PAPER www.rsc.

org/materials | Journal of Materials Chemistry

Synthesis of poly(vinyl alcohol)/reduced graphite oxide nanocomposites with

improved thermal and electrical properties†
Horacio Javier Salavagione,* Gerardo Martı́nez and Marian A. Gomez
Received 9th March 2009, Accepted 23rd April 2009
First published as an Advance Article on the web 1st June 2009
DOI: 10.1039/b904232f

Poly(vinylalcohol)/reduced graphite oxide nanocomposites have been synthesised by reducing graphite

oxide in the presence of the polymer matrix and coagulating the system with 2-propanol. It has been
observed that some interactions occur between the polymer and the reduced graphite oxide layers,
mainly by hydrogen bonding. These interactions are responsible for a remarkable change in the thermal
behaviour of the nanocomposites. In addition, high electrical conductivity has been achieved at
concentrations beyond 7.5 wt% of reduced graphite oxide (0.1 S cm1), with a percolation threshold
between 0.5 and 1 wt%.

Introduction (GO) has been treated with octadecylamine to produce octa-

decyl-amido graphite15 and with different isocyanates to produce
Polymer-based nanocomposites have been intensively studied in amides and carbamate esters on the GO surface.16 The authors
the last decades with the aim to obtain high-performance light- demonstrated that this treatment resulted in a new class of GO
weight materials. In order to achieve this purpose, clay minerals1 materials with reduced hydrophilic properties which can be
and carbon nanostructures1–4 have been proposed as candidates exfoliated in polar aprotic solvents to yield derivatized graphene
for the filler material. Several polymeric matrices have been oxide nanoplatelets. Thus, the modified GO can be intimately
employed as hosts for carbon nanofibers (CNF) and nanotubes mixed with polystyrene which is soluble in the same solvents and
(CNT) filler, including polyolefins,5 poly(acrylates),6 epoxy,7 reduced ‘‘in-situ’’ to produce polystyrene/graphene nano-
elastomers,8 etc. composites.17
However, the expensive multi-step methods used to prepare On the other hand, the high capacity of PVA to form both
and purify CNT limit their production on an industrial scale. intra- and interchain hydrogen bonding simultaneously makes it
Graphene, a two-dimensional monolayer of carbon atoms, the an interesting polymer which could interact with GO in the same
basic building block for naturally occurring graphite, emerges as way and could be used as a model for understanding the inter-
an interesting alternative to CNT due especially to their unique action of graphite oxide with other polar polymers. PVA/GO18–20
structure and novel electronic properties. and PVA/CNT12,13,21–23 composites have been prepared by mix-
As it occurs with CNT, processing and dispersion of graphene ing solutions of both components.
in the polymeric hosts constitutes the main challenge to use it in Here, we report a strategy for the synthesis of PVA/reduced
nanocomposites, the control of the interfacial interaction being graphite oxide (RGO) nanocomposites by reducing GO in
crucial. The polymer matrix and the filler are bonded to each presence of PVA and coagulating the system with 2-propanol to
other by weak intermolecular forces, and chemical bonding is precipitate the nanocomposite. In fact, the interaction between
rarely involved. If the reinforcing material in the composite could the polymer and the groups remaining on the graphene sheet
be dispersed on a molecular scale and interacted with the matrix significantly alters the thermal properties of the nanocomposites.
by chemical bonding, then significant improvements in the In addition, the nanocomposites display good electrical
mechanical properties of the material or unexpected new prop- conductivity with a low percolation threshold.
erties might be attained. In that sense, CNT have been success-
fully functionalized to improve their interaction with several
Experimental
polymeric matrices.9,10 Thus, amino-functionalized CNT have
been used as reinforcement of epoxy matrices,11 nylon-6 and Chemicals
other polyamides have been grafted from 3-caprolactam–func-
tionalized CNT12 and carboxylic functional groups in CNT have Graphite powder (45 mm), PVA (99+% hydrolyzed, Mw 
been used for etherification with poly(vinyl alcohol) (PVA).13,14 89000–98000) and hydrazine (35% water solution) were
Due to its recent discovery, just a few attempts to modify purchased from Aldrich. Potassium permanganate (KMnO4),
graphene layers have been reported. However, graphite oxide sulfuric acid (H2SO4), hydrogen peroxide (H2O2) and sodium
nitrate (NaNO3) were purchased from Panreac, Spain.

