In situ laser-furnace TOF mass spectrometry of C36 and the large-scale production by arc-discharge

Article abstract of DOI:10.1021/jp001263i

A greatly improved synthesis of C₃₆, obtained as its hydrides C₃₆H₄ and C₃₆H₆ and oxyhydrides C₃₆H₄O and C₃₆H₆O, via a high- temperature laser vaporization (laser-furnace) method and the conventional DC arcdischarge method is reported. We have found that doping metal catalysts such as Ni/Co or Ni/Y into the laser target graphite rod or the DC arc-discharge electrode, respectively, significantly enhanced the production of C₃₆ species; the methods have been known for efficient production methods of single-wall carbon nanotubes (SWNTs). The results suggest that the catalytic formation of C₃₆ species and SWNTs is closely related with each other in their early stage of formation. Furthermore, a molecular C₃₆H₆ gives C₁₈H₃ (in/z = 219) as a magic fragment upon desorption/ionization. By using a high-performance liquid chromatography (HPLC),. C₃₆H₆ was purified (99%) in milligram quantity for the first time. The purified C₃₆H₆ exists as a clusterassembled material such as oligomers or clusters in the solid state, not as a molecular crystal such as C₆₀.

Full text of DOI:10.1021/jp001263i

7908
J. Phys. Chem. B 2000, 104, 7908-7913
In Situ Laser-Furnace TOF Mass Spectrometry of C36 and the Large-Scale Production by
Arc-Discharge
Akira Koshio, Masayasu Inakuma, Zhong W. Wang, Toshiki Sugai, and Hisanori Shinohara*
Department of Chemistry, Nagoya UniVersity, Nagoya 464-8602, Japan
ReceiVed: March 31, 2000; In Final Form: June 14, 2000
A greatly improved synthesis of C36, obtained as its hydrides C36H4 and C36H6 and oxyhydrides C36H4O and
C36H6O, via a high- temperature laser vaporization (laser-furnace) method and the conventional DC arc-
discharge method is reported. We have found that doping metal catalysts such as Ni/Co or Ni/Y into the laser
target graphite rod or the DC arc-discharge electrode, respectively, significantly enhanced the production of
C36 species; the methods have been known for efficient production methods of single-wall carbon nanotubes
(SWNTs). The results suggest that the catalytic formation of C36 species and SWNTs is closely related with
each other in their early stage of formation. Furthermore, a molecular C36H6 gives C18H3 (m/z ) 219) as a
magic fragment upon desorption/ionization. By using a high-performance liquid chromatography (HPLC),
C36H6 was purified (99%) in milligram quantity for the first time. The purified C36H6 exists as a cluster-
assembled material such as oligomers or clusters in the solid state, not as a molecular crystal such as C60.
Introduction
tential of a solid C36 and predicted the possibility of a super-
conducting transition temperature larger than those of alkali-
doped C60 solids.17,18 Various experimental attempts to reproduce
the production of C36 by the DC arc-discharge method, however,
has not so far been successful, probably because the optimum
arc-discharge and sublimation conditions on the production and
purification of C36 are still not known.
Recently, we have reported a high-yield and preparative-scale
synthesis of C36 with purification and identification of C36
hydrides and the oxyhydrides (C36H6 and C36H6O) based on
the high-temperature laser-vaporization (laser-furnace) method.19
In this study, a distinct signal corresponding to C36H4 was
observed by laser desorption/ionization time-of-flight mass
spectrometry (LD TOF-MS) in the soot produced by the laser-
furnace method. We have found that the metal catalysts such
as Ni/Co and Ni/Y significantly enhance the production of C36
species. It is well-known that these metals can enhance the
production of single-wall carbon nanotubes (SWNTs).20 It was
also reported that purified C36 materials exist as a cluster-
assembled material in the solid state, not as a molecular form
such as C60.
Here, we report the first successful production and charac-
terization of C36-related materials (C36H4, C36H4O, C36H6, and
C36H6O) by a new system (the laser-furnace LD TOF mass
spectrometer) developed in this laboratory. We also report a
large-scale synthesis of C36H6 via the DC arc-discharge method
with metal-doped graphite anodes. A matrix-assisted laser
desorption/ionization (MALDI) TOF mass spectrometry with
cobalt ultrafine particles (Co-UFP) has revealed that C36H6 tends
to dissociate into C18H3 as a magic fragmentation on desorption/
ionization.
C60 has been considered as the smallest fullerene having the
cage-structure that can be produced, separated, and isolated in
a macroscopic quantity ever since Kra¨tschmer and Huffman’s
first macroscopic production of fullerenes.1 This is largely
because of the fact that C60 is the smallest fullerene which
satisfies the so-called isolated pentagon rule (IPR).2 Some
fullerenes and metallofullerenes smaller than C60, such as C20,
C28, and M@C28(M ) U, Ti, etc.), were found to be stable in
gas-phase cluster-beam experiments3, and their structures have
been predicted by theoretical calculations.4-6 However, the bulk
production and the isolation of these materials have not been
successful, and thus the extensive investigations of these clusters
have not been done yet.
The presence of C36 has also been predicted as one of the
smaller fullerenes. C36 and the other carbon clusters, C2n (2 e
2n e 100), have been observed experimentally in jet-cooled
cluster beam experiments by Cox and co-workers7 and Smalley
and co-workers.8 The structures of C36 and other carbon clusters
have been investigated by gas-phase mobility measurements by
ion chromatography,9,10 but the detailed structure and properties
of C36 have not yet been clarified.
It was not until quite recently that the first fullerene smaller
than C60, C36, has been produced by the electric DC arc-
discharge method by Zettl and co-workers.11 They reported that
a solid C36 was produced in a macroscopic quantity by the DC
arc-discharge method with a graphite electrode in a 400 Torr
helium atmosphere. They reported that the produced soot
contained C36. Their study suggested the presence of a fullerene-
like caged C36 which exists as a covalently bonded cluster-
assembled material in the solid state. Theoretical calculations12-15
also suggested that C36 form stronger inter-cage bonds than the
normal fullerenes and the presence of various types of C36-based
Experimental Section
-
2-
High-Temperature Laser Vaporization (Laser-Furnace)
TOF Mass Spectrometer. The high-temperature laser vaporiza-
tion (laser-furnace) method was originally developed by Smalley
and co-workers for production of fullerenes and metallo-
fullerenes.21 The laser-furnace method is also a useful method
oligomers in the solid state. The stability of C36 and C36
was also studied theoretically.16 Furthermore, Cohen and co-
workers have calculated the electron-phonon interaction po-
* To whom correspondence should be addressed.
10.1021/jp001263i CCC: $19.00 © 2000 American Chemical Society
Published on Web 07/21/2000

