Fluorination of single-wall carbon nanotubes

Article abstract of DOI:10.1016/S0009-2614(98)01026-4

Purified single-wall carbon nanotubes (SWNTs) were fluorinated at several different temperatures. Product stoichiometries were determined and electron microscopy was used to verify whether or not the fluorination was destructive of the tubes. SWNTs fluori

Full text of DOI:10.1016/S0009-2614(98)01026-4

30 October 1998
Ž
.
Chemical Physics Letters 296 1998 188–194
Fluorination of single-wall carbon nanotubes
a
a,b
a
E.T. Mickelson , C.B. Huffman , A.G. Rinzler a,b, R.E. Smalley a,b, R.H. Hauge ,
a,)
J.L. Margrave
a
Rice Quantum Institute and Department of Chemistry, MS-60, Rice UniÕersity, 6100 Main Street, Houston, TX 77005, USA
Center for Nanoscale Science and Technology and Department of Physics, MS-100, Rice UniÕersity, 6100 Main Street, Houston, TX
b
77005, USA
Received 26 May 1998; in final form 12 August 1998
Abstract
Ž
.
Purified single-wall carbon nanotubes SWNTs were fluorinated at several different temperatures. Product stoichiom-
etries were determined and electron microscopy was used to verify whether or not the fluorination was destructive of the
tubes. SWNTs fluorinated at three different temperatures were then defluorinated using hydrazine. Raman spectroscopy and
resistance measurements were utilized to verify whether or not the products of the defluorination were in fact SWNTs. It has
been determined that the bulk of the SWNTs survive the fluorination process at temperatures up to 3258C and that hydrazine
can be employed as an effective defluorinating agent to regenerate the unfluorinated starting material. q 1998 Elsevier
Science B.V. All rights reserved.
1. Introduction
materials. Recently, sono-chemistry was employed to
cut the long, intertangled tubes into independent,
Ž
Since the discovery of single-wall carbon nan-
macro-molecular scale, open tube fragments 50–300
Ž
. w x
. w x
otubes SWNTs 1 there has been a flurry of re-
search activity aimed at understanding their physical
nm long 6 . In that work, the high reactivity of the
dangling carbon bonds at the open tube ends was
exploited to tether the tubes to gold particles via
thiol linkages.
In contrast to the open tube ends, the side-walls of
the SWNTs, by virtue of their aromatic nature, pos-
sess a chemical stability akin to that of the basal
w x
properties 2 , elucidating their growth mechanisms
w x
w x
3 and developing novel uses for them 4 . Chem-
istry involving SWNTs is still in its infancy. This is
due, in large part, to previous difficulties in obtain-
ing reasonable quantities of highly purified SWNTs.
w x
w x
Progress in the bulk synthesis and purification 5
of SWNTs is now making available high-quality
samples in sufficient quantities to begin exploring
the chemical modification of this fascinating class of
ruption of the tubular structure is thus significantly
Using technology developed in the fluorination of
w x
graphite 8 , we report here on the chemical manipu-
lation of the SWNT side-wall by fluorinating high-
purity SWNTs and then defluorinating them. In a
)
Corresponding author. E-mail: margrav@rice.edu
0009-2614r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 98 01026-4

