The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones
-
Upload
yasser-ahmad -
Category
Documents
-
view
213 -
download
0
Transcript of The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones
![Page 1: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/1.jpg)
C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
.sc ienced i rec t .com
Avai lab le a t wwwjournal homepage: www.elsevier .com/ locate /carbon
The synthesis of multilayer graphene materialsby the fluorination of carbon nanodiscs/nanocones
Yasser Ahmad a,b, Elodie Disa a,b, Marc Dubois a,b,*, Katia Guerin a,b, Vincent Dubois a,b,Wei Zhang a,b,c, Pierre Bonnet a,b, Francis Masin d, Loıc Vidal e, Dimitri A. Ivanov e,Andre Hamwi a,b
a Clermont Universite, Universite Blaise Pascal, Institut de Chimie de Clermont-Ferrand, BP 10448, F-63000 Clermont-Ferrand, Franceb CNRS, UMR 6296, ICCF, BP 80026, F-63171 Aubiere, Francec Ecole Centrale de Pekin, Beijing University of Aeronautics and Astronautics (BUAA), Road 37, HaiDian District, Beijing 100191, Chinad Matiere Condensee et Resonance Magnetique, Universite Libre de Bruxelles (U.L.B.), CP 223, Boulevard du Triomphe,
B-1050 Bruxelles, Belgiume Institut de Sciences des Materiaux de Mulhouse, 15, rue Jean Starcky, BP 2488, 68057 Mulhouse Cedex, France
A R T I C L E I N F O
Article history:
Received 9 February 2012
Accepted 9 April 2012
Available online 13 April 2012
0008-6223/$ - see front matter � 2012 Elsevihttp://dx.doi.org/10.1016/j.carbon.2012.04.034
* Corresponding author at: Clermont UniverClermont-Ferrand, France. Fax: +33 4 73 40 7
E-mail address: marc.dubois@univ-bpcler
A B S T R A C T
Multilayer carbonaceous nanomaterial has been synthesized using a two-step process: car-
bon nanodiscs/nanocones were fluorinated using either the direct reaction with pure F2 gas
or the thermal decomposition of solid fluorinating agent (TbF4). Then the fluorinated parts
were removed by treatment at 600 �C in air. When the fluorine atoms are homogenously
dispersed, using fluorination by TbF4, thinning due to thermal defluorination results in
multilayer materials with 7–10 nm of thickness and 400–500 nm of width. Such resulting
materials and the fluorinated precursors have been characterized by solid state NMR,
TGA, XRD, SEM, TEM, AFM and Raman spectroscopy.
� 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Graphene, an infinite two-dimensional layer consisting of sp2
hybridized carbon atoms, has been attracting considerable
interest in recent years because of its unique band structure
and physical properties. Among them the extremely high car-
rier mobility makes it a promising material for applications in
the semiconductor industry, for composite materials, molec-
ular gas sensors, energy storage and conversion [1–3]. Never-
theless, graphene sheets are still difficult to fabricate. Several
methods are reported: (i) the micromechanical cleavage tech-
nique, graphene of high quality can be obtained but only in
minute quantities. Novoselov et al. [4] were able to isolate
graphene sheets using a simple micromechanical exfoliation
method placing a scotch-tape over commercial highly ordered
er Ltd. All rights reserved
site, Universite Blaise Pa1 08.mont.fr (M. Dubois).
pyrolytic graphite (HOPG). The latter technique allowed scien-
tists all over the world to isolate single and double-layered
graphene flakes (SLG and DLG). (ii) Intercalation–exfoliation
of graphite via thermal shock, acid treatment and intercala-
tion of graphite results in a few monolayer sheets (1%) and
mainly three-layer sheets. Exfoliation of natural graphite
using a liquid phase method has also been reported [5]. (iii)
‘Oxidation–exfoliation’ processes involve the synthesis of
graphite oxide (GO) and its subsequent reduction by exfolia-
tion at high temperature (thermal exfoliation), under vacuum
(vacuum exfoliation), or via chemical treatments (chemical
exfoliation) [6–8]. In particular, hydrogen induced reduction–
exfoliation of GO at a low temperature of 200 �C is reported
[9]. (iv) Heat treatments were also used, e.g. at 1600 �C of
diamond nanoparticles deposited on HOPG [10]. (v) Other
.
scal, Institut de Chimie de Clermont-Ferrand, BP 10448, F-63000
![Page 2: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/2.jpg)
3898 C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
methods to obtain graphene flakes were reported including
its epitaxial growth over SiC substrates [11], and chemical va-
pour deposition (CVD) growth over thin metal layers [12,13].
Moreover, there is a wide spectrum of methods available to
produce carbon nanoribbons (CNRs), from CVD, through
chemical treatments of graphite to the unzipping of carbon
nanotubes: (i) intercalation-exfoliation of MWCNT, involving
treatments in liquid NH3 and lithium, and subsequent exfoli-
ation using HCl and heat treatments [14], (ii) chemical routes,
involving acid reactions that start to break carbon–carbon
bonds (e.g. H2SO4 and KMnO4 as oxidizing agents) [15], (iii)
catalytic approach, in which metal nanoparticles ‘‘cut’’ the
nanotube longitudinally like a pair of scissors [16], (iv) the
electrical method, by passing electric current through a nano-
tube [17], (v) physicochemical method by embedding the
tubes in a polymer matrix followed by Ar plasma treatment
[18], and (vi) reducing the fluorinated parts of graphene using
electron beam irradiation [19]. The resulting structures are
either CNRs or graphene sheets.
