PRESSURE INDUCED DIMERISATION OF C70
M. Premila," C.S. Sundar,"'* P. Ch. Sahu," A. Bharathi," Y. Hariharan," D.V.S. Muthub and A.K. Soodb
"Materials Science Division, Indira Gandhi Centre for Atomic Research, Kalpakkam-603 102, India
bDepartment of Physics, Indian Institute of Science, Bangalore 560 01 2, India
Solid CT0 has been subjected simultaneously to high pressures and
temperatures (HPHT), with pressures upto 7.5 GPa and temperatures
upto 750°C. X-ray diffraction measurements on the recovered samples
indicate that the initial h.c.p. solid C70 transforms to a rhombohedral
structure which recovers to an f.c.c. structure on annealing. The associated
changes in the intra molecular vibrational modes have been probed through
infrared (IR) and Raman measurements. The 1R measurements on these
HPHT samples show splitting of some of the pristine modes and occurrence
of several new modes. These sharp IR modes in the HPHT treated samples,
which are seen to be different from that reported for photopolymerised C70,
have been attributed to the formation of C70 dimers. 0
Elsevier Science Ltd
1997 Published by
Keywords: A. fullerenes, D. phase transitions, D. optical properties.
The synthesis of polymeric fullerenes from solid state
precursors and an investigation of their properties has
been a topic of considerable interest [l], since the
first report of polymerisation of c 6 0 film under laser
irradiation [21. Polymerisation of c60 has been observed
 in RblCm and KIC60, as also in pristine c 6 0 when
subjected to high pressure (-5 GPa) and high temperature
(-800°C) treatment [4-71. Polymeric fullerenes consist
of covalently bonded cages which are arranged in linear
chains or two-dimensional sheets. The chains can be as
short as dimers. but can also extend to many units.
Polymerisation in C60 is associated with the (2 + 2)
cycloaddition reaction , wherein the parallel double
bonds of the neighbouring molecules react to form
polymerised structure, linked through a four-membered
cyclobutane ring. Under different conditions of pressure
and temperature, f.c.c. Cb0 is seen to transform to various
polymerised structures, such as an orthorhombic structure
consisting of one-dimensional polymeric chain or
rhombohedral and tetragonal structures consisting of
two-dimensional polymeric layers .
*Author to whom all correspondence should be
In contrast to extensive investigations on the
polymerisation of c60, there have been fewer studies on
C70. It has been shown that C70 can be photopolymerised
[8, 91, albeit with a much lower efficiency compared to
C60 and more recently there has been evidence [ 101 for
cross linking of C70 molecules in Ar plasma. The
difficulty in the polymerisation of C70 has been attributed
to the topochemical constraints specific to the C70
molecule  in that only the double bonds on the polar
caps of the molecule are reactive, whereas the cyclic
double bonds on the equatorial belt are ineffective in
undergoing (2 + 2) cycloaddition reaction. For the
polymerisation reaction to occur in C70 under HPHT
conditions, as in the case of Cb0, it is necessary that the
parallel configurations of C=C double bonds of the
neighbouring molecules be brought close together
under compression. In the case of solid C70, it is known
[ 111 that there are two orientational ordering transitions,
one corresponding to the free rotor phase going to a
long axis oriented rhombohedral phase and the second
corresponding to a completely oriented C70 in the mono-
clinic phase. Both these transitions are influenced by
pressure [ 12, 131; the rhombohedral to f.c.c. transition
temperature increases at the rate of 300 K GPa-I, whereas
the monoclinic to rhombohedral transition temperature
increases at a much slower rate of 50 KGPa-I.