(
)
E.T. Mickelson et al.rChemical Physics Letters 296 1998 188–194
189
w x
manner similar to that of fluorinated fullerenes 9 ,
of 325 and 4008C, the fluorination was begun at
2508C and after 1 h, the F2 flow was stopped and the
reactor temperature brought up to the appropriate
level after which the F2 flow was resumed for an
additional 4 h. For the reactions at 500 and 6008C,
the sample was fluorinated for 1 h at 2508C, 1 h at
4008C and then 3 h at the specified reaction tempera-
ture. The rationale behind this ‘stepped’ reaction
temperature procedure was to minimize, as much as
possible, the decomposition: CFn ™CF4, C2 F4, C2 F6,
etc., which has been well documented in the fluori-
we anticipate that SWNTs, once fluorinated, can
serve as a staging point for a wide variety of side-wall
chemical functionalizations.
2. Experimental
Single-walled carbon nanotubes were produced by
the dual-pulsed laser vaporization of CorNi doped
graphite rods and purified by techniques described
w x
w
x
w x
previously 5 . The purification product is
a
nation of graphite 12 and fullerenes 13 .
metastable colloidal suspension of SWNT ‘ropes’
x Ž
Once fluorinated, SWNTs were defluorinated with
w
Ž
.
10 bundles of hexagonally close-packed tubes
anhydrous hydrazine Aldrich, 98% . To the pieces
of ‘bucky paper’, fluorinated at 250, 325, and 4008C,
was added 5 ml of hydrazine under an inert atmo-
sphere at room temperature. The SWNTs were al-
lowed to sit in the hydrazine for 1 h prior to filtering
and washing with water.
.
ranging from a few to 200 SWNTs in a 0.2%
aqueous solution of Triton X-100 surfactant. Filter-
ing the solution through a PTFE filter membrane and
washing with methanol to remove residual surfactant
leaves a black film on the surface. If this layer is
Ž
sufficiently thick 10–75 mm, as measured by a
Product stoichiometries as a function of reaction
.
Ž
micrometer it can be peeled off to form a free-stand-
temperature were obtained both gravimetrically TA
.
ing film or ‘bucky paper’ of SWNTs. This form has
appreciable mechanical integrity and is convenient
for handling, and for electrical conductivity and Ra-
man scattering measurements. It is the fluorination of
this ‘bucky paper’ that we describe here.
Instruments TGA-DTA 2960 microbalance and via
electron microprobe analysis Cameca SX-50 . In-
frared spectroscopy Perkin-Elmer Paragon 1000 FT-
IR was used to confirm the presence of covalently
Ž
.
Ž
.
bound fluorine. Transmission electron microscopy
Ž
In fluorinating the SWNTs, elemental fluorine
JEOL model 2010 TEM using 100 keV beam en-
Ž
.
.
Air Products, 98% was used as the fluorinating
ergy was used to determine if, and at what tempera-
ture the tubes were being destroyed i.e., ‘unzipped’
by the fluorination. Raman spectroscopy Jobin
Yvon-Spex model HR460 monochrometer coupled
with an ISA Spectrum ONE series CCD detector and
Ž
.
agent. HF, being the major impurity in the fluorine,
was removed by passing it through an HF trap
Ž
Ž
.
Matheson Gas Products containing sodium fluoride
Ž
pellets. The fluorine, diluted with helium Trigas,
99.995% , was then passed through a temperature-
.
.
using a 532 nm Nd:YAG laser excitation source ,
scanning electron microscopy JEOL model JSM-
Ž
controlled Monel flow reactor containing the SWNT
sample.
6320F field emission SEM using 5 keV beam en-
.
Prior to fluorination, the purified ‘bucky paper’
ergy , and two-point resistivity measurements were
y6
Ž
.
was vacuum baked at 11008C 2=10
Torr for
used to analyze the untreated, fluorinated, and deflu-
several hours in order to desorb any residual surface
orinated SWNT samples.
contaminants. For each reaction a pre-weighed piece
Ž
of baked ‘bucky paper’ being ;10 mm thick and
Table 1
.
weighing 150–200 mg was used. F2 and He flow
Reaction product stoichiometries determined by both gravimetric
and electron microprobe analysis
rates for each reaction were 2 and 20 sccm, respec-
tively. In each case the reaction time was 5 h. The
only variable was reaction temperature. As the kinet-
ics of inorganic carbonqfluorine reactions are highly
Ž
.
Reaction temperature 8C
150 250 325
Gravimetric CF0.114 CF0.521 CF0.495 CF0.565
400
500
Ža.
600
Ža.
w
x
temperature dependent 11 , several reactions were
carried out at the following temperatures: 150, 250,
325, 400, 500, and 6008C. For reaction temperatures
Microprobe CF0.097 CF0.458 CF0.554 CF0.647 CF0.815 CF0.996
a No determination at these temperatures.

