Carbon 45 (2007) 744750

[Link]/locate/carbon

Br treated graphite nanoplatelets for improved electrical conductivity

of polymer composites
Jing Li a, Linda Vaisman b, Gad Marom b, Jang-Kyo Kim a,*

a
Department of Mechanical Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, Hong Kong
b
Casali Institute of Applied Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel

Received 5 September 2006; accepted 25 November 2006

Abstract

The graphite nanoplatelets (GNP) were treated by vapor-phase bromination. The increase in weight and atomic concentration of Br
indicated the bromine uptake. The intercalation of Br between graphene layers of GNP was confirmed by the X-ray diraction result,
showing an increase in the interlayer spacing from 3.342 A to 3.361 A. Two types of bonds between C and Br were introduced simul-
taneously, ionic and covalent bonds, both of them increased with bromination duration. The fraction of ionic bond reached the highest
value by 3 h Br exposure, which corresponded to the highest electrical conductivity of GNP. Although the bromination treatment did not
change the percolation threshold of composites, it increased the absolute value of electrical conductivity of composites when the filler
content was higher than the percolation threshold.
! 2006 Elsevier Ltd. All rights reserved.

1. Introduction electrical conductivity of composite materials is the most

critical requirement. Therefore, significant eorts have
Graphite nanoplatelet (GNP) is a nanometer-scale con- been directed towards improving the electrical conductivity
ductive filler produced by exfoliating natural flake graph- of conducting composites. It has been reported [4,5] that
ite, and has attracted significant attention as the low cost intercalation of guest atoms between the graphene planes
and lightweight alternative to metal- and carbon-based forming ionic bonds (charge transfer) can enhance the elec-
electrically conducting reinforcements for conducting poly- trical conductivity of graphitic material by adding electrons
mer composites. GNPs are produced from graphite flakes to its conduction band or holes to the valence band. Mean-
intercalated with high concentrated acids, which can be while, the formation of covalent bonds between the guest
expanded up to a few hundreds of times along its c-axis atoms and graphite can depress the carrier concentration
at a high temperature. The exfoliation of expanded graph- and displace some of the C atoms slightly away from the
ite results in separation of the graphite sheets into nano- highly planar structure of pristine graphite, which in turn
platelets with a very high aspect ratio [1,2]. The high decreases the electrical conductivity.
aspect ratio and the large surface area of GNPs are respon- Amongst many intercalation species, some compounds
sible for the much lower percolation threshold and better such as AsF5 would most probably generate higher con-
electrical conductivity of conducting polymer composites ductivity improvement than bromine [4]. We chose bro-
than micrometer-scale conventional reinforcements. [1,3] mine as the model compound for three reasons: (1)
Conducting polymer composites have many potential bromine is less toxic and easier to intercalate; (2) bromi-
applications in electromagnetic interference shielding for nated carbon fibres showed long term stability under
electronic devices and electrostatic dissipation, where high ambient condition [4] and a high thermal stability below
150 "C [4,5]; (3) bromine has intermediate chemical activi-
*
Corresponding author. Tel.: +852 2358 7207; fax: +852 2358 1543. ties in hostguest interactions, compared to iodine which
E-mail address: mejkkim@[Link] (J.-K. Kim). allows a weak hostguest reactivity, thus introducing only

0008-6223/$ - see front matter ! 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/[Link].2006.11.031
 J. Li et al. / Carbon 45 (2007) 744750 745