Thus, depending on the high pressure and temperature
conditions employed, solid C70 could be in different
orientationally ordered states, which would influence
the possibility of polymerisation. This together with the
topochemical features of the C70 molecule was used to
rationalise the non-observance of polymerisation in our
earlier investigations  on HPHT treated C70, which
had been carried out over a limited range of pressure and
In the present study, which is a continuation of our
earlier work , we have spanned a wider range of
pressure-temperature regimes, viz., P = 1-7.5 GPa and
T = RT-750"C, to look for conditions that would favour
polymerisation. In particular, we have investigated the
influence of keeping the system at high pressure
(7.5 GPa) and low temperature (RT), and low pressure
(1 GPa) and high temperature (75OoC), wherein solid C70
is expected to be in monoclinic and rhombohedral
structures respectively. The structural information on
the HPHT treated samples has been obtained through
X-ray diffraction measurements and the intramolecular
vibrational modes have been probed by Raman and IR
measurements. The significant result of the present
investigations is the dramatic splitting of IR modes in
the HPHT treated samples [ 141, which is seen to recover
to pristine modes on annealing. In order to find out if this
splitting of IR modes is associated with the formation of
orientationally ordered C70, in situ IR measurements across
the low temperature orientational ordering transitions [ 1 11
in pristine C70 have also been carried out. With support
from these experiments and Raman measurements, the
observed splitting of IR modes in the HPHT treated
samples has been attributed to the formation of C70
dimers [ 151.
2. EXPERIMENTAL DETAILS
Chromatographically purified polycrystalline C70
powder was synthesised using the procedure described
in [ 161 and this was heat treated in vacuum at 250°C for
over 24 h to get rid of the trapped organic solvents. A
pellet of 3 mm diameter and 2 mm thickness, wrapped in
a 7.5 pm stainless steel foil, was mounted along with
steatite disks inside a pyrophyllite gasket containing a
small annular graphite ring heater and this assembly was
loaded into an opposed anvil high pressure apparatus
[ 171. This whole procedure, which was carried out under
ambient conditions, entailed an exposure of the samples
to atmospheric oxygen for a period of -30 min, before
being sealed between the gaskets. After raising the
pressure upto 7.5 GPa maximum, over a period of
10 min, the temperature was increased upto a maximum
of 750°C. The sample was allowed to soak under high
pressure and temperature for -6 h, following which the
temperature was lowered gradually and the pressure was
subsequently released slowly. X-ray diffraction measure-
ments on finely chopped pellets were carried out using
Siemens D-500 diffractometer operating in the Bragg-
Brentano arrangement. Raman measurements were
carried out, under ambient conditions, using DILOR
X-Y spectrometer with liquid nitrogen cooled CCD
detector. Infrared absorbtion spectra were recorded
using a Bomem (Model MB 100) FTIR spectrometer,
on samples pelletized with KBr. Thin films of C70,
vacuum deposited on single crystal KBr substrates,
were used for low temperature FTIR measurements, the
temperatures being achieved using a Leybold Heraeus
closed cycle helium refrigerator.
3. RESULTS AND DISCUSSION
Figure 1 (a) shows X-ray diffraction (XRD) patterns
of the pristine C70 and those treated at various pressures
Fig. 1. (a) X-ray diffractograms of the starting C70
powder, indexed to h.c.p. structure and those of samples
treated at various pressures at 750°C, indexed to
rhombohedral structure. The XRD pattern of HPHT
treated samples subsequently annealed in vacuum at
250°C is indexed to f.c.c. structure. (b) XRD pattern of
C70 treated at various temperatures at 7.5 GPa. Note the
shift of the RT and 100°C treated spectra towards smaller
28 values, which are indexable to rhombohedral structure.
at 750°C. The XRD pattern of the starting C7,, powder
can be indexed to h.c.p. structure with lattice parameter
a = 10.60 A and c = 17.26 A, consistent with earlier
studies . In the case of the HPHT treated samples,
broad diffraction lines corresponding to a new crystalline
phase is observed. As reported earlier  these could be
indexed to a rhombohedral structure (space group R3m)
with the hexagonal lattice parameters, a = 10.10 A and
c = 26.88 A. In the course of our studies on various
HPHT treated CT0, we observed that in case of samples
treated at higher pressure (7.5 GPa) and low temperature,
viz., RT and lOO"C, the diffraction pattern was
shifted towards smaller angles, as shown in Fig. l(b).
Correspondingly these could be fitted to a rhombohedral
structure with a marginally expanded lattice parameters
(a = 10.23 A and c = 27.56 A) as compared to the rest
of the samples. The rhombohedral structure obtained
after HPHT treatments were seen to transform to f.c.c.
structure ( F m h ) with a = 14.94 A on annealing at
250°C for 24 h, as shown in the top panel of Fig. l(a).