(
)
190
E.T. Mickelson et al.rChemical Physics Letters 296 1998 188–194
3. Results
bound to the SWNT side-wall. Product stoichiom-
etries of the fluorination reactions are shown in
Table 1. Discrepancies between the gravimetric and
microprobe analyses can be attributed to product
decomposition as described above, especially at the
higher temperatures.
Fig. 1A shows a TEM image of the purified,
unreacted SWNT material. Fig. 1B shows a TEM
image of SWNTs fluorinated at 3258C. As can be
seen from the image, the tubes remain largely intact
Ž
.
Infrared spectroscopy KBr pellet method con-
firmed the presence of covalently bound fluorine
peaks in the 1220–1250 cm
y1
Ž
.
region in the sam-
ples fluorinated at temperatures of 2508C and higher.
No C–F stretching frequency was seen for the sam-
ple fluorinated at 1508C, and its two-point resistance
5 mm apart was ;100 V which therefore pre-
cludes large amounts of fluorine being covalently
Ž
.
Ž
.
Ž .
Ž .
Fig. 1. A TEM image of pure, unreacted SWNTs. B TEM of SWNTs after being fluorinated at 3258C. C TEM of SWNTs after being
Ž
.
fluorinated at 5008C. D Another TEM of SWNTs fluorinated at 5008C showing the formation of MWNTs.

(
)
E.T. Mickelson et al.rChemical Physics Letters 296 1998 188–194
191
after treatment under these conditions. Fig. 1C is a
TEM image of SWNTs fluorinated at 5008C. Here it
would appear that the tubes are essentially all de-
stroyed. However, a fair number of nested tube-like
graphitic structures reminiscent of multiwall carbon
temperatures of 2508C and above are insulators
Ž
.
two-probe resistance )20 MV .
We have found that, once fluorinated, SWNTs
can be defluorinated with anhydrous hydrazine via
1
the following reaction: CFn q 4 nN2 H4 ™ CqnHF
1
Ž
.
nanotubes MWNTs seem to have been generated as
a result of the high-temperature reaction. These
structures are shown in Fig. 1D.
q 2 nN2. As the Raman spectroscopy of SWNTs has
w
x
been well developed both theoretically 15 and ex-
w
x
perimentally 16 , we utilize it here to examine the
results of both the fluorination and subsequent deflu-
orination of the SWNTs. Fig. 2 shows the Raman
spectrum of the pure, unadulterated SWNT material.
The smaller peak at 186 cmy1 is due to a character-
istic breathing mode of the SWNTs. Raman spectra
of SWNTs fluorinated for 5 h at reaction tempera-
tures of 250, 325, and 4008C are shown in Fig. 3.
Trace A, corresponding to the reaction at 2508C,
shows only two broad peaks centered around 1340
and 1580 cmy1 and a broad band fluorescence. The
Raman peaks correspond to sp3 and sp2 carbon
stretching modes, respectively. At higher reaction
temperatures, yielding high F to C ratios, these peaks
disappear and the fluorescence is attenuated. As C–F
bonds are not very polarizable, it is understandable
that they are not seen in the Raman spectra presented
here.
The fluorination of MWNTs has been reported
w
x
previously 14 . This was done at two temperatures:
25 and 5008C. The room temperature reaction was
done with a F2 , HF, and IF5 mixture, and yielded an
intercalated type compound. The reaction carried out
at 5008C was done with F2 , and was determined to
have destroyed the tube structure to yield a graphite
fluoride compound of stoichiometry CF. In light of
this, it is not too surprising that we observe destruc-
tion of the SWNTs at 5008C, but somewhat surpris-
ing that MWNT-like structures are formed.
Not unexpectedly, the electronic properties of the
fluorinated tubes differ dramatically from those of
their unfluorinated counterparts. While the untreated
Ž
SWNTs are good conductors 10–15 V two-probe
resistance across the length of the ;10=3 mm=30
.
mm bucky paper samples , the tubes fluorinated at
Fig. 2. Raman spectrum of pure, unreacted carbon SWNTs.