weak charge transfer, and fluorine, which is so strong that phenylenediamine (mPDA, supplied by Sigma-Aldrich), was added into
covalent bonding can be easily formed with host materials. the mixture in the ratio of 14.5/100 by weight. Composite plates were cast
using an aluminum mould with a cavity of 3 mm in thickness, which was
Bromine intercalation has been successfully used to cured at 80 "C for 2 h, followed by a post cure at 150 "C for 3 h. Fig. 1
increase the electrical conductivity of carbon fibres [46]. shows the above described process flow of GNPs and the corresponding
The charge transfer was the main bonding type between composites.
Br and C, and covalent bonding was introduced only after
prolonged reactions [5]. A higher level of graphitization of 2.2. Characterizations
the host fiber induced better intercalation and higher elec-
The inter-graphene d-spacing of GNPs was determined using a powder
trical conductivity of doped carbon fibres [4,6]. Vapor-
X-ray diractometer (XRD, Philips PW1830). XRD patterns were
phase bromination treatment has also been applied to obtained using a Cu Ka radiation (k = 0.154 nm) with a step size of
nanometer-scale graphite, such as single wall carbon nano- 0.05" and at a scanning rate of 2.0"/min, working at 45 kV 40 mA. X-
tubes [7] and onion-like nanographite particles [8]. A ray photoelectron spectroscopy (XPS, PHI 5600), equipped with a mono-
remarkable 30 time increase in electrical conductivity was chromatic Al Ka X-ray source operated in a residual vacuum of 5 10!9
Torr, was employed to detect the changes in elemental compositions of
reported [7] for single wall carbon nanotubes through the
GNPs after the bromination treatment. The emission angle of the photo-
formation of charge transfer complexes. Bromine intercala- electron, the angle between the sample surface and the axis of the analyzer,
tion showed similar eects on the electronic properties of was 45", corresponding to a sampling depth of approximately 4.7 nm [10].
onion-like nanographite [8]. The pass energy was 187.85 eV for survey scans and 23.5 eV for high-res-
This paper is part of a larger project on fabrication, olution scans. The high-resolution spectra was fitted with multitak soft-
ware (provided by physical electronics). The bonding types between the
properties characterization and applications of conducting
C and Br atoms were analyzed using the time-of-flight secondary ion mass
GNP composites. In our previous studies [2,9], the UV/O3 spectrometry (ToF-SIMS, PHI 7200). The primary Cs+ ion source was
treatment technique was successfully applied to improve operated at 8 kV and the scanning area was 200 lm square. Secondary
the electrical conductivity as well as the thermo-mechanical ions in the mass range from 0 to 600 amu were collected. The XPS and
and mechanical properties of GNP/epoxy composites. Tof-SIMS results presented below are the mean values and the corre-
sponding standard deviations of three independent measurements
Here, GNPs were treated with bromine to improve the elec-
obtained for a given set of conditions.
trical properties of conducting polymer composites made The bulk electrical conductivities of GNPs and GNP/epoxy compos-
therefrom. With much larger density of open edges than ites were measured at room temperature using a resistivity/Hall measure-
the other carbon materials mentioned above, GNPs would ment system (Bio-Rad HL5500PC) based on the four probe van der Pauw
show significant synergy in improving the electrical con- method. The DC current source of 1 mA was applied for measuring the
conductivity of GNP and 0.1 lA for measuring the conductivity of
ductivity after the treatment. X-ray photoelectron spectros-
GNP/epoxy composites. GNPs were compacted into tablets of 5 mm in
copy (XPS), X-ray diraction (XRD) analysis and time-of- diameter and 0.5 mm in thickness using a 10 ton hydraulic press. Compos-
flight secondary ion mass spectrometry (ToF-SIMS) were ite plates were cut and polished into specimens of 10 mm square 1 mm in
employed to evaluate the intercalation states of Br atoms thickness. As the highest limit of the electrical resistivity measured using
and quantitatively analyze the bonding types between the the current resistivity/Hall measurement system was 108 X cm, the electri-
cal resistivity of composites higher than this limit was measured using a
C and Br atoms.
programmable curve tracer (Sony Tektronix 370A) with the lowest DC
current source of 1 nA. The initial linear part of IV curve was used to cal-
2. Experimental culate the electrical resistivity. Silver paste was applied onto the composite
specimens to ensure low contact resistance with the measuring probes.
2.1. Sample preparation Three-point flexure test was performed to measure the mechanical
properties of GNP/epoxy composites according to the specification,
GNPs were prepared according to the method described in our previ- ASTM standard D790-96. The moulded composite plates were cut into
ous studies [2,9]. Graphite intercalation compound (GIC) containing 12.7 mm wide 70 mm long 3 mm thick samples, which were subjected
2.8 wt% of sulfur as intercalant (supplied by Asbury Graphite Mills, to bending with a support span of 50 mm at a constant cross-head speed
USA) was put into an oven that was maintained at 1050 "C and taken of 1.3 mm/min on a universal testing machine. Five specimens were tested
out after 30 s to produce expanded graphite. Upon rapid heating the for each set of conditions.
GIC was expanded explosively several hundred times along the thickness
direction due to the evaporation of the intercalant and the thermal shock.
Rapid heating Expanded Ultrasonication
The expanded graphite was immersed in acetone and ultrasonicated at GIC GNP
graphite
70 W and 42 kHz for 8 h to obtain exfoliated GNPs. The thickness and
diameter of GNPs were estimated to be 4.5 nm and 46 lm on average [2].
Bromination
GNPs were exposed to Br2 vapor of the vapor pressure of liquid Br2 at
room temperature in a closed desiccator for varying durations. The bromi-
nated GNPs were transferred to the fume hood and ventilated to allow Shear mixing Brominated
desorption of physically absorbed bromine until a constant weight was Ultrasonication GNP
reached. It is the residual bromine intercalation compound that has tech- mPDA Mixture
Outgassing Epoxy
nological significance because of its stability in ambient air. The weights of
GNPs before and after the treatment were recorded in two sets of identical Curing
experiments.
Brominated GNPs were mixed with epoxy (Epon 828, supplied by
Shell Chemical) using a high-shear mixer for 30 min at 3000 rpm, followed Composite
by ultrasonication at 70 W 42 kHz for 30 min at 80 "C. The mixture was
then outgassed at 80 "C for 2 h in a vacuum oven. Curing agent, 1,3- Fig. 1. Process flow of GNPs and the corresponding composites.
746 J. Li et al. / Carbon 45 (2007) 744750