The broad diffraction patterns of the HPHT treated
C70 were not amenable to a detailed structural analysis, in
particular to obtain information on the formation of intra
molecular bonds. In order to find out if the rhombohedral
structure corresponds to a polymeric phase, we have
estimated the inter-fullerene distance. It must be
remarked that in the case of polymeric Cm obtained by
HPHT treatment, there is a distinct reduction [6,7] in the
interfullerene distance to 2.2 A, from 3 A in the f.c.c.
phase, providing a strong evidence for the formation of
intra-molecular bonds. With the known [ 181 molecular
dimensions of C70 and assuming that the C70 molecules
are aligned along the long axis, the smallest distance
between the double bonds on polar caps of neighbouring
C70 molecules at (0, 0, 0) and (213, 1/3, 1/3) can be
estimated  to be 2.61 A, while the distance between
the equatorial double bonds is 3.01 A. These inter-
fullerene distances in the rhombohedra1 phase are not
significantly different from that in the starting h.c.p.
phase or the annealed f.c.c. phase [ 181. Thus, our X-ray
diffraction measurements do not provide any direct
evidence for the polymerisation of C70. We may also
point out that qualitatively different diffraction patterns
have been obtained for polymerised C70 obtained by
other methods: For example, an amorphous XRD
pattern has been obtained [lo] in the study of plasma
polymerised C70 and Rao et al.  have observed an
increase in lattice parameter on UV polymerisation of
f.c.c. C70 film. While these differences need to be
investigated in more detail, they point out to a greater
variety in structures possible in the linking of C7(,
Figure 2 shows the results of IR studies on pristine
C7,, and those treated at 750°C at various pressures. In
lGPa, 750 OC
V l l O O 1200 1300 1400 1500
I! I,. 1, I
WAVE NUMBER (cm -1 )
Fig. 2. IR spectra of pristine and C70 treated at various
pressures at 750°C. The split IR pattern are seen to
recover to pristine modes on annealing at 250°C. The
IR modes marked with * correspond to Raman modes
measured at T = 23 K [23).
comparison with the IR spectrum of pristine C70, which is
in conformity with earlier results [19, 201, in the HPHT
treated samples the IR spectrum has developed fine
structure and several new modes have appeared. This is
particularly striking in the low wave number region: For
example, the pristine modes at 641 cm-' develops new
features on either sides, the pristine mode at 458 cm-l is
seen to develop a feature on the high wave number side at
467 cm-' while the mode at 794 cm-' is associated with
an additional mode on the lower wave number side at
776 cm-I. In addition, new modes are seen to emerge in
the spectra of the HPHT treated samples, for example the
mode at 605 cm-'. It is noted that similar features are
seen in all the pressure treated samples. Since some of
the modes in the HPHT treated sample are coincident
with those of pristine C 7 0 , it is possible that there is some
untransformed material. While, we have not carried out
quantitative studies on the solubility of HPHT treated
samples in toluene, which provides an indicator of the
residual C70, it was generally observed that the solubility
of the pressure treated samples were much lower
compared to pristine CT0. The sharp split IR spectrum
of HPHT treated samples are seen to recover to the
modes of pristine C70 on annealing at 250°C as shown
in the top panel of Fig. 2.
It was noted earlier that treatment at 7.5 GPa at low
temperatures (RT and 100°C) resulted in a rhombohedral
structure with marginally larger lattice parameters
I I I I
4. I. I
I, I, I
I ' " " " " " " " " " " " " " 1
200 300 400 500 600 700 000
WAVE NUMBER ( cm-1)
as compared to that obtained on treatment at high
temperatures, beyond 250°C [cf. Fig. I(b)]. Select results
of IR measurements on the samples treated at 7.5 GPa are
shown in Fig. 3. It is noted that whereas in the case of
sample treated at 250°C a split IR spectrum similar to that
in Fig. 2 is obtained, in the case of sample treated at RT,
the IR spectrum is unsplit and is similar to pristine C70.