(
)
192
E.T. Mickelson et al.rChemical Physics Letters 296 1998 188–194
Ž
.
Ž .
Ž .
Fig. 3. Raman spectra of SWNTs fluorinated at: A 2508C; B 3258C; and C 4008C.
Raman spectra of the defluorinated products of
true in trace C, indicating that the tubes are largely
destroyed at 4008C. Additionally, the peak at 1340
cmy1 grows relative to the characteristic SWNT
peaks with increasing fluorination temperature. This
can be attributed to one or both of the following
the SWNTs originally fluorinated at 250, 325, and
4008C are shown in Fig. 4. Traces A, B, and C
correspond to the material originally fluorinated at
250, 325, and 4008C, respectively. As can be seen in
traces A and B, the characteristic breathing mode at
186 cmy1 returns upon defluorination. This is not
Ž .
factors: 1 tubes are being ‘unzipped’ much more
Ž .
readily at the higher temperatures; and 2 at higher
Ž
.
Ž .
Ž .
Fig. 4. Raman spectra showing the defluorination of the nanotubes originally fluorinated at: A 2508C; B 3258C; and C 4008C.

(
)
E.T. Mickelson et al.rChemical Physics Letters 296 1998 188–194
193
reaction temperatures, a greater amount of decompo-
sition of the type: CFn ™CF4, C2 F4, C2 F6, etc., is
occurring.
SEM images and two-point resistivity measure-
ments were obtained on a single piece of ‘bucky
paper’ after each of the following stages: purifica-
tion, fluorination at 3258C, and defluorination in
hydrazine at room temperature for 1 h. Fig. 5A shows
the purified starting material. Fig. 5B shows the
same piece after having been fluorinated at 3258C
for 5 h. The image shows excessive charging due to
its insulating nature, but the ‘rope-like’ structure of
tubes is still evident. Finally, Fig. 5C shows the
same piece of ‘bucky paper’ after having been deflu-
orinated in hydrazine. The two-point resistance of
the purified starting material is 15–16 V measured 5
mm across the surface of the ‘bucky paper’. Identi-
cal measurements on the fluorinated and defluori-
nated ‘bucky paper’ yielded a resistance of )20
MV and 125–130 V, respectively. It is interesting
to note that the defluorinated material recovers most,
but not all of its original conductivity.
4. Conclusions
We have demonstrated via gravimetric and elec-
tron microprobe analysis that large amounts of fluo-
rine can be added to SWNTs. Resistance measure-
Ž
ments along with vibrational spectroscopy Raman,
.
IR confirm the formation of new chemical bonds to
the nanotube superstructure. Contributions of ad-
sorbed molecular fluorine to the overall fluorine
w
x
uptake are precluded at these temperatures 11 . We
conclude, therefore, that fluorine is being covalently
attached to the side-wall of the SWNTs.
TEM studies have shown that at fluorination tem-
peratures as high as 3258C, the majority of the
fluorination product maintains a tube-like structure.
These studies also indicate that at 5008C, the single-
wall tubular structure does not survive the fluorina-
tion process and that some MWNT-like structures
are being formed.
From the product stoichiometries, resistance mea-
surements and IR spectra we conclude that reaction
temperatures in excess of 1508C are necessary to
covalently add significant amounts of fluorine to the
tube wall. The small amount of fluorine that does
show up in the product of the 1508C fluorination
reaction could be attributed to a combination of
Ž
.
Ž .
Fig. 5. A SEM of pure, unreacted SWNTs. B SEM of SWNTs
Ž .
C SEM of
after having been fluorinated at 3258C for 5h.
SWNTs fluorinated at 3258C and then defluorinated in hydrazine.