3. Results and discussion

3.1. Bromine uptake and intercalation Pristine

1h
The uptake of bromine by GNP upon its exposure to 3h

Counts
molecular Br (generally termed bromination) is expected 10 h
to become evident by weight increase, which occurs by two 72 h
parallel routes forming either intercalation complexes or 168 h
covalent CBr groups. Because only the first is expected
to aect the p-electron interaction between the bromine
and GNP, it is important that in addition to the overall
uptake the proportion of the two compounds be deter- 25.6 26.0 26.4 26.8 27.2
mined. Fig. 2 shows the changes in weight and Br atomic Degree
concentration determined by XPS as a result of bromina- Fig. 3. XRD spectra of brominated graphite with dierent treatment
tion. Evidently, the weight increased rapidly with bromin- durations.
ation duration and saturated after about 10 h, and then
stabilized or decreased slightly. This phenomenon can be
Table 1
explained by the dynamic process of absorption and XRD results with bromination duration
desorption of bromine atoms into graphite. For treatments
Duration (h) 0 1 3 10 72 168
below 10 h, bromine was absorbed rapidly because of the
d-spacing (A) 3.342 3.355 3.348 3.361 3.361 3.361
ample availability of doping/reactive sites in the GNPs.
WHM (") 0.345 0.365 0.5 0.335 0.32 0.32
Upon filling of these sites, further treatment did not induce
additional absorption, rather, slight desorption of bromine
took place, which can be attributed perhaps to gradual
of XPS is up to approximately 4.7 nm, while the thickness
changes in the nature and balance of the CBr bond types
of GNP is about 4.5 nm.
with bromination duration.
The (0 0 2) d-spacing of graphite is an eective quantita-
Another evidence for bromine uptake was the increase
tive measure of the extent of intercalation. The XRD spec-
of Br atomic concentration with bromination duration
tra of brominated graphite, given in Fig. 3, indicate that
determined by XPS, as shown in Fig. 2. Although the point
the peak corresponding to the 3 h treatment diers in posi-
of maximum Br concentration was obtained for the 72 h
tion and breadth from those corresponding to the other
treatment, which is slightly dierent from the result of
bromination durations. Table 1 shows that with a 3 h treat-
the weight measurement, the two trends are virtually iden-
ment a point of instability is reached in which the d-spacing
tical. Noting that the weight of GNPs was measured in
exhibits a minimum while the width-at-half-maximum
ambient condition, but the XPS experiments were carried
(WHM) of the peak exhibits a maximum. Thereafter, for
out in a residual vacuum of 5 10!9 Torr, the dierences
longer treatments, they stabilize on 3.361 A and 0.32",
probably arose from the desorption of Br atoms during
respectively. The broadened peak is a reflection of the dis-
the evacuation for the XPS experiment. The XPS results
order of graphite crystals during the intercalation process,
can be taken to express the bulk rather than the surface
which is also responsible for the low d-spacing after the 3 h
chemical composition of GNP, because the sampling depth
treatment. It appears that after a transition period, the
graphite crystals become ordered again. As indicated by
the WHM values, the point of maximum at the 3 h treat-
ment is associated with the shortest coherence length of
Atomic concentration of Br by XPS (%)