Instituto de Ciencia y Tecnologı́a de Polı́meros, CSIC. Juan de la Cierva 3,

28006 Madrid, Spain. E-mail: horacio@ictp.csic.es; Fax: +34 915 644 853;
Methods
Tel: +34 915 622 900
Graphite oxide (GO) was obtained using Hummer’s method.24
† Electronic supplementary information (ESI) available:
Characterization of graphite oxide and FTIR spectra and DSC curves Briefly, 2 g of graphite was mixed with 1 g of NaNO3 and 50 mL
for PVA and nanocomposites. See DOI: 10.1039/b904232f of H2SO4, and the mixture was cooled down to 0  C. Then, 6 g

This journal is ª The Royal Society of Chemistry 2009 J. Mater. Chem., 2009, 19, 5027–5032 | 5027
KMnO4 was added slowly maintaining the temperature below 5 determined by considering the weight fraction of PVA in the

C. The cooling bath was removed and the suspension was nanocomposites.
maintained for 0.5 h. After that, 100 mL of water was added and Fourier transform infrared spectra (FTIR) of the nano-
the temperature increased to 90  C. The mixture was further composites in the transmission mode were obtained in KBr
diluted with 300 mL of water, treated with 50 mL of 5% H2O2, pellets using a Perkin-Elmer System 200 spectrometer with a 4
filtered and washed with hot water. The graphite oxide was dried cm1 resolution.
by vacuum obtaining a final mass of 3.2 g. The GO powder was DC-Conductivity measurements were carried out using the
examined by X-ray diffraction (XRD). It was observed that no four-probe method on films. The hot-pressed films were cut into
total oxidation was achieved suggesting that GO contains some rectangles (about 0.6 cm wide and 1.2 cm long) and dried under
original graphite (Fig S1, ESI).† vacuum for 24 h. The measurements were carried out using
a four-probe setup equipped with a dc low-current source (LCS-
02) and a digital micro-voltmeter (DMV-001) from Scientific
Nanocomposites preparation
Equipment & Services. The conductivity was calculated by using
PVA/reduced graphite oxide (RGO) nanocomposites with 0.5, 1, eqn (1).
1.5, 2, 5, 7.5 and 10% of filler were prepared as follows: GO was .  
V
dissolved in 20 mL of water and treated with ultrasound for 15 s ¼ 1 r ¼ 4:5324t f1 f2 (1)
I
min. PVA was dissolved in 80 mL water by heating it for 45 min
at 90  C and mixed with the GO solution. The mixture was where t is the thickness of the sample, f1 is the finite thickness
maintained under magnetic agitation for 3 days. Then, the correction for thick samples on an insulating bottom boundary
reduction of GO was accomplished by adding 5 mL of hydrazine and f2 is the finite width correction.25 The thickness correction is
to the PVA/GO solution and keeping it under magnetic stirring described by eqn (2):
overnight. The coagulation of the polymer nanocomposites was
lnð2Þ
accomplished by adding the aqueous solutions dropwise to 100 f1 ¼   (2)
sinhðt=sÞ
mL of 2-propanol under vigorous stirring. The solid nano- ln
sinhðt=2sÞ
composites were filtered, washed abundantly with methanol and
dried at 40  C under vacuum for ca. 24 h. The same procedure, where s is the probe distance (0.2 cm). Usually, the thickness is
without the reduction step, was employed to prepare PVA/ about 0.05 cm and f1 takes values close to 1. The width correction
graphite and PVA/GO composites f2 is a function of the width (d) and length (l) and is obtained