Our two-step strategy for producing multilayer graphene
materials consists first in the synthesis of sub-fluorinated
graphitized nanocarbons and then the thermal defluorination
in the air atmosphere. Sub-fluorination means that some car-
bon atoms are not fluorinated and the value x in the global
chemical composition CFx is lower than unity. A recent study
on the thermal stability of fluorinated carbon nanofibres
(CNF) [20] underlines that the fluorine atoms are thermally re-
moved from the outer tubes by a mechanism similar to a peel-
ing which preserves the non-fluorinated core. On the
contrary, if the fluorine atoms are homogenously dispersed
in the whole volume of the fibres, the gas evolution during
the thermal defluorination leads to the destructuration of
the carbonaceous matrix. As a matter of fact, the decomposi-
tion reaction of (CF)n graphite fluoride is described by the fol-
lowing reaction [21]:
2ðCFÞn ! nCþ XCF2 þ YC2F4 ð1Þ
where X + 2Y = n. The evolution of gaseous fluorocarbon spe-
cies such as CF2 and/or C2F4 during the decomposition of CxF
[22,23] and (CF)n [21–24] must be accompanied by a C–C bond
cleavage on the carbon sheets in the compounds. The decom-
position temperatures of the (CF)n samples prepared from var-
ious carbon materials is ranged from 320 to 610 �C; the higher
the graphitization degree of the carbonaceous precursor, the
higher the thermal stability, [25,26]. Moreover, in spite of sim-
ilar CFx chemical composition and C–F bonding, carbon nano-
fibres fluorinated using different ways, direct fluorination
with F2 or controlled process using solid fluorinating agent
TbF4, exhibit different thermal stabilities and different prod-
ucts after heating. Indeed, during the controlled fluorination,
the reactive species is mainly atomic fluorine FÆ and its diffu-
sion rate and reactivity allow a homogenous dispersion of the
fluorine atoms. The method using the thermal decomposition
of solid fluorinating agent (TbF4) is called controlled fluorina-
tion. Using a flux of F2 gas, first the outer tubes react and the
fluorination progresses then towards the core when the reac-
tion temperature increases. In the present study, both pro-
cesses were carried out and evaluated for the synthesis of
graphene materials. With this aim, our choice for the carbo-
naceous precursor was the graphitized nanocones/discs.
The commercially available product for nanocones/discs, pro-
vided by N-Tec Corporation, consists of a mixture of 70 weight
percent (wt.%) of nanodiscs (CND), 20 wt.% of nanocones and
10 wt.% of amorphous carbons. Due to the content of nano-
discs, the mixture will be called CND. In the present study,
fluorinated carbon nanocones/discs have been prepared from
CND post treated at 2700 �C under argon in order to increase
the graphitization degree because the order can be partly
maintained upon fluorination, although this process obliga-
tory results in a progressive disorganization of the carbona-
ceous matrix due to the accommodation of the fluorine
atoms and the change of carbon hybridization from sp2 to
sp3. Moreover, the more remarkable benefits of the use of
graphitized sample as starting material are, on one hand, that
cracks and irregularities of both cones and discs are avoided
contrary to as-prepared CND after fluorination and, in the
other hand, structural defects such as CF2 and CF3 and dan-
gling bonds, are significantly reduced [27]. Those defects act
on the thermal stability of the fluorinated CND.
Our wet chemistry method avoids the use of liquid-phase
extraction contrary to a recent work [26], for which the exfo-
liation of commercial CF�0.5 graphite fluoride in dimethyl-
formamide has been first carried out; the liquid-phase
extraction of graphene fluoride was then necessary. Finally,
a subsequent reduction of graphene fluoride with triethylsi-
lane or zinc particles results in graphene.
Carbon nanodisc/nanocones treated by direct fluorination
and by controlled fluorination using TbF4 as fluorinating
agent will be firstly compared using complementary tech-
niques including X-ray diffraction (XRD), scanning and trans-
mission electron microscopies (SEM and TEM), atomic force
microscopy (AFM), Raman spectroscopy, and 19F and 13C high
resolution nuclear magnetic resonance (NMR). These tech-
niques allow to investigate at various scales both the fluori-
nated parts (XRD, 19F and 13C NMR) and the non-fluorinated
carbons (13C NMR, Raman and XRD). The differences will be
discussed and used to explain the thermal defluorination
and the different final products, among which multilayer car-
bonaceous nanomaterials have been obtained.
2. Materials and methods
2.1. Synthesis and chemical composition of the fluorinatedCND
The carbon nanocones and nanodiscs were produced by pyro-
lysis of heavy oil using the Kvaerner carbon black and hydro-
gen process (CBH) [28]. The CBH is an emission-free industrial
process that decomposes hydrocarbons directly into carbon
and H2, based on a specially designed plasma torch, with a
plasma temperature above 2000 �C. The solid output was
found to consist of a significant amount of opened carbon
nanocones (20 wt.%), as well as a large number of flat carbon
discs (70 wt.%), and the rest being carbon black. Some impu-
rities are also present such as Fe (257 ppm), Si (85 ppm), Ca
(65 ppm) and Al (34 ppm) [29]. The size of the crystalline
domains, i.e. the coherence length along the c-axis (Lc) deter-
mined by XRD, is around 2 nm which means that the as-
product exhibits low degree of order. An annealing treatment
![Page 3: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/3.jpg)
C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8 3899
under argon at 2700 �C was then performed (the resulting
sample is denoted CND). It improves the crystallinity since
Lc reaches a much higher value of 39 nm.
Scanning electronic microscopy (SEM) images (Fig. 1a
and b) underline, firstly, the size dispersion of nanodiscs
and nanocones for the raw material. In order to ‘‘quantify’’
this dispersion, the frequency of occurrence was estimated
as a function of the diameter for discs and cones. Compared
to the data provided by N-Tec, no difference was observed be-
tween the as-prepared and graphitized samples indicating
that the post-treatment at 2700 �C did not significantly
change the geometry. Whatever the considered parameters,
the dispersion is rather large with a maximum giving the
average diameter centred at 1.5 lm for nanodiscs and at
1.0 lm for nanocones. Moreover, the same average thickness
of 35 nm is found for both cones and discs.
Fluorination was carried out with pure fluorine gas flow in
a Monel reactor. The reaction temperatures TF were ranged
between 480 and 520 �C. The duration of each experiment
was of 3 h. The resulting samples are denoted D�TF.
Terbium tetrafluoride was used as fluorinating agent. It
was obtained from TbF3 (Aldrich, 99.9%) in pure F2 gas at
500 �C. Its purity (i.e. the absence of residual TbF3) was sys-
tematically checked by X-ray diffraction. The thermogravi-
metric analysis (TGA) of TbF4, realized until 350 �C,
indicated that exactly one mole of FÆ was released per mole
of TbF4 between 100 and 300 �C. This experiment was carried
out under argon flow in order to avoid the formation of
oxyfluorides.
For the CND fluorination by TbF4, a nickel closed reactor
was used in order to preserve the defined fluorine amount
(atomic and/or molecular) released by the thermal decompo-
sition of TbF4. A two-temperature oven was used: the part
containing TbF4 was heated at 450 �C whatever the
Fig. 1 – SEM images of as-prepared CND (a) and treated at
2700 �C (b).
experiment whereas CND were heated at temperatures TF of
500 and 550 �C.