This difference is striking and as we shall see below, has
a bearing on the interpretation of the origin of the split IR
The changes in the vibrational modes on HPHT
treatment of CT0 have also been probed using Raman
spectroscopy. Figure 4(a) and (b) shows the results of
Raman measurements. In the case of HPHT treated
samples, the Raman spectra are not very different from
that of pristine C70 , except for a shoulder at
1426cm-', to the left of the 1444cm-l peak and a
weak new feature at 278 cm-I to the right of the
Raman mode at 255cm-I. These subtle changes in
Raman spectra are to be contrasted with the dramatic
splitting of IR spectra in HPHT treated C70. It may be
noted [4, 7, 211 that even in the case of pressure
polymerised Cm the changes in the Raman spectra are
minimal when compared to the changes in IR spectra.
In the following, we analyse the origin of the
sharp-split IR spectrum in C70 treated at 1 GPa, 750°C.
At the outset, we may attribute the splitting of IR modes
to a change in the site symmetry associated with the
1150 1250 1350 1450 1550 1650
WAVE NUMBER (cm-1)
Fig. 4. Raman spectra of pristine and C70 treated at various pressures at 750°C in the range (a) 200-900 cm-' and (b)
1150-1650 cm-'. Note the development of new mode at 1426 cm-l and 278 cm-'.
structural transformation from the initial h.c.p. to the
rhombohedral structure. However, in this assignment we
encounter a difficulty in that splitting is not seen in the
case of sample treated at 7.5 GPa, RT (cf. Fig. 3), in
which case also the sample has a rhombohedral structure
[cf. Fig. l(b)]. In order to explore further if the splitting
of IR pattern is in any way related to the structural
transition [ 1 11 associated with the orientational ordering
of C70 molecules, we have also carried out IR measure-
ments on pristine C70 films on KBr substrates at low
temperatures upto 12 K. These results are shown in Fig. 5.
It is seen that with the decrease of temperature, while
there is a systematic decrease in the line widths of most
of the IR modes, the modes at 458 and 565 cm-' show
indications of broadening and splitting and these results
are in conformity with earlier studies . These small
changes in the IR spectra at low temperatures are to be
contrasted with the large splitting and occurrence of
several new modes in the IR spectra (cf. Fig. 2) in the
case of HPHT treated C70. This clearly indicates that
the features seen in the IR spectra of HPHT treated
samples is not related to the structural transition
involving orientationally ordered C70 monomers.
450 500 550 600
WAVE NUMBER (cm -'
650 700 750 800 850
Fig. 5. IR spectra of pristine C70 at various temperatures.
The modes at 458 and 565cm-' broaden at low
temperatures, whereas the other modes are seen to sharpen.
In trying to find out if the split IR modes are
associated with the formation of polymeric phase, we
were led to compare with the earlier IR measurements 
on photopolymerised C70. The only distinctive feature in
the IR spectrum of photopolymerised C70, is a broad band
centered at 1086 cm-', which incidentally is also seen in
photopolymerised Cm and has been attributed to
disorder . Our IR spectra of HPHT treated C70,
shown in Figs 2 and 3, are very different from that
reported  for photopolymerised C70.
Given that the IR spectrum of HPHT treated C70
is distinct from that of photopolymerised C70 and
monomeric C70 in orientationally ordered state, it is
tempting to speculate that the splitting of IR modes
arises due to the formation of C70 dimers. In support of
this speculation, we may point out that some of the
new modes, marked with asterisks in Fig. 2 can be
identified with the Raman modes of C70 measured at
low temperature . Further, the mode at 1426cm-'
seen in Rarnan spectra [cf. Fig. 4(b)] can be associated
with the dominant IR mode at 1430cm-'. It is well
known  that Rarnan modes and IR modes are
no longer strictly complimentary once the inversion
symmetry is broken, i.e. Raman modes are also observed
in the infrared spectrum and vice versa. In fact such an
argument has been used to identify Cm dimer in RblC6,,
. To obtain more definitive confirmation on the
formation of C70 dimers under HPHT treatment, detailed
theoretical analysis of the vibrational spectrum of C70
dimers are called for, as has been done recently [2 1 ] for
dimers of C60.
To summarise, our present studies on solid C70
treated at high pressure and temperatures (> 250"C),
indicate transformation to a rhombohedral structure
and this is associated with sharp split IR spectra.