(
)
194
E.T. Mickelson et al.rChemical Physics Letters 296 1998 188–194
adsorbed fluorine and fluorination of the end caps of
the SWNTs.
References
We have also demonstrated that fluorine can be
effectively removed from the SWNTs using anhy-
drous hydrazine and that the rejuvenated product is
in fact a SWNT. From the results of these defluorina-
tion experiments and the Raman and SEM studies
associated with them, we would conclude that a
majority of the tubes are destroyed at fluorination
temperatures of around 4008C, whereas only a slight
amount of tube destruction occurs at 2508C.
We believe that for reactions in which only the
w x
Ž
.
Ž
.
1
w x
2
S. Iijima, T. Ichihashi, Nature London 363 1993 603.
J.-P. Issi, L. Langer, J. Heremans, C.H. Olk, Carbon 33
Ž .
7
Ž
.
1995 941.
w x
3
Ž
.
C.F. Cornwell, L.T. Wille, Chem. Phys. Lett. 278 1997
262.
A.C. Dillon, K.M. Jones, T.A. Bekkedahl, C.H. Kiang, D.S.
Bethune, M.J. Heben, Nature London 386 1997 377.
A.G. Rinzler, J. Liu, H. Dai, P. Nikolaev, C.B. Huffman, F.J.
Rodriguez-Macias, P.J. Boul, H. Lu, D. Heymann, D.T.
Colbert, R.S. Lee, J.E. Fischer, A.M. Rao, P.C. Eklund, R.E.
Smalley, Appl. Phys. A 1998, in press .
J. Liu, A. Rinzler, H. Dai, J. Hafner, R. Bradley, A. Lu, K.
Shelimov, C. Huffman, F.J. Rodriguez-Macias, P. Boul, T.
Iverson, D.T. Colbert, R.E. Smalley, Science 1998, in press .
w x
4
Ž
.
Ž
.
w x
5
Ž
.
w x
6
Ž
outside of the tube is being fluorinated the SWNTs
.
used in this study are closed at the ends , that there is
Ž
.
w x
Ž
.
a limiting stoichiometry of C2 F for which the fluori-
7
w x
8
J. Aihara, J. Phys. Chem. 98 1994 9773.
R.J. Lagow, R.B. Badachape, J.L. Wood, J.L. Margrave, J.
Chem. Soc., Dalton Trans. 12 1974 1268.
R. Taylor, J.H. Holloway, E.G. Hope, A.G. Avent, G.J.
Langley, T.J. Dennis, J.P. Hare, H.W. Kroto, D.R.M. Wal-
Ž
nated tube can still maintain its tube-like albeit
Ž
.
.
puckered structure. This is supported by the product
w x
9
stoichiometries we have obtained via elemental anal-
ysis and the evidence of significant tube destruction
at fluorination temperatures greater than 3258C. Fur-
ther addition of fluorine would then lead to the
breaking of carbon–carbon bonds and, hence, de-
struction of the tube.
Efforts to conceptualize a structure consistent with
this C2 F stoichiometry are ongoing and will be
presented shortly.
Ž
.
ton, J. Chem. Soc., Chem. Commun. 9 1992 665.
w
x
10 A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, C.
Xu, Y.-H. Lee, S.-G. Kim, A.G. Rinzler, D.T. Colbert, G.E.
Scuseria, D. Tomanek, J.E. Fischer, R.E. Smalley, Science
Ž
.
273 1996 483.
w
w
w
x
11 N. Watanabe, T. Nakajima, H. Touhara, Graphite Fluorides,
Elsevier, Amsterdam, 1988.
x
Ž
.
12 P. Kamarchik, J.L. Margrave, Acc. Chem. Res. 11 1978
296.
x
13 H. Selig, C. Lifshitz, T. Peres, J.E. Fischer, A.R. McGhie,
W.J. Romanow, J.P. McCauley, A.B. Smith, J. Am. Chem.
Ž
.
Soc. 113 1991 5475.
Acknowledgements
w
w
w
x
14 A. Hamwi, H. Alvergnat, S. Bonnamy, F. Beguin, Carbon 35
Ž . Ž
.
6
1997 723.
Ž .
15 E. Richter, K.R. Subbaswamy, Phys. Rev. Lett. 79 14
x
We would like to thank Mr. Milton Pierson for
his help with the electron microprobe analysis and
Miss Jane Bertone and Professor Vicki Colvin for
their assistance in obtaining the Raman spectra. This
work was supported by grants from the Robert A.
Ž
.
1997 2738.
x
16 A.M. Rao, E. Richter, S. Bandow, B. Chase, P.C. Eklund,
K.A. Williams, S. Fang, K.R. Subbaswamy, M. Menon, A.
Thess, R.E. Smalley, G. Dresselhaus, M.S. Dresselhaus,
Ž
.
Science 275 1997 187.
Ž
.
Welch Foundation and the ONR BMDO .