High-resolution 13C NMR studies of high-pressure-polymerized C60: Evidence for the [2+2] cycloaddition structure in the rhombohedral two-dimensional C60 polymer

Article abstract of DOI:10.1103/PhysRevB.54.R3676

We report on ¹³C NMR measurements of a rhombohedral two-dimensional polymer of C₆₀ obtained under high pressure. By spinning the sample up to 12 kHz, we were able to identify six resonances at 149.1, 147.9, 145.2, 139.6, 134.8, and 73.5 ppm. The static distortion of the C₆₀ molecules induced by the transformation under pressure must be at the origin of the observed inequivalent carbons in the polymer. The ¹³C NMR line shape simulation of the obtained spectrum is compatible with the suggested polymeric structures where the C₆₀ molecules are connected by [2 + 2] cycloadditions.

Full text of DOI:10.1103/PhysRevB.54.R3676

PHYSICAL REVIEW B
VOLUME 54, NUMBER 6
1 AUGUST 1996-II
High-resolution 13C NMR studies of high-pressure-polymerized C60: Evidence for the †2؉2‡
cycloaddition structure in the rhombohedral two-dimensional C60 polymer
C. Goze, F. Rachdi,* and L. Hajji
´
´
Groupe de Dynamique des Phases Condensees Universite des Sciences et Techniques du Languedoc,
Montpellier II Pl. E. Bataillon, 34095 Montpellier, France
´
˜
M. Nunez-Regueiro
ˆ
EPM-Matformag, CNRS, Boıte Postale 166 Cedex 09, 38042 Grenoble, France
L. Marques and J-L. Hodeau
ˆ
Laboratoire de Cristallographie, CNRS, Boıte Postale 166 Cedex 09, 38042 Grenoble, France
M. Mehring
¨
Physikalisches Institut, Universitat Stuttgart, Pfaffenwaldring 57, 70550 Stuttgart, Germany
͑Received 2 February 1996; revised manuscript received 23 May 1996͒
We report on 13C NMR measurements of a rhombohedral two-dimensional polymer of C60 obtained under
high pressure. By spinning the sample up to 12 kHz, we were able to identify six resonances at 149.1, 147.9,
145.2, 139.6, 134.8, and 73.5 ppm. The static distortion of the C60 molecules induced by the transformation
under pressure must be at the origin of the observed inequivalent carbons in the polymer. The 13C NMR line
shape simulation of the obtained spectrum is compatible with the suggested polymeric structures where the
C60 molecules are connected by 2ϩ2 cycloadditions. ͓S0163-1829͑96͒51630-9͔
͓
͔
I. INTRODUCTION
mg of each sample have been used. All the pristine C60
starting materials were 10% 13C enriched in order to increase
the NMR signal and the samples were kept in air.
Photoinduced polymerization of solid C60 films was
reported by Rao et al.1 The cycloaddition 2ϩ2 mechanism
NMR measurements were performed using a 200 MHz
and 400 MHz Bruker NMR spectrometers with 13C NMR
Larmor frequencies of 50.2 and 100.4 MHz, respectively.