A reaction time of 16 h was used. Prior to the heating, a
primary vacuum (�10�2 atm) was applied into the reactor.
The resulting samples are denoted C�TF, C means controlled
fluorination. The reactions involved during the fluorination
are the following ones:
TbF4!D
TbF3 þ F� or12
F2 ð2Þ
Cþ xF� orx2
F2 ! CFx ð3Þ
The total conversion of TbF4 into TbF3 was systematically
checked by both the weight loss and XRD.
The fluorination level ‘x’ (i.e. F:C molar ratio) of nanocar-
bons fluorinated by F2 (denoted D�TF) was determined by
weight uptake as for C�TF samples.
2.2. Physicochemical characterizations
SEM micrographs were recorded using a Cambridge Scan
360 SEM operating at 1 kV.
The different samples were characterized by Transmission
Electron Microscopy (TEM, FEI CM200 operating at 200 kV).
The nanomaterials were dispersed in chloroform using ultra-
sonic treatment and a few drops of suspension were depos-
ited onto copper observation grids covered with ultrathin
carbon/formvar films. The grids were subsequently dried at
ambient conditions.
The quantitative analysis of TEM images was performed in
reciprocal space. The details of the method can be found else-
where [30]. The two-dimensional power spectral density
function (P2(s)) was computed from TEM images (u(r)) up to
the critical, or Nyquist, frequency depending upon the exper-
imental sampling interval as:
P2 �1A
ZA
uðrÞwðrÞexpð2pisrÞd2r
�������� ð4Þ
where A denotes the image area, W(r) window function [31]
and s the 2D reciprocal space vector. The P2(s) function was
then transformed into the one-dimensional PSD (P1(s)), where
s stands for the norm of s, according to:
p1ðsÞ ¼ ð2psÞ�1Z
P2ðS0djS0j � sÞdS0 ð5Þ
The samples have been also investigated by AFM (Dimen-
sion V from Veeco) in tapping mode. CND have been dis-
persed in acetonitrile and sonicated for few minutes. Then,
they have been drop cast on freshly clived mica surfaces.
X-ray diffractogramms were obtained using a PHILIPS dif-
fractometer with a Cu(Ka) radiation (k = 1.5406 A).
NMR experiments were carried out with Bruker Avance
spectrometer, with working frequencies for 13C and 19F of
73.4 and 282.2 MHz, respectively. A Magic Angle Spinning
probe (Bruker) operating with 2.5 mm rotors was used. For
MAS spectra, a simple sequence was performed with a single
p/2 pulse length of 4.0 and 3.5 ls for 19F and 13C, respectively.13C chemical shifts were externally referenced to tetramethyl-
silane (TMS). 19F chemical shifts were referenced with respect
to CFCl3.
![Page 4: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/4.jpg)
3900 C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
Raman spectra were recorded at room temperature using a
Jobin Yvon T64000 with a charge coupled device multichannel
detector. The radiation source was a 514.5 nm Argon laser
line. The laser power was tuned to 10 mW.
2.3. Thermal treatment
In order to select the most efficient thermal treatment for the
synthesis of multilayer graphene, thermogravimetric analysis
(TGA) experiments were carried out on Shimadzu TGA-50.
Measurements were made in air at a heating rate of
2 �C min�1 from room temperature to 600 �C on about 10 mg
of sample (accuracy of 0.001 mg). The same conditions were
used for each sample. Different parameters are extracted
from the TGA curve: (i) T10 is the temperature determined
from the TGA curve for 10% weight loss, (ii) TC–F is determined
from the derivative TGA curve and corresponds to the decom-
position temperature of the C–F bond (defluorination of the
fluorinated CND) and (iii) TC represents the temperature of
the oxidation of the non-fluorinated carbon parts, i.e. burning
in air atmosphere.
Fig. 2 – SEM images of C-500 (a), C-550 (b) and D-500 (c). The
fluorine contents are 0.70, 0.95 and 0.96, respectively.
3. Results and discussion
3.1. Fluorination
The fluorine contents x in CFx, after fluorination, obtained by
weight uptake, are 0.70, 0.95, 0.78 and 0.96 for C-500, C-550, D-
480 and D-500, respectively. The fluorination level increases
with TF. Among the different samples, the highly fluorinated
CND were obtained at 550 �C using TbF4 and 500 �C using F2
with a corresponding fluorination level of 0.95 and 0.96,
respectively. Increase of the reaction temperature to 520 �Cduring the direct fluorination results in partial exfoliation of
the sample, which is already visible by the macroscopic vol-
ume expansion and confirmed at the nanoscale by SEM.
The volume expansion due to hyperfluorination and gas evo-
lution of volatile species such as CF4, and C2F6 results in
cracks and swelling on the sheet edges. Any further investiga-
tion was carried out with this damaged sample. Nevertheless,
this observation underlines the efficient reaction temperature
range for the direct fluorination, up to 500 �C. Similar phe-
nomenon has been found for fluorination of nanofibres at
temperature higher than 480 �C.
Scanning electron microscopy (SEM) images (Fig. 2) under-
line the maintaining of the geometry for both direct and con-
trol fluorinations, whatever the reaction temperature.
Nevertheless, the disc and cone edges are more irregular in
the case of D-500. Moreover, a nanocarbon swelling occurred
due to the accommodation of the fluorine atoms. This swell-
ing differs after the direct and controlled processes as re-
vealed by atomic force microscopy.
AFM was performed in tapping mode. Topography and er-
ror images of the raw nanodiscs (D-500) and fluorinated nano-
discs thermally treated at 2700 �C (C-550) show a clear
difference in the surface morphology (Fig. 3). The accommo-
dation of the fluorine atoms occurs by a swelling, which is
more pronounced on the edges for D-500 (Fig. 3a); the maxi-
mum extent of swelling is observed in the overall perimeter.
The edge seems then to be the preferential area for the direct
fluorination. On the contrary, the swelling is homogenous for
C-550 and the accommodation of fluorine atoms results in
small cracks homogenously dispersed on the disc surface
(Fig. 3c).
The fluorination progressively changes the structure as a
function of the fluorination temperature (TF). The XRD pat-
terns of the fluorinated CND are compared in Fig. 4. The pris-
tine CND pattern is similar to graphite. The main peaks
correspond to the graphite structure (002), (100), (101) and
(004) diffraction lines for 2h values of 26.4� (interlayer dis-
tance d = 0.337 nm), 42.9� (0.211 nm), 44.8� (0.202 nm) and
54.4� (0.169 nm). CND exhibit a high structural organization
thanks to the post treatment at 2700 �C (coherence length
![Page 5: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/5.jpg)
Fig. 3 – AFM images of D-500 (a), and C-500 (b and c).