from experimental curves in the literature (in our case, l ¼ 2d).
Techniques
Results and discussion
The X-ray diffractograms (XRD) of the powder nanocomposites
were obtained using a Bruker D8-Advanced X-ray generator (Cu Nanocomposites of PVA/RGO with 0.5, 1, 1.5, 2, 5, 7.5 and 10%
Ka) operated at 40 kV and 40 mA. of the filler were prepared. In all cases, the starting brownish
The dispersion of the nanofiller was examined using a Philips solution of PVA/GO changed immediately to black when
XL30 SEM equipment. The nanocomposite samples were cryo- hydrazine was added, suggesting some graphene formation by
fractured from film specimens and then were coated with ca. 5 reduction of GO. However, no precipitate was observed a few
nm Au/Pd overlayer to avoid charging during electron irradia- hours after the magnetic stirring had been stopped, confirming
tion. that no aggregates of the RGO sheets were formed. This
The thermogravimetric analysis was performed in a TGA behaviour suggests the presence of interactions between the
Q500 equipment from TA instruments. The samples were dried matrix and the RGO which is critical for avoiding RGO sheets
under dynamic vacuum before the experiments and then placed aggregation.
in a platinum pan. The loss of weight was monitored from room Fig. 1 shows the XRD patterns of PVA powder and its
temperature to 950  C using a heating rate of 10  C min1. The nanocomposites with different contents of RGO. PVA shows
experiments were carried out in nitrogen atmosphere. diffraction peaks at 11.3, 19.4 and 22.4 that correspond to the
The crystallization and melting behaviour were investigated by crystalline phase of the polymer.26 In the nanocomposites, the
DSC using a Mettler TA4000/DSC30. The experiments were intensity of the main peaks of the PVA decreased and became
carried out in nitrogen atmosphere using 5 mg sample sealed in broader, showing a decrease in crystallinity and indicating that
aluminium pans. The samples were heated from room tempera- some interactions between polymer chains and the filler may take
ture to 240  C, maintained at this temperature for 5 min, then place.
cooled to room temperature and heated again to 240  C. The In order to characterize the nanocomposites and to determine
heating and cooling rates were 10  C min1 in all cases. The the interaction among polymer chains and with the filler, FTIR
transition temperatures were taken as the peak maximum in the experiments were carried out. It is known that both the –OH
calorimetric curves. The degree of crystallinity (Xc) was calcu- stretching and the –C–OH stretching bands are sensitive to the
lated from the ratio DHa/DHu, which are the apparent and the hydrogen bonding degree. The FTIR spectra of PVA and the
100% crystalline melting enthalpy, respectively. The apparent nanocomposites contain typical bands of PVA (Fig S2†).
heat of fusion DHa is taken from the area of the fusion peak. The However, the broad band around 3400 cm1, involving the
melting enthalpy of a 100% crystalline PVA, DHu, was taken as strong hydroxyl band for free and hydrogen bonded alcohols,
161.6 J g1. The crystallinity of PVA in the nanocomposites was shifts to a lower wavenumber as the amount of filler increases.