We used probes with 4 mm rotor diameter that allows to spin
the samples up to 15 kHz. The spectra were obtained by a
one pulse sequence ␲/2 acquisition with repetition time of
900 and 200 s. The samples were obtained by compressing
the enriched C60 ͑encapsulated in a Pt container͒ in a belt
apparatus for periods not exceeding two hours. The rh-2D
sample was made at 5 GPa and 700 °C and the graphite
phase was recovered from a C60 sample treated at 4 GPa and
800 °C. The x-ray-diffraction pattern of the polymerized
rh-2D sample is nearly the same as the one published
earlier.4
͓
͔
between two adjacent C60 molecules was proposed to be
at the origin of the observed lattice contraction and increase
of disorder on phototransformed C60 films. The reaction was
shown also to be reversible, as the pristine fcc C60 structure
is recovered by heating the polymer above 500 K. Bulk
polymerized C60 phases obtained at high pressure and
high temperature have also been reported.2,3 At least three
phases were synthesized, characterized, and theoretically
investigated.4–7 The pressure, temperature, and duration syn-
thesis conditions are essential to obtain one of the rhombo-
hedral, orthorhombic, or tetragonal phase. As with other non-
saturated organic molecules similar reactions as in
photopolymerization1 were expected to appear in solid C60
under pressure and temperature. By this technique bulk or-
dered polymerized C60 samples can be easily obtained giving
access to a larger range of characterization experiments. In
this paper we report on 13C NMR MAS measurements on
the rhombohedral two-dimensional ͑rh-2D͒ polymerized
phase. The observed inequivalent carbons in the studied
polymer are discussed in terms of the bonding nature be-
tween adjacent C60 molecules and distortion from the spheri-
cal shape of the molecules. Also, we discuss the molecular
dynamic in this system.
III. RESULTS AND DISCUSSION
Figure 1 shows the 13C NMR MAS spectra of the rh-2D
sample with different spinning rates and repetition times ͑a͒
at 900 s, ͑b͒ and ͑c͒ at 200 s. In both spectra ͓see Figs. 1͑a͒
and 1͑b͔͒ we observed two isotropic lines at the positions
146 and 73.5 ppm and two sidebands corresponding to the
line at 146 ppm, which appear at 266 and 26 ppm. The line
at the position 146 ppm exhibits a structure. A deconvolution
of this group of lines ͑see Fig. 2͒ gives us 6 isotropic lines at
the following positions 134.8, 139.6, 142.8, 145.2, 147.9,
and 149.1 ppm. We suspect the line at 142.8Ϯ1 ppm coming
from unreacted pristine C60 and evaluate its contribution to
be about 5% of the total intensity ͑that could be probably
undetected by the x-ray characterization͒. The relative inten-
II. EXPERIMENT
We performed 13C NMR measurements at ambient tem-
perature on rh-2D polymerized fullerene C60 and amorphous
graphite obtained by extreme compression of C60 . About 30
0163-1829/96/54͑6͒/3676͑3͒/$10.00
54
R3676
© 1996 The American Physical Society

54
HIGH-RESOLUTION 13C NMR STUDIES OF . . .
R3677
As expected in such frozen systems of polymerized
C
60 , we found long spin-lattice relaxation time T1 and very
broad static powder spectra. As it has been shown for the
A6C60 phase with AϭK, Rb, and Cs ͑Ref. 8͒ the large extent
of anisotropy of about 300 ppm indicates the freezing in of
the C60 molecules. The latter can be blocked by electrostatic
forces in A6C60 or by bonding between the C60 molecule and
its neighbor C60 molecules in polymer configurations. The
13C MAS NMR spectrum reveals inequivalent carbons sites
which can be interpreted as sp2 carbons around the 146 ppm
line position and sp3 bonding carbons between C60 mol-
ecules at 73.5 ppm. Such covalent bonding have been sug-
gested also in A1C60 polymerized phase where 13C NMR
resonances have been observed in the range 60–80 ppm.9–12
The disagreement in the relative intensities of the sp2 and
sp3 carbons calculated from structural studies and obtained
from experiment with repetition time of 200 s comes from
the very long T1 of the sp3 carbons, see Table I. The inten-
sity of sp2 carbon is always overestimated in comparison
with the sp3 carbon intensity due to the chemical shift an-
isotropy relaxation mechanism which is usually more effi-
cient for sp2 carbons. Such behavior is visible in the 13C
MAS NMR spectrum of rh-2D polymer phase obtained with
a repetition time of 900 s which shows the increase of the
sp3 carbon resonance intensity compared with the one ob-
tained with 200 s repetition time ͑see Fig. 1͒. But here again
all the intensity of sp3 carbons is not totally covered because
of the long T1, nevertheless the obtained lines intensities are
very close to what is expected from the cycloaddition model.
