Transmutation of fullerenes

Article abstract of DOI:10.1021/ja045521r

Fullerenes were pyrolyzed by subliming them into a stream of flowing argon gas and then passing them through an oven heated to ～1000°C. C ₇₆, C₇₈, and C₈₄ all readily lost carbons to form smaller fullerenes. In the case of C₇₈, some isomerization was seen. Pyrolysis of ³He@C₇₆ showed that all or most of the ³He was lost during the decomposition. C₆₀ passes through the apparatus with no decomposition and no loss of helium.

Full text of DOI:10.1021/ja045521r

Published on Web 02/12/2005
Transmutation of Fullerenes
R. James Cross* and Martin Saunders*
Contribution from the Department of Chemistry, Yale UniVersity, PO Box 208107,
New HaVen, Connecticut 06520-8107
Received July 26, 2004; E-mail: james.cross@yale.edu
Abstract: Fullerenes were pyrolyzed by subliming them into a stream of flowing argon gas and then passing
them through an oven heated to ∼1000 °C. C76, C78, and C84 all readily lost carbons to form smaller
fullerenes. In the case of C78, some isomerization was seen. Pyrolysis of 3He@C76 showed that all or most
3
of the He was lost during the decomposition. C60 passes through the apparatus with no decomposition
and no loss of helium.
Introduction
The formation of fullerenes by laser ablation of graphite or
in a carbon arc is a complicated process. Initially, carbon atoms
are produced, and they condense to form fullerenes, carbon
nanotubes, and other less well characterized products. Although
there has been much discussion about how these conversions
might occur, it is fair to say that we do not know the detailed
mechanisms. However, the products do not appear to be at
equilibrium. For example, much more C60 is formed than C70.
Both theory and experimental thermodynamic results lead one
to expect at least as much C70 to be present at equilibrium. Very
high energy species, such as those which violate the isolated
pentagon rule (IPR), are not found. Within each isomeric class
of the higher fullerenes, the ratios of amounts seem to be much
closer to the expected equilibrium distributions. It might well
be that after the distribution of fullerene sizes is established,
there is some kind of annealing process that interconverts
isomeric structures. In an effort to detect such processes and to
probe the latter stages of this complex sequence of reactions,
we have heated fullerenes in the gas phase to study their
reactions. To our surprise, we have found loss of carbon atoms
at temperatures above 1000 °C as well as conversion of one
isomer to another.
Figure 1. Schematic of the apparatus. The long tube is made of fused
silica (quartz), 9 mm od and 7 mm id. It passes through two tube furnaces.
The sample is placed inside the tube in the middle of the first oven. It
sublimes into the argon stream and is carried through the second oven,
condensing on the far side.
products are dissolved and analyzed by high-pressure liquid
chromatography (HPLC), using a PYE column.
Experimental Details
Higher fullerenes were purchased from MER Corp. as a mixture.
They were separated with HPLC using a semipreparative PYE column
[2-(1-pyrenyl)ethylsilyl] (Cosmosyl 10 × 250 mm), with toluene as
the eluent. The mixture contained a small amount of C70, larger amounts
of C76 and C78, and some C84. C70 and C76 are cleanly separated. C78
consists of three isomers that elute in two incompletely separated peaks,
which we refer to as C78A and C78B. Previous work,1 using 13C NMR,
has shown that C78A is a mixture of one of the C2V isomers and the D3
isomer, while C78B is the other C2V′ isomer and a small amount of the
C2V isomer. C84 is notoriously hard to resolve by straight HPLC
analysis.2,3 It comes off our HPLC as a broad peak containing a mixture
of several isomers which blend into each other.
The fused silica tube is 9 mm od and 7 mm id. We introduce the
fullerene sample as a solution in CS2 into the tube using a long
hypodermic needle and then evaporate the CS2. The flow rate of argon
is roughly 20 mL/s. The length of the second oven is 38 cm, so that
the residence time at high temperature is about 40 s. Since the ends of
The current price of C60 is very roughly the price of gold.
The price of C70 is about an order of magnitude larger, and that
of higher fullerenes very much higher. A process to interconvert
fullerenes could be useful in making higher fullerenes available
at a reasonable price. So far, however, we have succeeded in
the alchemist’s equivalent of transmuting gold into lead.
The apparatus is shown in Figure 1. A long tube made of
fused silica (quartz) is attached to an argon cylinder and passes
through two ovens. A flow meter at the exit end measures the
flow rate. A few milligrams of the fullerene sample are placed
inside the tube in the middle of the first oven. At temperatures
of 600-700 °C, the fullerene sublimes into the argon stream
and is carried through the second oven. Reaction products then
condense in the bend after the second oven. After the run, the
(1) Sternfeld, T.; Saunders, M.; Cross, R. J.; Rabinovitz, M. Angew. Chem.,
Int. Ed. 2003, 42, 3136-3139.
(2) Dennis, T. J. S.; Kai, T.; Tomiyama, T.; Shinohara, H. Chem. Commun.
1998, 619-620.
(3) Dennis, T. J. S.; Kai, T.; Asato, K.; Tomiyama, T.; Shinohara, H.; Yoshida,
T.; Kobayashi, Y.; Ishiwatari, H.; Miyake, Y.; Kikuchi, K.; Achiba, Y. J.
Phys. Chem. A 1999, 103, 8747-8752.
9
3044
J. AM. CHEM. SOC. 2005, 127, 3044-3047
10.1021/ja045521r CCC: $30.25 © 2005 American Chemical Society