5028 | J. Mater. Chem., 2009, 19, 5027–5032 This journal is ª The Royal Society of Chemistry 2009
 PVA.28 The relative intensity of this band (related to the band at
ca. 1100 cm1) decreases as the amount of RGO increases
showing the same trend described by the shifts of the C–OH and
O–H bands (inset in Fig. 2). These results may reinforce the
hypothesis for the existence of hydrogen bonding between the
polymer and the filler, to the detriment of hydrogen bonding
among polymer chains, that diminishes the crystalinity.
Before we analyze the physical properties of the nano-
composites, it is important to note that the filler is composed of
graphite and reduced graphite oxide (see ESI†). In order to
investigate the influence of non-exfoliated graphite on the
physical properties of the nanocomposites, 5 wt% PVA/graphite
and PVA/GO composites were prepared. These compounds will
be used to compare their properties with the ones of PVA/gra-
phene nanocomposite.
In order to determine the degree of exfoliation and the quality
Fig. 1 X-Ray diffraction patterns of PVA and PVA/RGO nano- of the nanofiller dispersion in PVA, scanning electron micros-
composites. copy (SEM) was employed. Fig. 3 shows the SEM images of the
fracture surface of 2 and 5 wt% PVA/RGO nanocomposites.
Composites of PVA/graphite and PVA/GO (not reduced) are
This behaviour has been attributed to the dissociation of the also shown for comparison. A well-dispersed and densely packed
hydrogen bonding among the hydroxyl groups in the polymer.27 distribution of GO and RGO can be seen in 5 wt% nano-
Furthermore, the band assigned to the –C–OH stretching composites. In both cases, graphene platelets more than 5 mm
(around 1100 cm1) displayed a similar behaviour, indicating long and about 50–60 nm thick can be observed. However, it
hydrogen bonding between the –OH in PVA and the oxygenated seems that the voids between the laminates are slightly higher in
groups remaining in RGO.20,27 PVA/GO. Also, some agglomerates appear (around 0.3 mm) due
These band displacements are represented in Fig. 2. A signif- to not total oxidation of the graphite (Fig S1a†). In fact, the
icant effect is observed at low filler concentration. The effect 5 wt% PVA/graphite composite shows filler dimensions of about
decreased when the filler concentration increased over 2% 0.2–0.3 mm thick and less packed distribution (Fig. 3b). Thus, it
probably due to the association between RGO laminates which can be stated that the agglomerates are due to original graphite
became important and should be considered at these loadings. materials and the polymer plays a key role to prevent further
Although the trend in the FTIR bands is clear, factors other sheet agglomeration after the reduction step. The 2 wt% PVA/
than the hydrogen bonding interactions can also be involved RGO nanocomposite displayed separated nanoplatelets of gra-
(e.g., steric effects). However, it is clear that the addition of RGO phene (the white dots are due to the low quality of the fracture
decreases the crystallinity of the polymer (Fig. 1). The band and the cracks are due to the sensitivity of the PVA to the elec-
around 1155 cm1 has already been related to the crystallinity of tron beam)