Unfortunately such experiments requiring large repetition
time were too long to be correctly performed in order to get
a good resolution in the structure of the five isotropic lines
around 146 ppm.
We turn now to the splitting of the line at 146 ppm, see
the 13C MAS NMR detailed spectrum in Fig. 2 and Table I
for the relative intensities. It has been shown4 that the
fullerene molecule are no longer spheres, but are deformed
by the new bonding and are flattened perpendicular to the
plane of polymerization. It is tempting to interpret the five
isotropic lines as coming from this distortion from sphericity
of the C60 molecule. We can expect on the same C60 mol-
ecule inequivalent carbons due to different neighboring. The
carbons involved here are all sp2 hybridized and they have
three neighbors which can be at slightly different positions
due to the nonsphericity. The small changes in the angles and
the distances are at the origin of the five inequivalent carbons
observed. From x-ray studies we know the space group
FIG. 1. 13C MAS NMR spectra of rhombohedral ͑2D͒ two-
dimensional phase with ͑a͒ 13C NMR Larmor frequency of 50.2
MHz, spinning rate of 6 kHz, repetition time of 900 s and 1250
scans; ͑b͒ 13C NMR frequency 100.4 MHz, spinning rate of 12 kHz
and repetition time of 200 s, 3000 scans; ͑c͒ 13C NMR frequency
50.2 MHz, spinning rate of 1.5 kHz and repetition time of 200 s,
3000 scans. The stars indicate the sidebands of the line at 146 ppm.
sities of the others five isotropic lines including the side
bands are reported in the Table I. The 13C NMR spectrum
͓see Fig. 1͑c͔͒ using a low spinning rate of 1.5 kHz shows
the two isotropic lines at 146 and 73.5 ppm and sidebands
separated by the spinning frequency. The observed sidebands
correspond to the line at 146 ppm and they mimic the enve-
lope of the static spectrum which shows a large extent of
anisotropy of about 300 ppm. No sideband was detected for
the line at 73.5 ppm indicating a very weak chemical shift
anisotropy of the corresponding carbons. In order to avoid
confusion in the interpretation of the obtained NMR spec-
trum of rh-2D polymer sample, we performed 13C NMR
measurements on amorphous graphite obtained at high pres-
sure. The latter NMR spectrum, obtained using the same
conditions as for the polymerized samples presents a very
broad line ͑not shown here͒ and no isotropic lines similar to
the rh-2D polymerized C60 were observed.
TABLE I. Relative intensities of the isotropic lines including the
sidebands in rh-2D phase.
experimental ͑900 s͒ experimental ͑200 s͒
calculated
ppm
spectrum 1͑a͒
spectrum 1͑b͒
from Fig. 3
149.1
147.9
145.2
139.6
134.8
73.5
13.9
16.1
13.3
12
12
12
6
6
49
53.9
48
ͮ ͮ ͮ
FIG. 2. The line around 146 ppm of the 13C NMR spectrum ͑b͒
of rhombohedral ͑2D͒ phase. The structure can be deconvoluted
using 6 Lorentzian line shapes. Note the presence of about 5% of
unreacted pristine C60 at 143 ppm.
5.4
5.2
11
6.1
12

R3678
C. GOZE et al.
54
therefore provide evidence that the carbons 2 and 4 are less
numerous and are the closest one to the plane where the
bondings are formed. This scenario leads us to think that the
carbons close to this plane are more shielded, in agreement
with the ratio and the shift of the 13C NMR line positions.
The other three carbons labelled 3, 5, and 6 are the numerous
ones and are shifted to the low field. We found here an
evidence for the distortion of the molecule from the spheric-
ity. The number of inequivalent sp2 carbons predicted by the
structural studies are in good agreement with our NMR re-
sults.