Fig. 4 – XRD diagrams of C-500, C-550, D-480 and D-500.
C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8 3901
along the c axis, LC = 39 and 2 nm for post-treated and raw
CND, respectively). Considering that the value of 39 nm is
close to the average thickness of the graphitized sample, the
rebaked discs and cones can be considered to be nearly sin-
gle-crystal-like in the c-direction. For C–TF a new phase ap-
pears with corresponding peaks at 2h values centred at 14.8�and 40.9� attributed to the (001) and (100) peaks of a fluoro-
graphite matrix in an hexagonal system. The corresponding
interlayer spacing of 0.60 nm is characteristic of (CF)n-type
structure. On the contrary, an intermediate phase is deduced
for D-480 and D-500. Because of the coexistence of (C2F)n and
(CF)n phases, the value d equal to 0.68 nm is measured instead
of 0.81–0.82 nm and 0.6 nm for (C2F)n and (CF)n structural
types, respectively. Moreover, because of this coexistence of
phases, the full width at half maximum is larger for D-480
and D-500 (Dh = 3.7�) than for C-500 and C-550 (Dh = 2.7�).In order to complete the investigation on the non-fluori-
nated carbon atoms and their content in the samples, 19F
MAS NMR operating at high spinning rates of 15.0 and
30.0 kHz was carried out; the spectra are shown in Fig. 5. This
technique can highlight the nature of the interaction between
carbon and fluorine atoms, i.e. the C–F bonding. The compar-
ison of samples with similar fluorination level obtained by di-
rect and controlled fluorination using TbF4 indicates different
fluorination mechanisms. The spectra of covalent graphite
fluorides with (C2F)n and (CF)n structural types are added to
this figure for comparison. Whatever the sample, fluorinated
C-500, C-550, D-480 and D-500 or graphite fluorides, the main
line at �190 ppm (Fig. 5a and b) is assigned to covalent C–F
bonds [32–36]. Thanks to high spinning rate, a second line
at �175 ppm can be observed, present only for D-500 and
(C2F)n type graphite fluoride (Fig. 5b). This line can be unam-
biguously assigned to fluorine nuclei present only in the
(C2F)n structure: 19F in weak interaction with non-fluorinated
sp3 carbon atoms. The structure of (C2F)n can be described
using a F/C/C/F stacking sequence for fluorocarbon sheets
whereas for (CF) the sequence is F/C/F/F/C/F. All carbon atoms
exhibit a sp3 hybridization. Half of the carbon atoms are fluo-
rinated in (C2F)n while the others being linked to carbon
atoms. By analogy with the hyperconjugation effect in fluo-
rine-graphite intercalation compounds (F-GICs), that involves
C–C in non-fluorinated parts and C–F bonds [37], the cova-
lence of the C–F bonds located close to these diamond-like
atoms could be weakened in the same manner. The 19F chem-
ical shift is then of �175 ppm instead of �190 ppm for pure
covalent C–F. In the case of F-GICs, �170 ppm is recorded.
The line at �175 ppm, which is observed thanks to a high
![Page 6: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/6.jpg)
Table 1 – Key parameters deduced from TGA curves.
Sample F/C T10 (�C)±5 �C
TC–F (�C)±5 �C
C-500 0.70 432 521C-550 0.95 517 569D-480 0.78 460 520D-500 0.96 430 520
0 -50 -100 -150 -200 -250 -300 -350
D-500
D-480
C-550
C-500
C-F
CF2
*
*
*
*
*
*
*
*
*
*
* *
δ19F/ CFCl3 (ppm)
CF3
(a)
150 100 50 0
C-C-F sp3 C
δ13C / TMS (ppm)
(c)C-F
CF2
C-C-F
sp2 C D-500
D-480
C-500
C-550
(b)
Fig. 5 – 19F MAS NMR spectra of fluorinated CND with 14 kHz
(a) and 30 kHz (b) spinning speed ( * are the spinning
sidebands). 13C MAS spectra (10 kHz) (c).
3902 C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
spinning rate, can be considered as an indicator of the pres-
ence of (C2F)n structural type. (C2F)n type is maintained in a
wide temperature range for the direct fluorination whereas
(CF)n type is formed during the controlled fluorination using
TbF4 whatever the temperature. This is due both to progres-
sive generation of atomic fluorine by thermal decomposition
of TbF4 and to its high reactivity and diffusion.
The 19F lines close to �80 and �120 ppm are assigned
respectively to CF3 and CF2 groups, their intensities are higher
for D-480, D-500, (C2F)n and (CF)n prepared with F2 gas at high
temperature (380 and 600 �C for (C2F)n and (CF)n, respectively).
Indeed, the control of fluorination kinetic avoids the hyperflu-
orination at high temperature. The generated atomic fluorine
penetrates towards the core without formation of important
structural defects. Moreover, excess of fluorine is avoided
contrary to the case of a F2 gas flux. This results in a decrease
of structural defects such as CF2, CF3 and dangling bonds and
could change the thermal stability. These defects are always
well correlated with structural disorder [38–41].13C NMR confirms the structural phase formed during fluo-
rination. As a matter of fact, in the F/C/C/F stacking sequence
of (C2F)n, half of the carbon atoms are fluorinated. The other
atoms, also with sp3 hybridization, are linked only to other
carbon atoms, as in a diamond phase. The typical 13C NMR
line of those 13C nuclei is observed at 42 ppm and appears
as a good indicator of the presence of (C2F)n phase. This line
is observed only for D-480 and, with lower intensity, for D-
500, in accordance with both 19F NMR and XRD. A common
line whatever the synthesis way appears at 88 ppm and is as-
signed to covalent C–F bonds.
The high relative intensity of the line at 110 ppm related to
CF2 groups also confirms the large amount of those groups in
D-500 and D-480. Finally, for the lower fluorination tempera-
ture using both F2 and TbF4, an additional line is measured
at 141 ppm. Such a chemical shift, which differs from the
one of pure graphite at 120 ppm, is explained by the interac-
tion of non-fluorinated carbon atoms with neighbouring C–F
bonds (the corresponding atoms are denoted C–C–F on
Fig. 5c). The C–F bonding is weakened due to this interaction.