Fig. 2 Displacements of the –OH and –C–OH stretching bands with the Fig. 3 SEM images of PVA/GO 5 wt% (a), PVA/graphite 5 wt% (b),
RGO content in the nanocomposites. Inset: variation of the height ratio PVA/RGO 2 wt% (c) and 5 wt% (d) nanocomposites. The scale bar
between 1155 and 1100 cm1 bands with the RGO contents. corresponds to 2 mm.

This journal is ª The Royal Society of Chemistry 2009 J. Mater. Chem., 2009, 19, 5027–5032 | 5029
Table 1 DSC crystallization and melting parameters of PVA and PVA/ decrease, effecting more compact packaging, which increases the
RGO nanocomposites rigidity of the system.
Sample Tg/ C Tm/ C Tc/ C DHm/J g1 Xc (%)a
The DSC cooling scans show a strong dependence on the filler
nature and content. The Tc decreases significantly as the filler
PVA 82 225 203 83 51 loading increases in the RGO composites, which indicates that
PVA/RGO 1 wt% 84 214 196 60.9 37 the crystallization is retarded. On the other hand, no changes are
PVA/RGO 2 wt% 94 200 180 54.5 33
PVA/RGO 5 wt% 94 189 169 33.2 19 observed in Tc for PVA/GO and PVA/graphite nanocomposites.
PVA/RGO 10 wt% 103 160 5.2 3 The differences in crystallization behaviour observed for the
PVA/graphite 5 wt% 84 225 206 70.1 43 composites can be explained considering the existence of inter-
PVA/GO 5 wt% 87 221 202 55.0 34 actions between the filler and the matrix in the reduced graphite
a
Obtained from the melting endotherms. oxide nanocomposites as was previously pointed out from other
experimental evidences (FTIR and DRX), and they suggest that
some steric effects are eliminated through the reduction step.
In order to investigate the effects of the RGO on the thermal
The crystallization behaviour and crystalline parameters of the stability of the polymer matrix, thermo-gravimetric analysis was
polymer matrix in the composite were also investigated because performed on all composites (Fig. 4). TGA curves for PVA/RGO
the macroscopic properties of the nanocomposites strongly suggests that the filler plays an important role to improve the
depend on them. The crystallization and melting behaviour were thermal stability of the nanocomposites. The temperature of the
analyzed by cooling down from the melt to 30  C at 10  C min1 maximum degradation rate for the nanocomposites (obtained
and subsequent heating up, in the same experiment, to 240  C from the derivative of TGA curves, Fig 4b) increases by more
(Fig S3†). The parameters derived from the experimental DSC than 100  C. This behaviour has been attributed, in similar
curves are summarized in Table 1. Significant changes can be systems, to the stability of the hydrogen bonding between the
observed in the crystallization temperature (Tc), the melting polymer and the GO.20 In our case, most of the oxygenated
temperature (Tm) and the glass transition temperature (Tg) of
PVA in the nanocomposites depending on filler content.
Regarding the heating scan, it can be observed that Tm and the
melting enthalpy (DHm) are strongly influenced by the filler
loading. As the RGO content increases, the melting endotherm
of PVA in the nanocomposite broadens and Tm decreases. In
addition, the crystallinity of PVA in the nanocomposites
decreases from 51% (for PVA) to an almost amorphous
material at 10 wt% of RGO (Table 1), supporting the existence of
some interaction between the polymer and the filler to the
detriment of interactions among polymer chains. These results
are confirmed by the DSC heating scan of PVA/graphite
composite, where a sharp melting peak with the same Tm and
slightly lower Xc with respect to PVA is observed (Fig S3 in ESI,†
Table 1). Therefore, the rests of the graphite in the filler behaves
as inert material in the nanocomposite and the reduced graphite
oxide platelets are responsible for the changes in the properties.
However, the melting behaviour of PVA/GO displays a broad
melting peak (similar to the nanocomposites), but with Tm values
close to that for PVA and PVA/graphite. It appears that the
reduction step, which eliminates oxygenated groups, changes the
interaction between the filler and the polymer. Therefore, addi-
tional effects to the hydrogen bonding are also involved (e.g.,
steric effects). In addition, the Xc of PVA/GO lies between the
values for PVA/RGO 1 and 2 wt%, suggesting that the rigidity of
the nanocomposite is altered after the reduction step. These
changes should affect the mobility of the chains.
In order to investigate this effect, the glass transition region
was analyzed (Fig S3, ESI).† A remarkable increase in the Tg is
observed as the filler loading increases over 1 wt% (Table 1). This
effect can be ascribed to an effective attachment of the polymer
to the layers of RGO that prevents the segmental motions of the
polymer chains, as occurs with different polymer matrices
intercalated in laminar montmorillonite (MMT).29 However, the
Tg for PVA/GO composite is similar to that for PVA. Therefore, Fig. 4 Thermogravimetric curves of PVA and PVA/RGO nano-
the removal of oxygenated groups causes the steric effect to composites (a) weight loss percent and (b) derivative curves.