IV. CONCLUSION
We have shown here that the ͓2ϩ2͔ cycloaddition reac-
tions predicted in polymerized C60 is clearly verified by our
13C MAS NMR measurements on the rhombohedral ͑2D͒
polymer. The static spectrum indicates the freezing in of the
C60 molecules. MAS NMR measurements reveal very long
spin-lattice relaxation time of carbons and the observed reso-
nance positions and relative intensities are in good agree-
ment with the expected behavior of sp2 and sp3 carbons in
the polymer. Finally, the splitting of the line at 146 ppm in
five components around the position of pristine C60 indicates
the deformation of the C60 molecule spherical shape leading
to the observed inequivalent carbons sites. Further magnetic
shielding calculations taking into account the nonsphericity
of the C60 molecule in order to get a better understanding of
the observed five components are under progress.
FIG. 3. Schematic representation of rhombohedral ͑2D͒ phase
for zϾ0, empty circles and dashed lines represent some carbons
with zϽ0 participating to the cycloaddition connections.
R3m and the six inequivalent coordinates are needed to gen-
erate all the other carbons positions. See Fig. 3 for the dif-
ferent positions of the inequivalent carbons labeled from 1 to
6. The same notations as Ref. 4 are used. Our NMR results
ACKNOWLEDGMENT
L.M. was supported by JNICT.
*
Author to whom correspondence should be addressed.
7 F. Rachdi, C. Goze, M. Nunez-Regueiro, L. Marques, J.-L.
Hodeau, and M. Mehring ͑unpublished͒.
1 A. M. Rao, P. Zhou, K.-A. Wang, G. T. Hager, J. M. Holden, Y.
Wang, W.-T. Lee, X.-X. Bi, P. C. Eklund, D. S. Cornett, M. A.
Duncan, and I. J. Amster, Science 259, 955 ͑1993͒.
8 F. Rachdi, J. Reichenbach, L. Firlej, P. Bernier, M. Ribet, R.
Aznar, G. Zimmer, M. Helme, and M. Mehring, Solid State
Commun. 87, 547 ͑1993͒.
2 Y. Iwasa, T. Arima, R. M. Fleming, T. Siegrist, O. Zhou, R. C.
Haddon, L. J. Rothberg, K. B. Lyons, H. L. Carter, Jr., A. F.
Hebard, R. Tycko, G. Dabbagh, J. J. Krajewski, G. A. Thomas,
9 H. Alloul, V. Brouet, Y. Yoshinari, L. Malier, E. Lafontaine, and
L. Forro, in Physics and Chemistry of Fullerenes and Deriva-
tives, Proceedings of the International Winterschool in Elec-
tronic Properties of Novel Materials ͑World Scientific, Sin-
gapore, 1995͒, p. 443.
and T. Yagi, Science 264, 1570 ͑1994͒.
3
´
˜
O. Bethoux, M. Nunez-Regueiro, L. Marques, J-L. Hodeau, and
´
10 K.-F. Thier, G. Zimmer, M. Mehring, and F. Rachdi, Phys. Rev.
M. Perroux, in Proceedings of the Materials Research Society,
Boston, 1993, Abstract of Contributed Papers ͑Materials Re-
search Society, Pittsburgh, 1993͒, Abstract No. G2.9, p. 202.
B 53, 496 ͑1995͒.
11
¨
T. Kalber, G. Zimmer, and M. Mehring, Phys. Rev. B 51, 16 471
4
˜
´
M. Nunez-Regueiro, L. Marques, J.-L. Hodeau, O. Bethoux, and
͑1995͒.
12 C. Goze, F. Rachdi, M. Apostol, M. Mehring, G. Zimmer, and J.
E. Fischer, Proceedings of the European Materials Research
Society, Strasbourg ͓Synth. Met. 77, 115 ͑1995͔͒.
M. Perroux, Phys. Rev. Lett. 74, 278 ͑1995͒.
5 C. H. Xu and G. E. Scuseria, Phys. Rev. Lett. 74, 274 ͑1995͒.
6 C. Goze, Ph.D. thesis, University of Montpellier, France, 1996.