graphite crystals [11].

20 2.0

3.2. The nature of the CBr bond

Weight change (%)

15 1.5

There are two possible types of CBr bonds in the bro-

10 1.0
minated graphite, namely charge transfer complexes and
covalent bonds. They are distinguishable by ToF-SIMS
5 Weight change of GNP 0.5 as shown in typical ToF-SIMS spectra of brominated
Weight change of second experiment
Atomic concentration of Br GNP in Fig. 4. Atomic bromine exhibits two isotopic peaks
0 0.0 at molar masses 79 and 81 with an intensity ratio of
0 20 40 60 160 180 50.5:49.5, respectively; molecular bromine, Br2, exhibits a
Bromination duration (h) triplet at 158, 160 and 162 with a respective intensity ratio
Fig. 2. Weight changes and Br atomic concentration determined by XPS of about 1:2:1; CBr exhibits a doublet at 91 and 93; and
as a function of bromination durations. C2Br exhibits a doublet at 103 and 105. The reading of
 J. Li et al. / Carbon 45 (2007) 744750 747

Fig. 4. ToF-SIMS spectra of brominated GNP.

CBr and C2Br species indicates existence of covalent bond. lently-bonded Br, and the results are shown as a function
The normalized intensities of the peaks at 91, 93, 103 and of bromination duration in Fig. 7. The fraction of covalent
105 are proportional to the concentration of covalently- Br showed the lowest value after the 3 h treatment, which
bonded Br. The intensity was normalized by dividing the coincides with the highest WHM value obtained from the
areas under the CBr and C2Br peaks by the area under XRD analysis.
the C2 peak (at 24), and the result is shown in Fig. 5. It An additional quantitative study was made by plotting
is seen that the normalized intensity of covalent Br gradu- the atomic concentration of covalent and ionic Br as a
ally increases with increasing bromination duration.
Although the existence of various covalently-bonded Br
species was confirmed by the ToF-SIMS analysis, this tech-
nique could not provide quantitative information about
their concentrations and relative intensities compared to
ionic bonded Br. Therefore, an XPS analysis was carried
out, and typical XPS spectra corresponding to the Br 3d
peaks after dierent bromination durations are shown in
Fig. 6a. The split peak originates from two dierent dou-
blets superimposed in such a way that the peak located
at the lower binding energy corresponds to the ionic state
of Br, while the upper one is associated with covalently-
bonded Br. It is apparent that the intensities of both the
total Br and the fraction of covalent Br increase with
increasing the bromination duration. Quantitative analysis
was performed by deconvoluting the spectrum into two
doublet sets (1 + 2 and 3 + 4), as shown in Fig. 6b. In each
doublet, the separation between the two peaks was 1.05 eV,
the intensity ratio was 3:2 and the WHM was equal [12].
The area under the 3 + 4 doublet divided by the total area
under the Br 3d peak was defined as the fraction of cova-
Normalizaed intensity of covalent Br

0.045
0.040
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
0 20 40 60 160 180
Bromination duration (h)