The fluorination using TbF4 decomposition generates the
formation of a unique highly fluorinated phase, i.e. (CF)n type,
whatever the fluorination temperature, the C–F bonding being
covalent. This fluorinated phase is always mixed with some
residual non-fluorinated CND and a high fluorination level
can be obtained (x = 0.95) without hyperfluorination contrary
to the F2 process. The higher homogeneity of the materials
obtained using this controlled process by comparison with
the direct fluorination using fluorine gas can be explained
by the low kinetic of decomposition of TbF4 that allows a con-
tinuous addition of fluorine to carbon matrix and a more pro-
gressive fluorination.
3.2. Thermal treatment
Fig. 6 displays the TGA curves for samples fluorinated by F2
and TbF4. Table 1 summarizes the key values to evaluate the
thermal stability, namely T10 and TC–F. Whatever the fluorina-
tion way, the TGA curves may exhibit two weight losses as-
signed to the defluorination at lower temperature and to
oxidation of non-fluorinated carbon atoms with dioxygen
from air (burning) at higher temperature. The derivative
curves clearly underline one main process by a well-defined
peak (insert on Fig. 6 for C-550 as a representative example).
The temperature of burning in air (TC) is higher than 600 �C
![Page 7: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/7.jpg)
Fig. 6 – Evolution of the residual weight of CND fluorinated
by TbF4 decomposition (straight lines for C-500 and C-550)
and by F2 ( d for D-480 and s for D-500) versus the
temperature (air atmosphere, 2 �C min�1); insert: derivative
curve of TGA curve of C-550 sample showing TC–F.
C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8 3903
for all the samples. The derivative TGA curve does not exhibit
a second peak related to TC because the experiment is con-
ducted up to 600 �C (Fig. 6). TC–F does not depend on the fluo-
rination temperature using F2, being equal to 520 ± 5 �C. The
sample fluorinated by controlled fluorination exhibits an en-
hanced thermal stability when compared to the ones ob-
Fig. 7 – SEM images after thermal treatment at 600 �C of C-550 (a–
Image of exfoliated D-520 is added for comparison (h).
tained by molecular fluorine F2. The physico-chemical
characterization underlines that the controlled process using
TbF4 results in more homogeneous fluorine atom distribution
than the direct fluorination and leads to lower structural
defects such as CF3, CF2 groups and dangling bonds (always
related to disorder). Such disorder favours the thermal
decomposition. The higher defluorination temperature TC–F
was then measured for C-550 with a value of 569 ± 5 �C(Table 1).
The thermal defluorination occurs at lower temperatures
for C-500 (TC–F = 521 ± 5 �C) than for C-550 (TC–F = 569 ± 5 �C).
An additional effect must be considered. The presence of
weakened C–F bonds has been underlined by 13C NMR and
their thermal defluorination must take place at lower temper-
ature because of the weakening of the C–F bonding. Such pro-
cess could destabilize the fluorocarbon matrix and decrease
both TC–F and T10. As a consequence of the presence of weak-
ened covalent C–F, the temperatures T10 for 10% weight loss
are 432 ± 5 �C for C-500, lower than 517 ± 5 �C for C-550. Similar
phenomenon has been found for fluorinated nanofibres [20].
With the aim to produce multilayer carbonaceous nanom-
aterials, the treatment temperature has been selected taking
into account both the TGA derivative curves and TC–F and TC
values. Both C-550, and D-500 were thermally defluorinated
at 600 �C. It is to note that the residual weight% is calculated
equal to 12n/[2n(12 + 19)] � 20% taking into account reaction
(1); such value is in good accordance with the TGA data for
these samples. For comparison, similar treatment was per-
formed with non-fluorinated nanodiscs/nanocones. The
c), D-500 (d–f) and non-fluorinated nanodiscs/nanocones (g).
![Page 8: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/8.jpg)
Fig. 8 – TEM images of C-550 (left) and D-500 (right) heated at 600 �C in air. The magnifications are 5000, 50 000, 115 000 and
310 000.
3904 C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
resulting samples have been characterized by Raman spec-
troscopy, SEM and TEM.
Discs and cones are still observed by SEM (Fig. 7 d–f) after
the thermal treatment of sample fluorinated by F2. Neverthe-
less, their thickness is irregular because of the thermal deflu-
orination. The sheet edges appear thinner on the discs
(Fig. 7e) and the cones (Fig. 7f). On the contrary, most of the
cones and discs disappear during the defluorination of
C-550 (Fig. 7a–c). The resulting particles of about 400–500 nm
of width and 7–10 nm of thickness seem like flakes folded
up on themselves.
When the non-fluorinated nanodiscs/nanocones were
treated at 600 �C in air, strong damages were generated
(Fig. 7g): the edges are irregular on both discs and cones (see
arrows on Fig. 7g) and pseudo-hexagonal holes are observed
(insert). Another route to produce carbon multilayers could
have been exfoliation during a fast increase of the tempera-
ture up to 520 �C in F2 atmosphere. But the resulting sample
(D-520 in the same notation than before) exhibits inhomoge-
neity in sizes, both in thickness and width (Fig. 7h). The flakes
have also a curvature. This method has not been retained.
Contrary to C-550 heated at 600 �C, numerous discs are still
observed by TEM for CND fluorinated by F2 and thermally
defluorinated. In addition, smaller particles are present. Their
shapes are similar to those of the defluorinated C-550. A high-
er magnification, the particles reveal their substructure
![Page 9: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/9.jpg)
Fig. 9 – TEM images of heat-treated C-550 (a) and D-500 (b) showing the presence of graphitic layers and corresponding power
spectral density (PSD) functions (c). The curves are offset vertically for clarity. The dotted line indicates the periodicity of
graphitic layers. One-dimensional correlation functions computed from the TEM images (d).
C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8 3905
consisting of multilayers folded up on themselves. As the ini-
tial shape is the disc; the thinning occurring during the ther-
mal defluorination results in the discs folding up. This
process systematically occurs for C-550 during the defluorina-
tion and punctually for D-500 resulting in the smaller parti-
cles. The thickness of the edges is ranged in between 7 and
10 nm corresponding to about 20–30 layers (Fig. 9).