Transmutation of Fullerenes
A R T I C L E S
Figure 2. HPLC traces for C76. The top panel shows the starting material.
In the middle panel, the second oven is kept at 650 °C, the same temperature
as that in the first oven. In the bottom panel, the second oven is at 1110
°C, and extensive decomposition occurs.
Figure 3. HPLC traces for C78A. The original sample contains a small
amount of C76 due to incomplete separation of the original mixture.
the oven are colder than the middle and since the flow rate fluctuates,
this time is only approximate. The ovens are tube furnaces made by
Omega Engineering and by Thermolyne. The temperatures are con-
trolled by Omega controllers using Chromel-Alumel thermocouples
positioned in the middle of the ovens.
After the run, the products are dissolved in CS2 and reinjected into
the HPLC. Much of the original fullerene sample is left unsublimed at
the beginning of the tube. Crystals of fullerene almost always contain
solvent molecules as well as small amounts of oxygen. When heated
to the temperatures required for sublimation, reactions occur that
produce polymeric, nonvolatile products, so that only about 25%
actually sublimes. We frequently find a trace of black residue at the
end of the tube which is insoluble in CS2. After removal of the product,
the tube is filled with oxygen and flamed with a glass-blowing torch
to burn out any residue. The process is then repeated to ensure a clean
surface in the reactor for the next run.
Results and Discussion
Figure 2 shows the results for C76. The top panel is the HPLC
trace of the starting material. In the middle panel, the second
oven is held at 650 °C, the same temperature as that in the first
oven. The C76 emerges unchanged. In the bottom panel, the
second oven is at 1110 °C. Here, there is extensive decomposi-
tion to form C70 and a small amount of C60. The detector on
the HPLC apparatus uses visible light at 400 nm. By injecting
known amounts of C70 and C76, we calibrated the detector. The
molar extinction coefficient for C70 is 20% lower than that for
C76, so roughly a third of the C76 has decomposed. The
temperature is critical. If the oven is 50 °C colder, there is little
decomposition; at 50 °C hotter, most of the C76 is gone.
Figure 4. HPLC traces for C78B. The original sample contains a small
amount of C78A.
Figures 3 and 4 show the results for the two peaks for C78:
C78A and C78B. As the top panels show, the samples are not
9
J. AM. CHEM. SOC. VOL. 127, NO. 9, 2005 3045