Fig. 5. Normalized intensity of covalent Br as a function of bromination Fig. 6. XPS spectra (a) of Br 3d for varying bromination durations and
duration. (b) typical deconvoluted curve fitting.
748 J. Li et al. / Carbon 45 (2007) 744750

0.7 ing eect on the electrical conductivity of GNP because the

ionic Br acts as an ionic acceptor. Conversely, covalent
0.6
bonds decrease the electrical conductivity of GNP due to
Fraction of covalent Br

0.5 depressed carrier concentrations. As these two mechanisms

0.4
cancel out each other, the final conductivity of GNP is to
be determined by the ionic/covalent bond balance.
0.3

0.2 3.3. Electrical conductivity of brominated GNP

0.1 and its epoxy composites
0
1 10 100 1000 The electrical conductivity of GNP is plotted as a func-
Bromination duration (h) tion of bromination duration in Fig. 9a. After a 3 h treat-
ment, the electrical conductivity rose from 353 S/cm to
Fig. 7. Fraction of covalently-bonded bromine as a function of bromin-
ation duration. reach a maximum of 540 S/cm (an increase of 53%), corre-
sponding to the spike in d-spacing for the 3 h treatment
(Table 1) and to the lowest fraction of covalent Br
function of bromination duration as shown in Fig. 8. The (Fig. 7). In comparison, the electrical resistivity of dierent
atomic concentrations of covalent and ionic Br are equiva- types of carbon fibers decreased by 0%70% after Br treat-
lent to the respective products of the fractions of covalent ment [6]. The dependence of electrical conductivity of GNP
or ionic Br and the total atomic concentration of Br. It is is plotted as a function of the fraction of covalent Br in
interesting to note that the atomic concentrations of both Fig. 9b. The electrical conductivity of GNP decreased with
covalent and ionic Br increase with bromination duration increasing the fraction of covalent Br, which confirmed the
and reach a plateau value after about 10 h, indicating the deteriorating eect of covalent bonding on electrical con-
saturation of bromine uptake (Fig. 2). The trend of cova- ductivity as discussed above. It is interesting to note that
lent Br with respect to bromination duration is consistent the electrical conductivity of GNP spiked for a 3 h bromin-
with the ToF-SIMS results shown in Fig. 5. The ionic bro- ation treatment which coincides with an abrupt change in
mine concentration increased rapidly for 1 h3 h of bro- the 0 0 2 d-spacing and WHM.
mination due to the formation of intercalation complexes Based on these observations, the 3 h treatment was cho-
between Br and C, as reflected by the broadened XRD sen to prepare brominated GNP for GNP/epoxy compos-
peak at 3 h of bromination (Fig. 3 and Table 1). ites with dierent GNP contents, whose electrical
It has been reported for carbon fibers that bromination conductivity was measured. Fig. 10 presents the electrical
treatment generates mainly charge transfer (ionic) bonding conductivity as a function of GNP content for both treated
between C and Br and only prolonged treatment durations and untreated GNPs. A typical percolation phenomenon is
can generate covalent bonds [5]. Covalent CBr bonds can observed with increasing GNP content: the electrical con-
form more readily with C atoms at the edges of the gra- ductivity changes abruptly at around 1 wt% GNP, where
phitic basal plans, hence the GNP with more edge C the composite undergoes an insulatorconductor transi-
atoms are more susceptible than carbon fibers to forma- tion. To determine the percolation thresholds of the com-
tion of covalent CBr bonds. So, in Fig. 8, covalent and posite, the first derivatives of the conductivity plots were
ionic Br was introduced simultaneously to GNPs. As dis- calculated, and the results are plotted as a function of
cussed in the Introduction, ionic bonds have an ameliorat-

800

0.9 700
Atomic concentration (%)

Electrical conductivity (S/cm)

0.8
0.7 600

0.6 500
0.5
0.4 400
0.3
300
0.2 Covalent Br
0.1 Ionic Br 200
0.0
0 20 40 60 160 180 100
0 20 40 60 160 180 0.3 0.4 0.5
Bromination duration (h) Bromination duration (h) Fraction of covalent Br