5030 | J. Mater. Chem., 2009, 19, 5027–5032 This journal is ª The Royal Society of Chemistry 2009
groups were reduced to produce graphene sheets. However, it has CNT, which generates a higher degree of structural defects that
been proposed that the carboxylic groups on the border of the causes the material to lose its special electronic properties.33,34
GO laminates are not reduced under the reaction conditions we Also, under these conditions very short tubes with a low aspect
used.30,31 In fact, the PVA/GO composite displays the same ratio are produced.34 Therefore, although original graphite is
behaviour with a temperature of maximum degradation rate present, the exfoliation and reduction of the oxidised laminates
around 365  C (not shown). are very important steps.
Fig 4 also shows that although a small weight loss is observed In addition the second increase of conductivity occurs between
around 200  C commonly in these systems; the onset of PVA 5 and 10 wt% (3 to 6.2 vol%). This transition has also been
decomposition for the composites is retarded. Similar behaviour observed in PMMA/graphite32 and in epoxy/bromine-treated
has been observed for PVA/CNT composites with less than 20% graphite nanoplatelets composites35 and has been attributed to
of filler and has been attributed to the absorption, by the carbon hopping of electrons between filler laminates by tunnelling
surface, of free-radicals generated during polymer decomposi- through the polymer layers.32 We think that tunnelling also
tion.21 Here, a low quantity of RGO (0.5 wt%) produces the same occurs in our system but we have not explanation for the fourth
effect, due to their higher aspect ratio. order of magnitude increase.
One of the main challenges of nanocomposites of insulating Here, it is very important to emphasize that for 5 wt% PVA/
polymers and very conductive fillers is to achieve materials with graphite and PVA/GO composites no measurable conductivity
improved electrical properties. The fillers, normally CNT and values were obtained. It is well known that GO is intrinsically an
graphite nanosheets, exhibited a high aspect ratio which is insulating material. However, although graphite is a conductive
advantageous for forming conducting networks in a polymer material, this did not confer conductivity to the composite.
matrix. The conductivity of the PVA/RGO nanocomposites as Although the maximum conductivity value obtained for 7.5
a function of the filler loading has been measured by the four- wt% PVA/RGO is slightly lower than those reported for other
probe method (Fig. 5). Three-stage behaviour can be observed in systems (PS/graphene17 and PMMA/graphite32), it is rather
which two different steps occurred in conductivity. The first higher than the one reported for PVA/CNT composites, which
increase can be assigned to the typical insulating-conductive requires more than 20 wt% of CNT to achieve conductivity in the
percolation behaviour. That is, conductivity remains at very low same order of magnitude.23
values (1013 to 1014 S cm1) until a critical loading content is Taking into account the results described above, it can be
achieved. Then, it increases quickly by several orders of magni- noted that (i) the presence of non-exfoliated graphite is not
tude up to 8  106 S cm1. The percolation threshold lies responsible for the increase in the conductivity of PVA/RGO,
between 0.5 and 1 wt% of RGO. If we use the density of PVA and (ii) the reduction step is crucial to eliminating the functional
(1.312 g cm3) and assume the graphite oxide density as groups that limit conductivity, and (iii) the polymer plays a key
2.2 g cm3 (similar to ref. 17) to convert the values to vol% this role in avoiding filler aggregation.
value changes to 0.3–0.6% which is a little higher than those In addition, the system described in this paper could be opti-
obtained for polystyrene/graphene,17 and close to poly(methyl mized by controlling some other parameters such as the degree of
methacrylate)(PMMA)/graphite nanosheets.32 The higher oxygenated groups in the graphene layers and the type of inter-
percolation threshold in our case seems reasonable because of the action with the matrix (e.g., covalent bonds). In fact, we used
presence of non-exfoliated graphite, which diminishes the filler partially oxidized graphite as starting material. Thus, by
aspect ratio. However, the reported percolation threshold in increasing the degree of oxidation (and thus, the degree of
similar systems23 composed of PVA and CNT (between 5 and exfoliation) better nanocomposite properties are expected.
10 wt%) is quite higher than the one obtained in this work. This
difference could be attributed to the previous acidic treatment of
Conclusions
A successful method to produce PVA/RGO nanocomposites,
based on the reduction of graphite oxide in the presence of the
polymer was developed. The nanocomposites prepared by this
procedure display interactions among the polymer and the filler
due to the remaining oxygenated groups in graphene. These
interactions significantly alter the thermal properties of the
nanocomposites, lowering the Tm more than 35  C and
decreasing crystallinity down to a half for nanocomposites with 5
wt% of RGO. The materials display good electrical conductivity
with a percolation threshold below 1 wt% of filler and conduc-
tivities as high as 0.1 S cm1 for composites with 7.5% wt of
RGO.