Magnification of TEM bright-field images of the heat-trea-
ted fluorinated CND underlines the presence of graphitic lay-
ers (Fig. 9a and b). The periodicity of the layers is reflected by a
Bragg peak in the corresponding PSD curves (Fig. 9c), which is
positioned at about 0.35 nm for both samples. This value is in
accordance with non-fluorinated graphene layers. For fluori-
nated nanofibres, in addition to the usual graphene layer peri-
odicity the PSD function also displays a broad peak with a
maximum at about 1.5–2.0 nm�1. This additional feature in
the PSD curve was assigned to the presence of layers, which
were less ordered and are more spatially separated due to
the accommodation of fluorine atoms [42]. Importantly, the
additional line is absent for treated C-550 and D-500 in accor-
dance with a quasi-total defluorination.
The one-dimensional correlation function (CF) denoted
here as c (r) was computed as the real part of the Fourier
transform of the one-dimensional PSD function, P1(s), cor-
rected for the so-called sigmoidal-gradient transition zones
with thickness r:
cð1Þ ffi Ref2pZ 1
0
P1ðsÞs expð2pislÞexpð4p2r2s2Þds ð6Þ
For the sake of simplicity the CFs were normalized to unity
at the origin. The method is described in full detail in Ref. [30].
The CFs show that treated D-500 generates the first subsidiary
maximum corresponding to the nearest-neighbor distance,
which is located at a slightly smaller distance than the one
of the C-550 sample. Also, the subsidiary maximum itself is
somewhat broader for this sample (Fig. 9d). Therefore the
structure of treated D-500 reveals less order than the one of
heated C-550. Raman spectroscopy in the following section
will confirm this fact.
The Raman spectra of the treated samples are displayed
on Fig. 10. First, it is to note that the spectra of the fluorinated
CND cannot be recorded because of the strong fluorescence
![Page 10: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/10.jpg)
1000 1200 1400 1600 1800
CNDs
Heated C-550
Raman Shift (cm-1)
Heated D-500
Fig. 10 – Raman spectra of thermally treated C-550 and D-
500 and raw non-fluorinated CND.
3906 C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
typical of highly fluorinated carbons. The microRaman spec-
tra were recorded in numerous regions of the samples, heated
C-550 and D-500. Any difference was found whatever the
location. For heated C-550 and D-500, three narrow vibration
bands with full width at half maximum of about 20–40 cm�1
are visible: two of them around 1356 and 1620 cm�1 are as-
signed to the D and D 0 modes. For their activation D and D 0
resonances require a defect, such as bond dislocations,
Fig. 11 – Effect of thermal treatment of nanodisc; the graphitic c
after defluorination. Contrary to C-550, the thinning is incompl
fluorine atoms.
missing atoms at the edges of the sample and sp3-hybridized
carbon atoms; their presence is associated with an increased
degree of disorder [43–45]. The third band at 1589 cm�1 is as-
signed to the conventional G mode, which is related to the
graphitization degree of carbon material and is also called
tangential modes (where the carbons in sp2 hybridization vi-
brate in parallel to the axis of the nanotube) [46]. The value
measured for heated fluorinated samples differ from the ones
of the raw CND, for which only D and G are recorded at 1345
and 1565 cm�1. On the contrary, some similitudes appear with
fully fluorinated graphene; in those materials the D and D 0
peak appear at 1350 and 1620 cm�1 and their intensity is high-
er than that of the G band at 1580 cm�1. Contribution to the D
peak of fully fluorinated graphene comes from sp3-bonded
carbon atoms rather than the edges of the fully fluorinated
flakes. The fluorination/defluorination induces defect into
the resulting graphene sheets.
The relative integrated intensities of D and G bands ID/IGare 2.0 and 2.3 for heated C-550 and D-500, respectively, the
value for the raw CND being 0.06. Those ratio are in accor-
dance with TEM study that underlines the slightly lower dis-
order for heated C-550 than treated D-500. For comparison,
fully highly fluorinated graphene exhibits ID/IG of 3.8 [47].
3.3. Discussion
Using a freeze fracturing setup, Naess et al. [48] observed for
the same nanodiscs from N-Tec an in-plane cleavage that
was usually parallel to the disc surface. The particles were
ore (in white), which acts as a reinforcement, is maintained
ete for D-500 because of inhomogeneous distribution of the
![Page 11: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/11.jpg)
C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8 3907
fractured in a plane close to the geometrical centre of the
disc. The splitting occurs between the graphitic layers, which
are only bonded by weak van der Waals forces, and not in the
amorphous carbon surrounding the disc. The authors con-
cluded then that the as-synthesized discs can be described
as a graphitic core within an envelope of non-crystalline car-
bon. The graphitic core is first grown to its full size, where-
upon layers of amorphous carbon are deposited, the
thickness of the graphitic core being estimated to only
2–5 nm. Moreover, if the material is heat-treated at 2700 �C,
the graphitization takes place simultaneously in different
parts of the amorphous carbon. Thus the temperature-
induced conversion from amorphous carbon to layers of
graphite is not fully coupled geometrically to the original
graphite core. This history of the nanodiscs has important
consequence during the fluorination/defluorination. The level
of fluorination may differ between the core and the outer
parts and the difference is maintained after fluorination
(Fig. 11). Moreover, the core could act in a reinforcement ef-
fect. Regarding the thermal defluorination, a peeling of the
fluorinated layer occurs in a similar way than fluorinated
nanofibres. A thinning of the discs progressively occurs by
gaseous removal of the fluorinated parts. The final thickness
is close to 7–10 nm, not so far from the initial core size. Two
cases must be distinguished: When the fluorine atoms are ini-
tially homogenously dispersed in the whole outer volume
thanks to the controlled fluorination (TbF4), the thinning is
equal whatever the location along the diameter. Moreover,
the discs fold up itself when the thickness is about ten nm
and appear as ruffled paper (Figs. 7 and 8).
On the contrary, as irregular swelling takes place upon di-
rect fluorination (see AFM section), the thermal defluorina-
tion results in different thickness and the planar
configuration is partly maintained. The planarity is main-
tained because of the thicker parts.
The homogenous thinning of the discs cannot be per-
formed by a simple burning at 600 �C in air atmosphere be-
cause the resulting discs are damaged with holes and edge
cut down (Fig. 7g). Fluorination allows the decomposition
temperature to be decreased and the gap between defluorina-
tion and burning in air is sufficient to separate the two pro-
cesses (for C-550, TC–F is equal to 569 ± 5 �C whereas TC is
higher than 600 �C). The chemical thinning is then possible.
