magnetism aspects in asubsequent paper.

In summary we have demonstrated the catalyst-free efficientroute for the synthesis of carbon nanoscrolls starting frompyrolysis of polymer. Pyrolysis of poly(acrylic acid-co-maleicacid) sodium salt leads to formation of carbon form and sodiumcarbonate. The carbon part of decomposition products undergoself-assembly upon its aqueous soaking to form carbon nanosc-rolls. We find that the as synthesized carbon nanoscrolls haveordered structure with interlayer distance of 0.34 nm.

Acknowledgements

One of us (PAY) acknowledges Fellowship support from theCouncil of Scientific and Industrial Research (CSIR), Govt. of India.

SBO acknowledges funding support by the Department of Scienceand Technology (DST) and Department of Information Technology(DIT), Govt. of India.

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.ssc.2012.08.029.

References

[1] T. Ebbesen, P. Ajayan, Nature 358 (1992) 220.[2] T. Guo, P. Nikolaev, A. Thess, D. Colbert, R. Smalley, Chem. Phys. Lett. 243

(1995) 49.[3] A. Eftekhari, P. Jafarkhani, F. Moztarzadeh, Carbon 44 (2006) 1343.[4] H. Kroto, J. Heath, S. O’brien, R. Curl, R. Smalley, Nature 318 (1985) 162.[5] S. Iijima, Nature 354 (1991) 56.[6] G. Tibbetts, J. Cryst. Growth 73 (1985) 431.[7] A. Geim, K. Novoselov, Nat. Mater. 6 (2007) 183.[8] R. Bacon, J. Appl. Phys. 31 (1960) 283.[9] L. Viculis, J. Mack, R. Kaner, Science 299 (2003) 1361.

[10] M. Savoskin, V. Mochalin, A. Yaroshenko, N. Lazareva, T. Konstantinova,I. Barsukov, I. Prokofiev, Carbon 45 (2007) 2797.

[11] Y. Gao, X. Chen, H. Xu, Y. Zou, R. Gu, M. Xu, A. Jen, H. Chen, Carbon 48 (2010)4475.

[12] G. Mpourmpakis, E. Tylianakis, G. Froudakis, Nano Lett 7 (2007) 1893.[13] V. Coluci, S. Braga, R. Baughman, D. Galvao, Phys. Rev. B 75 (2007) 125404.[14] A. Lu, W. Li, E. Salabas, B. Spliethoff, F. Schuth, Chem. Mater. 18 (2006) 2086.[15] Y. Yang, A. Centrone, L. Chen, F. Simeon, T. Hatton, G. Rutledge, Carbon 49

(2011) 3395.[16] V. Pol, P. Thiyagarajanb, J. Environ. Monit. 12 (2010) 455.[17] W. Cho, E. Hamada, Y. Kondo, K. Takayanagi, Appl. Phys. Lett. 69 (1996) 278.[18] Z. Sun, Z. Yan, J. Yao, E. Beitler, Y. Zhu, J. Tour, Nature 468 (2010) 549.[19] S. Byun, H. Lim, G. Shin, T. Han, S. Oh, J. Ahn, J. Phys. Chem. Lett. 2 (2011) 493.[20] T. Zhao, Y. Liu, J. Zhu, Carbon 43 (2005) 2907.[21] N. Kobayashi, M. Kijima, J. Mater. Chem. 17 (2007) 4289.[22] C. Nguyena, D. Kim, J. Mater. Chem. 21 (2011) 14226.[23] A. Oya, S. Yoshida, J. Alcaniz-Monge, A. Linares-Solano, Carbon 33 (1995)

1085.[24] J. Yamashita, M. Shioya, T. Kikutani, T. Hashimoto, Carbon 39 (2001) 207.[25] S. Shin, J. Janga, S.H. Yoon, I. Mochida, Carbon 35 (1997) 1739.[26] H. Hines, A. Bagreev, T. Bandosz, Langmuir 20 (2004) 3388.[27] J. Zheng, H. Liu, B. Wu, Y. Guo, T. Wu, G. Yu, Y. Liu, D. Zhu, Adv. Mater. 23

(2011) 2460.[28] X. Xie, L. Ju, X. Feng, Y. Sun, R. Zhou, K. Liu, Nano Lett. 9 (2009) 2565.[29] D. Graf, F. Molitor, K. Ensslin., C. Stampfer, A. Jungen, C. Hierold, Nano Lett. 7

(2007) 238.[30] C. Casiraghi, S. Pisana, k. Novoselov, A. Geim, A. Ferrari, Appl. Phys. Lett. 91

(2007) 233108.[31] M. Dresselhaus, G. Dresselhaus, R. Saito, A. Jorio, Phys. Rep. 409 (2005) 47.[32] H. Zhou, C. Qiu, H. Yang, F. Yu, M. Chen, L. Hu, Chem. Phys. Lett. 501 (2011)

475.[33] L. Liu, S. Ryu, M. Tomasik, E. Stolyarova, N. Jung, M. Hybertsen, Nano Lett. 8

(2008) 1965.[34] K. Kudin, B. Ozbas, H. Schniepp, R. Prudhomme, I. Aksay, R. Car, Nano Lett. 1

(2008) 36.[35] R. Silverstein, F. Webster, Spectrometric Identification of Organic Com-

pounds, 6th edn, Wiley, New York, 1998.[36] G. Eda, Y. Lin, C. Mattevi, H. Yamaguchi, H. Chen, I. Chen, C. Chen,

M. Chhowalla, Adv. Mater. 22 (2010) 505.[37] L. Minati, G. Speranza, I. Bernagozzi, S. Torrengo, L. Toniutti, B. Rossi,

M. Ferrari, A. Chiasera, J. Phys. Chem. C 114 (2010) 11068.[38] W. Chen, Z. Fan, L. Gu, X. Baoc, C. Wang, Chem. Commun. 46 (2010) 3905.[39] C. Du, N. Pan, Nanotechnology 17 (2006) 5314.[40] S. Rao, A. Stesmans, K. Keunen, D. Kosynkin, A. Higginbotham, J. Tour, Appl.

Phys. Lett. 98 (2011) 083116.[41] M. Zaka, Y. Ito, H. Wang, W. Yan, A. Robertson, Y. Wu, M. Rummeli,

D. Staunton, T. Hashimoto, J. Morton, A. Ardavan, G. Briggs, J.H. Warner,ACS Nano 4 (2010) 7708.

[42] K. Shen, D.L. Tierney, T. Pietrab, Phys. Rev. B 68 (2003) 165418.