On the Synthesis of Fullerenes

1
On the Synthesis of Fullerenes
N.R. Conley and J.J. Lagowski, Department of
Chemistry and Biochemistry, The University of
Texas at Austin, Austin, TX 78717 Telephone and
fax [phone], e-mail Lagarto2_at_aol.com
C60, also known as buckyball
C70, slightly larger than buckyball
Abstract
Experimental
Results
Discussion
In July of 1991, the potential of fullerene
combustion synthesis made itself known to the
world. Chemists working at the Massachusetts
Institute of Technology reported fullerene yields
up to 9 using low-pressure, premixed
benzene/oxygen/argon laminar flames 3.
Continuing the work of J.B. Howard, J.T.
McKinnon, Y. Markarovsky, A. Lafleur and M.
Johnson, we have investigated fullerene
production in high-temperature flames at
atmospheric pressure using various hydrocarbons,
heterocyclic compounds, and ferrocene. Mass
spectral data is presented.
The following table provides a list of the doping
compounds, the amounts used for soot formation,
the total amount of soot formed, and the
corresponding mass spectrum numbers for analysis
of the aromatic soot.
We have identified C60 and C70 in soot from a
benzene/oxy-acetylene flame (spectrum 1). While
we had no device to measure the temperature, NMDO
models predict the window of thermodynamic
stability for C60 to be between 2200 and 2600 K
at one atmosphere 2. We have also identified
C60 and C70 in soot from the combustion of
dicyclopentadiene (spectrum 2). However, the
lower peak intensities suggest that fullerenes
are formed to a lesser extent in the combustion
of dicyclopentadiene. When we injected a mixture
of benzene and dicyclopentadiene in the
stoichiometric ratio of ten moles of benzene to
three moles of dicyclopentadiene, resembling the
20 hexagon to 12 pentagon ratio of C60, we also
produced fullerenes. A wide band of peaks
appeared around 655 on this spectrum ( 2.1). In
spectrum 3a, obtained from the combustion of
pyridine, a similar band of peaks appears around
653. In addition, smaller groups of peaks are
observed at masses 722, 799, and 877. Spectrum
3b, the CI spectrum of pyridine soot, reveals a
large peak at 680. High-resolution mass
spectroscopic analysis resolves this peak to
679.5988. Elemental composition suggests the
presence of the compound C40H72N9 in this sample.
There is a relatively low-intensity peak at 679
on spectrum 4.1b, the negative CI spectrum of
soot formed from a mixture of thiophene and
benzene. In the combustion of a benzene solution
saturated with ferrocene, there are no peaks
indicative of an iron-encapsulated fullerene
(spectrum 5).
1(-)
2 (-)
Introduction
Since the introduction of Kratschmers graphite
vaporization method in 1990, the first procedure
for macroscopic synthesis of fullerenes, chemists
have worked diligently to modify these new
fullerene structures--trapping atoms, adding
elements, and attaching functional groups.
Despite all of the hard work, several ideal
modifications have eluded organic chemists for
years. It has come to the attention of the
authors that fullerene combustion synthesis
offers the advantage of flame-doping with the
possibility of modifying the fullerene products,
an area which has not been thoroughly explored.
Below is a table with the compounds we introduced
into a hot, non-sooting oxy-acetylene flame, the
method by which we introduced them, and the
products we intended to form.
2.1 (-)
3a (-)
3.1 (Ø)
 
 
Conclusion
3b ()
3b ()
Until these experiments can be carried out at
lower pressures, producing larger fullerene
yields that will allow for separation 2, it is
difficult to make conclusions concerning the
products that are formed. However, peak
intensities from spectra 1 and 2 suggest that
it is easier to form the five-membered carbon
rings necessary in the mechanism for fullerene
formation from six-membered carbon rings than it
is to form six-membered carbon rings from
five-membered carbon rings. The majority of the
products in these experiments are large
polycyclic aromatic hydrocarbons (PAH), often
referred to soot. PAH are believed to be
precursors in fullerene formation 3. The large
peak at 655 in spectrum 2.1 may represent a
stable PAH intermediate in the formation of
fullerenes. In the spectra of soot obtained from
the combustion of pyridine or thiophene, there
are no peaks with an isotopic distribution that
would suggest a fullerene structure, although the
presence of these large PAH indicates the
potential for their formation. The
ferrocene-doping of a benzene flame seems to
suppress the formation of C60 and C70. Again
however, due to the presence of these PAH, it is
not possible to rule out the potential of
combustion synthesis in forming an
iron-encapsulated fullerene.

• Neither spectrum is provided.
• Only the positive CI spectrum is provided.
• Only the negative CI spectrum is provided.
• The high-resolution mass spectrum is provided.
• SPECTRA NOT PUBLISHED ON POSTER ARE AVAILABLE FOR
  VIEWING UPON REQUEST.

4.1a ()
4.1b (-)
References
Procedure
1. Howard, J.B., McKinnon, J.T., Makarovsky, Y.,
Lafleur, A.L., and Johnson, M.E., Nature 352
139-141 (1991). 2. McKinnon, J.T., and Bell,
W.L., Combustion and Flame 88 102-112
(1992). 3. Bachmann, M., Wiese, W., and Homann
K.-H., Twenty Sixth Symposium (International)
on Combustion, The Combustion Institute, 1996,
pp. 2259-2267.
To achieve the high temperature necessary for
fullerene combustion synthesis, we used a Victor
Medalist XL oxy-acetylene torch with a brazing
tip. The hydrocarbons/heterocyclic compounds
were injected, 10 mL per experiment, into the
non-sooting oxy-acetylene flame using a glass
syringe. Preliminary experiments in benzene
combustion revealed the optimal oxy-acetylene
flame conditions at atmospheric pressure 19 cm.
flame length, a cone 4 cm. in length, and
flame/stainless steel interaction at 14 cm. below
the torch tip. Soot from these experiments was
collected on a 19 cm. x 24 cm. x ¾ cm. stainless
steel plate that was placed on top of a
water-cooled brass block. The cooling block
allowed the stainless steel plate to remain below
800 K, the temperature at which C60 sublimes.
Because of its low boiling point and its ability
to dissolve fullerenes, CS2 was chosen as a
solvent. After the experiments, the soot was
scraped off of the stainless steel plate, 10 mL
of CS2 were added, and the samples were placed in
an ultrasonic bath for at at least thirty
minutes. Following ultrasonication, the samples
were filtered and the filtrate was collected.
After removing enough solvent to concentrate the
samples, they were sent for mass spectral
analysis using chemical ionization.
5 (-)

Special Thanks
The authors would like to extend special thanks
to all of those involved in this project. First
and foremost, we acknowledge Robin Rogers,
Longfei Jiang, and Dr. Mehdi Moini in the
Department of Mass Spectrometry for their
diligent work in sample analysis. Also, we would
like to thank Robert A. Lewandowski III for his
unequivocally skillful glassblowing and all of
the machinists whose insight and hard work has
made an enormous impact on our continuing
research. We are thankful for the help of Michael
Klysik, who has assisted in a portion of this
research. We are forever indebted to Rita
Wilkinson for her timely filing of all forms
necessary to make possible our trip to San
Francisco, CA. Finally, we would like to
recognize the Welch Foundation for their funding
of this research.
oxy-acetylene torch/injector set-up