This discussion focuses on 70 wt.% of the sample, corre-
sponding to the disc percentage. 10% amorphous parts are
probably burned under fluorine atmosphere [49]. 20 wt.% of
the sample are nanocones. They are structurally similar to
flat carbon discs, which can be regarded as carbon cones with
apex angle equal to 180, or alternatively, with zero pentagon
at the tip. The change of the geometry of the cones is more
difficult to be observed than for discs using SEM and TEM
but we believe that the mechanisms are nearly similar due
to the structural similarities.
4. Conclusions
The fluorination of opened structures such as nanodiscs (or
nanocones) may be the first step for the production of multi-
layer carbonaceous nanomaterials. The fluorinated parts
must be homogenously dispersed in the carbon lattice with
a non-fluorinated core (sub-fluorination). A controlled
fluorination using solid fluorinating agent must be then pre-
ferred rather than the direct process using molecular fluorine
F2. Whatever the fluorination method, the removal of the
fluorinated parts during a process similar to a peeling with
CF4 and C2F6 gas evolution does not result in damages of
the non-fluorinated region when the structure is opened con-
trary to fluorinated nanofibres. The structure of those latter
consists of either concentric cylinders of graphite sheets
(Russian doll model) or of a single graphite sheet rolled in
around itself (parchment model). Gas evolution during the
thermal defluorination into this confined structure may cause
damages in the resulting material such as exfoliation and loss
of the tubular shape. Multilayer carbonaceous nanomaterials
were synthesized with 7–10 nm thickness in a form of ruffled
paper. The discs fold up on themselves during thinning. Fur-
ther studies are necessary to decrease the thickness by using
the controlled fluorination and by increasing the fluorine con-
tent closer to the limit of CF1. A few non-fluorinated layers
must be conserved in order to form thinner multilayers. The
present work is a promising first stage towards the synthesis
of multilayer graphene materials via fluorination.
Acknowledgements
The authors are thankful to Elodie Petit for her helpful contri-
bution for Raman spectra.
R E F E R E N C E S
[1] Chae HK, Siberio-Perez DY, Kim J, Go Y, Eddaoudi M, MatzgerAJ, et al. A route to high surface area, porosity and inclusionof large molecules in crystals. Nature 2004;427(6974):523–7.
[2] Stankovich S, Dikin DA, Dommett GHB, Kohlhaas KM, ZimneyEJ, Stach EA, et al. Graphene-based composite materials.Nature 2006;442(7100):282–6.
[3] Wang G, Wang B, Wang X, Park J, Dou S, Ahn H, et al. Sn/graphene nanocomposite with 3D architecture for enhancedreversible lithium storage in lithium ion batteries. J MaterChem 2009;19(44):8378–84.
[4] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y,Dubonos SV, et al. Electric field effect in atomically thincarbon films. Science 2004;306(5696):666–9.
[5] Hernandez Y, Nicolosi V, Lotya M, Blighe FM, Sun Z, De S,et al. High-yield production of graphene by liquid-phaseexfoliation of graphite. Nat Nano 2008;3(9):563–8.
[6] Lv W, Tang DM, He YB, You CH, Shi ZQ, Chen XC, et al. Low-temperature exfoliated graphenes: vacuum-promotedexfoliation and electrochemical energy storage. ACS Nano2009;3(11):3730–6.
[7] Schniepp HC, Li JL, McAllister MJ, Sai H, Herrera-Alonso M,Adamson DH, et al. Functionalized single graphene sheetsderived from splitting graphite oxide. J Phys Chem B2006;110(17):8535–9.
[8] Tung VC, Allen MJ, Yang Y, Kaner RB. High-throughputsolution processing of large-scale graphene. Nat Nano2009;4(1):25–9.
[9] Kaniyoor A, Baby TT, Ramaprabhu S. Graphene synthesis viahydrogen induced low temperature exfoliation of graphiteoxide. J Mater Chem 2010;20(39):8467–9.
![Page 12: The synthesis of multilayer graphene materials by the fluorination of carbon nanodiscs/nanocones](https://reader035.fdocuments.net/reader035/viewer/2022080401/575073021a28abdd2e8d4041/html5/thumbnails/12.jpg)
3908 C A R B O N 5 0 ( 2 0 1 2 ) 3 8 9 7 – 3 9 0 8
[10] Affoune AM, Prasad BLV, Sato H, Enoki T, Kaburagi Y,Hishiyama Y. Experimental evidence of a single nano-graphene. Chem Phys Lett 2001;348(1–2):17–20.
[11] Berger C, Song Z, Li X, Wu X, Brown N, Naud C, et al.Electronic confinement and coherence in patterned epitaxialgraphene. Science 2006;312(5777):1191–6.
[12] Kim KS, Zhao Y, Jang H, Lee SY, Kim JM, Kim KS, et al. Large-scale pattern growth of graphene films for stretchabletransparent electrodes. Nature 2009;457(7230):706–10.
[13] Reina A, Jia X, Ho J, Nezich D, Son H, Bulovic V, et al. Largearea, few-layer graphene films on arbitrary substrates bychemical vapor deposition. Nano Lett 2008;9(1):30–5.
[14] Cano-Marquez AG, Rodriguez-Macias FJ, Campos-Delgado J,Espinosa-Gonzalez CG, Tristan-Lopez F, Ramirez-Gonzalez D,et al. Ex-MWNTs: graphene sheets and ribbons produced bylithium intercalation and exfoliation of carbon nanotubes.Nano Lett 2009;9(4):1527–33.
[15] Kosynkin DV, Higginbotham AL, Sinitskii A, Lomeda JR,Dimiev A, Price BK, et al. Longitudinal unzipping of carbonnanotubes to form graphene nanoribbons. Nature2009;458(7240):872–6.
[16] Elias AL, Botello-Mendez AR, Meneses-Rodriguez D, JehovaGonzalez V, Ramirez-Gonzalez D, Ci L, et al. Longitudinalcutting of pure and doped carbon nanotubes to formgraphitic nanoribbons using metal clusters as nanoscalpels.Nano Lett 2009;10(2):366–72.
[17] Kim K, Sussman A, Zettl A. Graphene nanoribbons obtainedby electrically unwrapping carbon nanotubes. ACS Nano2010;4(3):1362–6.
[18] Jiao L, Zhang L, Wang X, Diankov G, Dai H. Narrow graphenenanoribbons from carbon nanotubes. Nature2009;458(7240):877–80.
[19] Withers F, Bointon TH, Dubois M, Russo S, Craciun MF.Nanopatterning of fluorinated graphene by electron beamirradiation. Nano Lett 2011;11(9):3912–6.