Acknowledgements
Financial support from the Spanish Ministry of Science and
Fig. 5 Plot of the electric conductivity of PVA/RGO nanocomposites as Innovation, MICINN (MAT 2006-13167-C02-01) is gratefully
a function of the RGO content. acknowledged. H.J.S. thanks the Consejo Superior de

This journal is ª The Royal Society of Chemistry 2009 J. Mater. Chem., 2009, 19, 5027–5032 | 5031
Investigaciones Cientı́ficas (CSIC) for a postdoctoral contract 17 S. Stankovich, D. A. Dikin, G. H. B. Dommett, k. M. Kohlhass,
JAE-Doc. The authors thank Dr Gary Ellis and Dr Carlos E. J. Zimney, E. A. Stach, R. D. Piner, S. T Nguyen and
R. S. Ruoff, Nature, 2006, 442, 282–286.
Marco for helpful discussions. 18 H. Kaczmarek and A. Podg orski, Polym. Deg. & Stab., 2007, 92, 939–
946.
19 J. Xu, Y. Hu, L. Song, Q. Wang, W. Fan and Z. Chen, Carbon, 2002,
40, 445–467.
References 20 Y. Matsuo, K. Hatase and Y. Sugie, Chem. Mater., 1998, 10, 2266–
2269.
1 D. R. Paul and L. M. Robeson, Polymer, 2008, 49, 3187–3204. 21 K. P. Ryan, M. Cadek, V. Nicolosi, S. Walker, M. Ruether,
2 J. N. Coleman, U. Khan and Y. K. Gun’ko, Adv. Mater., 2006, 18, A. Fonseca, J. B. Nagy, W. J. Blau and J. N. Coleman, Synth.
689–706. Met., 2006, 156, 332–335.
3 M. Moniruzzaman and K. I. Winey, Macromolecules, 2006, 39, 5194– 22 J. N. Coleman, M. Cadek, R. Blake, V. Nicolosi, K. P. Ryan,
5205. C. Belton, A. Fonseca, J. B. Nagy, Y. K. Gun’ko and W. J. Blau,
4 N. Grossiord, J. Loos, O. Regev and C. E. Koning, Chem. Mater., Adv. Funct. Mater., 2004, 14, 791–798.
2006, 18, 1089–1099. 23 M. S. P. Shaffer and A. H. Windle, Adv. Mater., 1999, 11, 937–941.
5 A. Linares, J. C. Canalda, E. Cagiao, M. C. Garcı́a-Gutierrez, 24 W. S. Hummers, Jr. and R. E. Offeman, J. Am. Chem. Soc., 1958, 80,
A. Nogales, I. Martı́n-Gull on, J. Vera and T. A. Ezquerra, 1339.
Macromolecules, 2008, 41, 7090–7097. 25 Semiconductor Material and Device Characterization, ed. D. K.
6 G. L. Hwang, Y.-T. Shieh and K. C. Hwang, Adv. Funct. Mater., Scroedeer, John Willey and Sons, New Jersey, 2006.
2004, 14, 487–491. 26 L. A. Garcı́a-Cerda, M. U. Escare~ no-Castro and M. Salazar-
7 B. L. Wardle, D. S. Saito, E. J. Garcı́a, A. J. Hart, R. Guzman de Zertuche, J. Non-Cryst. Solids, 2007, 353, 808–810.
Villoria and E. A. Verploegen, Adv. Mater., 2008, 20, 2707–2714. 27 L. Lu, H. Sun, F. Peng and Z. Jiang, J. Membrane Sci., 2006, 281,
8 S. V. Ahir, Y. Y. Huang and E. M. Terentjev, Polymer, 2008, 49, 245–252.
3841–3854. 28 S. K. Mallapragada and N. A. Peppas, J. Polym Sci. Part B: Polym.
9 S. Wang, R. Liang, B. Wang and C. Zhang, Chem. Phys. Lett., 2008, Phys., 1996, 34, 1339–1346.
457, 371–375. 29 A. Leszczynska and K. Pielichowski, J. Therm. Anal. Calorim., 2008,
10 L. Dai and A. W. H. Mau, Adv.Mater., 2001, 13, 899–913. 93, 677–687.
11 J. Shen, W. Huang, L. Wu, Y. Hu and M. Ye, Composite Sci. 30 S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas,
Technol., 2007, 67, 3041–3050. A. Kleinhammes, Y. Jia, Y. Wu, S. T. Nguyen and R. S. Ruoff,
12 Y. Lin, M. J. Meziani and Y.-P. Sun, J. Mater. Chem., 2007, 17, 1143– Carbon, 2007, 45, 1558–1565.
1148. 31 D. Li, M. B. M€ uller, S. Gilje, R. B. Kaner and G. G. Wallace, Nat.
13 M. C. Paiva, B. Zhou, K. A. S. Fernando, Y. Lin, J. M. Kennedy and Nanotechnol., 2008, 3, 101–105.
y.-P. Sun, Carbon, 2004, 42, 2849–2854. 32 G. Chen, W. Weng, D. Wu and C. Wu, Eur. Polym. J., 2003, 39, 2329–
14 Y. Lin, B. Zhou, K. A. S. Fernando, P. Liu, L. F. Allard and Y.- 2335.
P. Sun, Macromolecules, 2003, 36, 7199–7204. 33 M. A. Hamon, J. Chen, H. Hu, Y. Chen, M. E. Itkis, A. M. Rao,
15 S. Nigoyi, E. Bekyarova, M. E. Itkis, J. L. McWilliams, M. A. Hamon P. C. Eklund and R. C. Haddon, Adv. Mater., 1999, 11, 834–840.
and R. C. Haddon, J. Am. Chem. Soc., 2006, 128, 7720–7721. 34 A. Hirsch, Angew. Chem. Int. Ed., 2002, 41, 1853–1859.
16 S. Stankovich, R. D. Piner, S. T. Nguyen and R. S. Ruoff, Carbon, 35 J. Li, L. Vaisman, G. Marom and J.-K. Kim, Carbon, 2007, 45, 744–
2006, 44, 3342–3347. 750.

5032 | J. Mater. Chem., 2009, 19, 5027–5032 This journal is ª The Royal Society of Chemistry 2009