Fig. 8. Atomic concentration of covalent and ionic Br as a function of Fig. 9. Electrical conductivity of GNP as a function of (a) bromination
bromination duration. duration; (b) fraction of covalent Br.
 J. Li et al. / Carbon 45 (2007) 744750 749

1.E+00
1.E-01
1.E-02
1.E-03

Electrical conductivity (S/cm)

1.E-04
1.E-05

Differentiation of Log (conductivity)

1.E-06
6
1.E-07
1.E-08
4
1.E-09
1.E-10 2
1.E-11 Without Br treatment
1.E-12 With Br treatment for 3h 0
0 1 2 3 4
1.E-13 GNP content (wt%)
1.E-14
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
GNP content (wt%)

Fig. 10. Electrical conductivity of GNP/epoxy composites with and without Br treatment as a function of GNP content. Insertion is the absolute values of
dierentiation of the upper two curves.

GNP content in the inset of Fig. 10. The GNP contents covalently-bonded Br at the edges of the graphitic basal
corresponding to the highest absolute derivatives were planes improves the interfacial adhesion between GNP
taken as the percolation thresholds [2], giving "1 wt% (or and epoxy. This is manifested by the results of the mechan-
"0.5 vol%) for the composites containing both the treated ical properties for treated and untreated GNPs-filled epoxy
and untreated GNPs. Apparently, the percolation thresh- composites. Fig. 11 presents the results of the flexural
old remained unchanged after the bromination treatment strength as a function of GNP content, showing signifi-
because this property of conducting polymer composites cantly higher values for the brominated GNP composites.
depends mainly on the aspect ratio of the filler. As the bro- In fact, it exhibits (in the lower range of GNP content)
mination treatment is not expected to have a significant strength values that are higher than those of the pure epoxy
eect on the aspect ratio of GNP, the percolation threshold matrix. It is thought that the enhanced interfacial adhesion
too is expected to remain unchanged. due to the presence of Br results also in improved electrical
Conversely, the bromination treatment is expected to conductivity by eliminating unbonded gaps or micro-voids
have a significant eect on the conductivity values beyond that are highly resistant to electrical conduction and are
the percolation threshold. It is seen that whereas the con- detrimental to the formation of a conductive network by
ductivity values of the composites below the percolation GNPs. In particular, when the GNP content is close to
threshold are insensitive to the bromination treatment of the percolation threshold, there are only a few continuous
GNP, those of the composites with more than 1 wt% of conduction pathways throughout the bulk composite, as
GNP are enhanced significantly after bromination. The shown in the corresponding schematic drawing in Fig. 10.
maximum eect is noted for the 2.0 wt% GNP composites Any unbonded interface or micro-voids that tend to gather
for which two orders of magnitude increase is recorded near the reinforcements may block the electron hopping
for the brominated GNP composite. A combination of sev- and break up the conducting pathway, causing a significant
eral mechanisms is responsible for these observations as fol- reduction in the overall electrical conductivity. Fig. 10 also
lows. Below the percolation threshold, where GNP
conducting networks in the polymer matrix do not exist 140 Without Br treatment
yet (Fig. 10), the electrical conductivity of the composite With Br treatment for 3h
Flexural strength (MPa)

is controlled mainly by the conductivity of the matrix itself. 120

With the formation of GNP conducting networks after the
percolation threshold, the electrical conductivity of com- 100
posites is more controlled by the corresponding properties
of the filler; hence, the improved electrical conductivity of 80
GNP aects the electrical conductivity of the composites.
Even beyond percolation, the conductive fillers are still 60
wrapped by a thin layer of polymer matrix if they are well 0 1 2 3
dispersed. Electrons hop between adjacent fillers by a tun- GNP content (wt%)
neling mechanism, so that the electrical conductivity is Fig. 11. Flexural strength of GNP/epoxy composites with and without Br
aected also by fillermatrix interactions. The presence of treatment as a function of GNP content.
750 J. Li et al. / Carbon 45 (2007) 744750