[20] Disa E, Dubois M, Guerin K, Kharbache H, Masin F, Hamwi A.The effect of nanostructure on the thermal properties offluorinated carbon nanofibres. Carbon 2011;49(14):4801–11.
[21] Watanabe N, Koyama S, Imoto H. Thermal decomposition ofgraphite fluoride. I. Decomposition products of graphitedefluoride, (CF) in a vaccum. Bull Chem Soc Jpn1980;53:2731–4.
[22] Moguet F, Bordere S, Rabardel L, Tressaud A, Rouquerol F,Llewelly P. Deintercalation process of fluorinated carbonfibres. Part I - Controlled rate evolved gas analysis. Mol CrystLiq Cryst Sci Technol Section A 1998;310(1):111–8.
[23] Sato Y, Hagiwara R, Ito Y. Thermal decomposition of 1st stagefluorine–graphite intercalation compounds. J Fluorine Chem2001;110(1):31–6.
[24] Kuriakose AK, Margrave JL. Mass spectrometric studies of thethermal decomposition of poly(carbon monofluoride). InorgChem 1965;4(11):1639–41.
[25] Watanabe N, Shibuya A. Reaction of fluorine and carbons,and properties of their compounds. J Chem Soc Jpn1968;71(7):963–7 [in Japanese].
[26] Takashima M, Watanabe N. Formation and structure ofcrystalline graphite fluoride. J Chem Soc Jpn 1975;3:432–6 [inJapanese].
[27] Zhang W, Dubois M, Guerin K, Bonnet P, Petit E, Delpuech N,et al. Effect of graphitization on fluorination of carbonnanocones and nanodiscs. Carbon 2009;47(12):2763–75.
[28] Lynum S, Hugdahl J, Hox K, Hildrum R, Nordvik M. patentEP1017622.
[29] Krishnan A, Dujardin E, Treacy MMJ, Hugdahl J, Lynum S,Ebbesen TW. Graphitic cones and the nucleation of curvedcarbon surfaces. Nature 1997;388(6641):451–4.
[30] Basire C, Ivanov DA. Evolution of the lamellar structureduring crystallization of a semicrystalline-amorphous
polymer blend: time-resolved hot-stage SPM Study. Phys RevLett 2000;85(26):5587–90.
[31] Press WH, Flannery BP, Teukolsky SA, Vetterling WT.Numerical Recipes in C, The Art of ScientificComputing. Cambridge University Press; 1988. p. 47–48.
[32] Panich AM. Nuclear magnetic resonance study of fluorine–graphite intercalation compounds and graphite fluorides.Synth Metals 1999;100(2):169–85.
[33] Dubois M, Giraudet J, Guerin K, Hamwi A, Fawal Z, Pirotte P,et al. EPR and Solid-State NMR Studies of Poly(dicarbonmonofluoride) (C2F)n. J Phys Chem B 2006;110(24):11800–8.
[34] Giraudet J, Dubois M, Guerin K, Delabarre C, Hamwi A, MasinF. Study of the post-fluorination of (C2.5F)n Fluorine�GIC. JPhys Chem B 2007;111(51):14143–51.
[35] Giraudet J, Dubois M, Hamwi A, Stone WEE, Pirotte P, Masin F.Solid-State NMR (19F and 13C) study of graphitemonofluoride (CF)n: 19F Spin�Lattice Magnetic Relaxationand 19F/13C distance determination by Hartmann�HahnCross Polarization. J Phys Chem B 2004;109(1):175–81.
[37] Mallouk T, Hawkins BL, Conrad MP, Zilm K, Maciel GE, BartlettN Raman. Infrared and n.m.r studies of the graphitehydrofluorides. Philosl Trans R Soc London A1985;314(1528):179–87.
[37] Sato Y, Itoh K, Hagiwara R, Fukunaga T, Ito Y. On the so-called‘‘semi-ionic’’ C-F bond character in fluorine–GIC. Carbon2004;42(15):3243–9.
[38] Dubois M, Guerin K, Pinheiro JP, Fawal Z, Masin F, Hamwi A.NMR and EPR studies of room temperature highly fluorinatedgraphite heat-treated under fluorine atmosphere. Carbon2004;42(10):1931–40.
[39] Guerin K, Pinheiro JP, Dubois M, Fawal Z, Masin F, Yazami R,et al. Synthesis and characterization of highly fluorinatedgraphite containing sp2 and sp3 carbon. Chem Mater2004;16(9):1786–92.
[40] Zhang W, Guerin K, Dubois M, Fawal ZE, Ivanov DA, Vidal L,et al. Carbon nanofibres fluorinated using TbF4 asfluorinating agent. Part I: Structural properties. Carbon2008;46(7):1010–6.
[41] Zhang W, Guerin K, Dubois M, Houdayer A, Masin F, HamwiA. Carbon nanofibres fluorinated using TbF4 as fluorinatingagent. Part II: Adsorption and electrochemical properties.Carbon 2008;46(7):1017–24.
[42] Chamssedine F, Dubois M, Guerin K, Giraudet J, Masin F,Ivanov DA, et al. Reactivity of carbon nanofibers withfluorine gas. Chem Mater 2006;19(2):161–72.
[43] Ferrari AC. Raman spectroscopy of graphene and graphite:disorder, electron–phonon coupling, doping andnonadiabatic effects. Solid State Commun 2007;143(1–2):47–57.
[44] Ferrari AC, Meyer JC, Scardaci V, Casiraghi C, Lazzeri M, MauriF, et al. Raman spectrum of graphene and graphene layers.Phys Rev Lett 2006;97(18):187401.
[45] Ferrari AC, Robertson J. Interpretation of Raman spectra ofdisordered and amorphous carbon. Phys Rev B2000;61(20):14095–107.
[46] Wang Z, Huang X, Xue R, Chen L. Dispersion effects of Ramanlines in carbones. J Appl Phys 1998;84(1):227–31.
[47] Withers F, Dubois M, Savchenko AK. Electron properties offluorinated single-layer graphene transistors. Phys Rev B2010;82(7):073403.
[48] Naess SN, Elgsaeter A, Helgesen G, Knudsen KD. Carbonnanocones: wall structure and morphology. Sci Technol AdvMater 2009;10(6):065002.
[49] Zhang W, Moch L, Dubois M, Gu, rin K, Giraudet J, et al. Directfluorination of carbon nanocones and nanodiscs. JNanoscience Nanotech 2009;9(7):4496–501.