A R T I C L E S
Cross and Saunders
Table 1. Calculated Energies
namic data. At 1050 °C and 1 atm, ∆G ) 681 kJ/mol (163
kcal/mol). This corresponds to an equilibrium constant of 1 ×
10-27 (using a standard pressure of 1 atm), far too small for the
reaction to be important.
species
symmetry
E/C (kJ/mol)a
C2
C6
D∞h
D∞h
Ih
D5
D2
D3h
D3h′
D3
426
167
0.0
-3.8
-4.6
-5.2
-4.1
-4.9
-5.2
C60
C70
C76
C78
C78
C78
C78
There are two ways to overcome the large endothermicity.
One is a radical chain reaction. A small amount of decomposi-
tion could give rise to small fragments that then add to other
fullerene molecules causing them to decompose as well.
Alternatively, the reaction could occur on the surface of the
fused silica tube. The binding energy of the fragment to the
silica surface greatly reduces the endothermicity. We ran the
reaction with C76 as before, except that the tube was loosely
packed with quartz wool, greatly increasing the surface area. If
the reaction goes by way of radical intermediates, the rate should
decrease substantially since collisions with the surface should
remove the radicals. On the other hand, if the reaction occurs
on a surface, the rate should increase. The experimental results
were unchanged.
C2V
a Energies are calculated as energy per carbon atom, relative to C60.
initially pure, due to incomplete separation in the initial HPLC
run. C78A has a small amount of C76, and C78B has a small
amount of C78A. Again, there is extensive decomposition to
smaller fullerenes: C76, C70, and even C60. In the case of C78A,
there is apparently a small amount of isomerization to C78B
since the size of the C78B peak is significantly larger than it is
in the starting sample.
One mechanism, which avoids these problems, is where two
fullerene molecules collide, join, and then separate into a larger
and a smaller fragment. The problem with this mechanism is
that we have never seen any fullerenes larger than the starting
material. Of course, these larger species could be, or ultimately
form, the insoluble residue that we find at the end of the tube.
To distinguish between a unimolecular and a bimolecular
reaction, we ran two samples of C76 under conditions as closely
matched as possible. One sample was four times as large as the
other. If the reaction is unimolecular, the fractional decomposi-
tion should remain unchanged. If it is bimolecular, the decom-
position should increase. The larger sample gave much less
decomposition than the smaller sample, 7 versus 50%. Clearly,
the mechanism is more complicated than a simple unimolecular
or bimolecular reaction.
We ran the first part of the peak for C84 at 1100 °C. As in
the other cases, there was extensive decomposition to C78A,
C76, C70, and a small amount of C60. There also appeared to be
rearrangement to isomers appearing later in the HPLC separa-
tion. We ran C70 at 1110 °C and found no trace of C60. Since
our oven is rated at 1200 °C, we did not attempt a higher
temperature. Since we saw small amounts of C60 formed at lower
temperatures from C76, C78A, and C78B, it would appear that
this C60 cannot have been formed by a process that passes
through C70 as an intermediate.
There are several possible mechanisms for the observed
reactions. The simplest is that the fullerene molecule eliminates
one or more carbon fragments, such as C2 or C6. To test this,
we ran calculations using Gaussian 20034 using the B3LYP
method and a 6-31G(d,p) basis set. The results are shown in
Table 1. They are tabulated in energy per carbon atom relative
to C60 to facilitate comparison. We see that the reaction
Our group has developed techniques for putting noble gases
inside fullerenes,5,6 so it is natural to ask if the decomposition
reactions open up a hole in the fullerene large enough to allow
a helium atom to escape. We labeled a sample of higher
C78 (C2V) f C76 (D2) + C2
(1)
3
fullerenes with He using our high-pressure method. Since the
yield of soluble product is low for higher fullerenes, we did
not use the cyanide catalyst which works so nicely for C60.7
The labeled fullerenes were separated, as before, with HPLC.
is endothermic by 905 kJ/mol (216 kcal/mol), and that the
reaction
3
3
We then ran He-labeled C76. A He NMR spectrum showed
C76 (D2) f C70 (D5h) + C6
(2)
only the peaks for 3He@C76 and 3He2@C76, both seen before,8
3
but no signal for He@C70. The sample was separated with
is endothermic by 1090 kJ/mol (260 kcal/mol). Reactions that
are this endothermic are unlikely to occur to any extent in the
gas phase. However, both reactions are favored in terms of
entropy. The calculations for reaction 1 were extended to
calculate the vibrational frequencies and, hence, the thermody-
HPLC and clearly showed the presence of C70 and a trace
amount of C60, although the extent of decomposition was not
as much as that shown in Figure 2. From the peak area in the
HPLC trace and our calibration of the detector sensitivity to
C
70 and C76, we can predict both the position and the magnitude
of the expected NMR signal for 3He@C70. This predicted value
is approximately four times the noise level. Thus, most or all
the helium escaped in the decomposition reaction. In a similar
(4) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,
J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.;
Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo,
J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 2003;
Gaussian, Inc.: Pittsburgh, PA, 2003.
4
experiment, we ran He@C60 through the reactor at 1100 °C.
(5) Saunders, M.; Cross, R. J.; Jime´nez-Va´zquez, H. A.; Shimshi, R.; Khong,
A. Science 1996, 271, 1693.
(6) Saunders, M.; Cross, R. J. In Endofullerenes: A New Family of Carbon
Clusters; Akasaka, T., Nagase, S., Eds.; Kluwer: Dordrecht, The Nether-
lands, 2002; Vol. 3, pp 1-11.
(7) Cross, R. J.; Khong, A.; Saunders, M. J. Org. Chem. 2003, 68, 8281-
8283.
(8) Wang, G.-W.; Saunders, M.; Khong, A.; Cross, R. J. J. Am. Chem. Soc.
2000, 122, 3216-3217.
9
3046 J. AM. CHEM. SOC. VOL. 127, NO. 9, 2005

Transmutation of Fullerenes
A R T I C L E S
The fraction of C60 containing a helium atom (about 1%) was
identical to that of the starting material. We have evidence7 that
when we put a noble gas atom inside C60 at 650 °C, some
promoter, such as cyanide, is added reversibly to the C60 to
open up the cage. We do not understand the detailed mechanism
here either, but perhaps the two are related.
We set out to investigate the late reactions involving the
formation of fullerenes in a carbon arc. We find at tempera-
tures above 1000 °C that carbons are easily lost, and that
rearrangements of isomers occur, but the mechanism is clearly
complicated.
Acknowledgment. We thank Mr. Christopher Stanisky for
taking the 3He NMR spectrum. We are grateful to the National
Science Foundation for the support of this research under Grants
CHE-0071379 and CHE-0307168.
JA045521R
9
J. AM. CHEM. SOC. VOL. 127, NO. 9, 2005 3047