suggests, however, that the electrical conductivity became 4. Bromination did not change the percolation threshold of
rather insensitive to bromination when the GNP content composites. The ameliorating eect of bromination on
was increased to 4 wt% GNP or above. It is envisaged that electrical conductivity of composite became more signif-
at high GNP contents far above the percolation threshold, icant when the GNP content was higher than the perco-
there are many conduction pathways that crisscross the lation threshold.
bulk composite, and the interfacial adhesion becomes
rather unimportant. In this case, the dierence in electrical
conductivity between the composites containing GNP with References
and without bromination becomes negligible, only by 2.3
times in absolute terms, compared to 122 times for GNP [1] Chen G, Wu C, Weng W, Wu D, Yan W. Preparation of polystyrene/
content of 2 wt%. graphite nanosheet composite. Polymer 2003;44:17814.
[2] Li J, Kim JK, Sham ML, Marom G. Morphology and properties of
UV/ozone treated graphite nanoplatelet/epoxy nanocomposites.
4. Conclusion Comp Sci Techol 2007;67:296305.
[3] Jia W, Tchoudakov R, Narkis M, Siegmann A. Performance of
The GNPs were treated by vapor-phase bromination at expanded graphite and expanded milled-graphite fillers in thermoset-
room temperature for dierent durations. Weight increase ting resins. Polym Compos 2005:52633.
[4] Enoki T, Suzuki M, Endo M. Graphite intercalated compounds and
was closely monitored, and the XPS, XRD and ToF-SIMS applications. Oxford: Oxford University Press; 2003. p. 388394.
were employed to evaluate the intercalation of Br atoms [5] Klimenko IV, Zhuravleva TS, Geskin VM, Jawhary T. Study of the
and quantitatively analyze the bonding types between C bromination of pitch-based carbon fibres. Mater Chem Phys
and Br. Electrical conductivities of the brominated GNPs 1998;56:1420.
[6] Mathur RB, Bahl OP, Kannan A, Flandrois A, Marchand A, Gupta
and the epoxy composites were measured and compared
V. In situ electrical resistivity changes during bromine intercalation in
between those with and without bromination. Major find- carbon fibres. Carbon 1996;34(10):121520.
ings are highlighted as following: [7] Lee RS, Kim HJ, Fischer JE, Thess A, Smalley RE. Conductivity
enhancement in single-walled carbon nanotube bundles doped with K
1. Systematic increases in weight and atomic concentration and Br. Nature 1997;388:2557.
of Br after bromination indicate gradual bromine [8] Prasad BLV, Sato H, Enoki T. Intercalated nanographite: structure
and electronic properties. Phys Rev B 2001;64:235407-1235407-10.
uptake. The intercalation of Br between the graphene [9] Li J, Kim JK, Sham ML. Conductive graphite nanoplatelet/epoxy
layers of GNP was further confirmed by the XRD nanocomposites: eects of exfoliation and UV/ozone treatment of
result, showing an increase in d-spacing from 3.342 A graphite. Scripta Mater 2005;53:23540.
to 3.361 A. [10] Weng LT, Ng KM, Cheung ZL, Lei Y, Chan CM. Quantitative
2. Two types of bonds were introduced simultaneously analysis of styrene-pentafluorostyrene random copolymers by ToF-
SIMS and XPS. Surf Interf Anal 2006;38(1):3243.
between C and Br, namely ionic and covalent bonds, [11] Nysten B, Issi JP, Barton R, Boyington DR, Lavin JG. Determina-
both of which increased with increasing bromination tion of lattice defects in carbon fibres by means of thermal-
duration. conductivity measurements. Phys Rev B 1991;44(5):21428.
3. The fraction of ionic bond reached the highest value [12] Beamson G, Briggs D. High resolution XPS of organic polymers: the
after 3 h bromination, which corresponded to the high- scienta ESCA300 database. Chichester, UK: Wiley; 1992. pp. 289.
est electrical conductivity of GNP with a 52.9%
improvement of electrical conductivity compared to
the untreated GNP.