These IR spectra, which are different from that seen in
photopolymerised C 70 and in orientationally ordered
monomeric C70 have been attributed to the formation
of C70 dimers. In the case of C70 treated at high pressure
(7.5 GPa) and low temperatures (RT and 100°C). the
IR spectra are similar to that in pristine C70. This is
consistent with our initial premise that under these
treatment conditions, solid C70 can be expected to be in
monoclinic structure, whose orientational order is not
conducive to the formation of covalent linkages between
molecules. In our earlier studies , we had inferred that
C70 does not polymerise, based on a limited scan of
pressure and temperature range. Further, in these studies,
the vibrational modes were mainly monitored through
Raman measurements, which as we have seen presently
is far less sensitive to the formation of intra molecular
bonds. Our present conclusion on the dimerisation of C70
on HPHT treatment is principally based on dramatic
changes seen in the IR spectra.
The formation of only the dimers in the case of
HPHT treated C70 is to be contrasted with the Occurrence
of polymeric chains and sheets in the case of similarly
treated Cm [6,7]. This may be due to the stereochemical
constraints, specific to C70 molecule, which prohibit the
formation of long range ordered chains. Alternatively, it
is possible that C70 polymers have limited metastability
in that they break up during the slow cooling and release
of pressure after HPHT treatment. Further investigations
on pressure quenched C70, instead of slow cooling as in
the present series of experiments, will be of interest.
Acknowledgements-The authors would like to thank
V. Vidhya, K. Sankaran, K.S. Viswanathan, G.V.N.
Rao, V.S. Sastry and T.S. Radhakrishnan for their help
with respect to IR and X-ray diffraction measurements.
One of us (AKS) would like to thank the Department of
Science and Technology for their financial support.
Fischer, J.E., Science, 264, 1994, 1548; see also
Rao, A.M. and Eklund, P.C., in Cluster Assembled
Materials (Edited by Klaus Sattler), Materials
Science Forum, 232, 1996, 173.
Rao, A.M. et al., Science, 259, 1993, 955.
Chauvert, 0. etal., Phys. Rev. Lett., 72, 1994,2721.
Iwasa, Y. et al., Science, 264, 1994, 1570.
Nunez-Regueiro, M. et al., Phys. Rev. Lett., 74,
Marques, L. etal., Phys. Rev., B54, 1996, R12633.
Sundar, C.S. et al., Phys. Rev., B53, 1996, 8180.
Rao, A.M. etal., Chem. Phys. Lett., 224,1994,106.
Rao, C.N.R. etal., J. Phys. Chem., 99,1995,16814.
Atta, M. et al., J. Phys. Chem., B101, 1997, 5.
Verheijen, M.A. et al., Chem. Phys., 166, 1992,
287; Vaughan, G.B.M. et al., Chem. Phys., 178,
Kawamura, H. ef al., Solid State Commun., 83,
Lundin, Solatov, A. and Sundqvist, B., Europhys.
Lett., 30, 1995, 469.
Premila, M. et al., Proc. DAE Solid State Physics
Symposium, BARC, Bombay, 39C, 1996, 342.
For a preliminary account of this, see, Sundar, C.S.
et al., in Proc. Conference on Frontiers in
Materials Modelling and Design, Kalpakkam,
Aug. 1996 (Edited by V. Kumar et at.). Springer-
Verlag (in print).
Hariharan, Y. et al., Curr. Sci., 63, 1992, 65.
Govindarajan, K. et al., Ind. J. Pure and Appl.
Phys., 27, 1989, 461.
Valsakumar, M.C. et al., Phys. Rev., B48, 1993,
19. Bethune, D.S. et al., Chem. Phys. Lett., 179, 1991,
20. Ragavachari, K. and Rohlfing, C.M., J. Phys.
Chem., 95, 1991, 5768; Jishi, R.A. et al., Chem.
Phys. Lett., 206, 1993, 187.
21. Venkateswaran, U.D. et al., Phys. Status Solidi,
198, 1996, 545; Rao, A.M. et al., Phys. Rev. B.
22. Varma, Vijay et al., Chem. Phys. Lett., 203, 1993,
23. van Loosdrecht, P.H.M. et al., Phys. Rev., B47,
24. Cotton, F.A., Chemical Applications of Group
Theory. Wiley-Eastern, New Delhi, 1970.
25. Martin, M.C. etal., Phys. Rev., B51, 1995, 3210.