TOF Mass Spectrometry of C36
J. Phys. Chem. B, Vol. 104, No. 33, 2000 7909
Figure 1. A schematic diagram of the high-temperature laser vaporiza-
tion LD TOF mass spectrometer.
for SWNT production when one employs a Ni/Co/graphite
composite rod as a target rod.20 However, details of the laser-
furnace products besides SWNTs have not been thoroughly
investigated to date.
Investigations of SWNT intermediates and its precursors are
particularly interesting in understanding the growing mechanism
of SWNTs. We have developed a new apparatus to produce
and analyze soot containing C36 materials,19 which consists of
a laser-furnace combined anaerobically with a LD TOF mass
spectrometer. Figure 1 shows a schematic of the apparatus: the
high-temperature laser vaporization LD TOF mass spectrometer.
It consists of an electric furnace with a target rod in a quartz
tube (28 mm inner diameter), a transferring chamber of a sample
plate (15 × 15 mm) on which soot deposits, and a linear
reflectron TOF mass spectrometer. Ni(0.6 at. %)/Co(0.6 at. %)-
or Ni(4.2 at. %)/Y(1.0 at. %)-doped graphite composite rods
(Toyo Tanso Co. Ltd., 5 × 5 × 20 mm) were used for target
rods.
The second-harmonic output of a Nd:YAG laser (532 nm,
10 Hz, 6-7 ns, 600 mJ/pulse, ca. 1 mm diameter spot; Spectra-
Physics GCR-250) was used to vaporize the metal-doped
graphite composite rod at a temperature range of 30-1000 °C
in an Ar flow (1-760 Torr, 300 mL/min) condition. Special
precaution was taken to vaporize the fresh surface of the target
rod.
The soot containing C36 species together with fullerenes and
SWNTs were collected on a water-cooled nickel plate. The plate
was anaerobically transferred to the acceleration region of the
LD TOF mass spectrometer, which enable us to perform in situ
TOF mass analyses. The mass analysis was done in the LD
TOF MS chamber (∼10-7 Torr). The linear reflectron-type LD
TOF mass spectrometer employs vertical laser desorption. The
third-harmonic output of the other Nd:YAG laser (355 nm, 2
Hz, HOYA Continuum MINILIGHT II) desorbs and ionizes
the soot product on the plate. The pulse sequence, such as the
delayed time and the acceleration time of ions, was controlled
by a pulse generator (SRS Inc. DG535). The acceleration voltage
was typically set at 2400 V. The time-of-flight signals were
analyzed by a transient digitizer (LeCroy 9400 A). The
conversion from time-to-mass scale and the analysis of the
obtained TOF spectra were performed on a personal computer
(NEC PC9801-VX).
Figure 2. In situ mass spectra of the laser-furnace products. The
production conditions are (a) pure graphite target rod, 500 Torr Ar;
(b) Ni/Co/C composite target rod, 100 Torr Ar; (c) Ni/Co/C composite
target rod, 500 Torr Ar.
with an anaerobic sampling/collection mechanism are de-
scribed elsewhere.22-25 Similar Ni/Y- or Ni/Co-doped graphite
rods as those used for the laser-furnace experiments were used
as the anode. The gap of the anode (15 × 15 × 250 mm) and
the cathode (a carbon block) was typically set to 1 mm. The
arc was sparked at 300 A in a He atmosphere of 400 Torr.
The soot containing C36 species was extracted by solvents such
as toluene and CS2. The extracts were analyzed by a home-
made LD TOF mass spectrometer and a mass spectrometer
(SHIMADZU KOMPACT MALDI IV) at 337.1 nm. The soot
containing SWNTs was also characterized by a scanning
electron microscope (SEM) (JEOL JSM 6340F).
The separation and isolation of C36 species from fullerenes
was done by a high-performance liquid chromatography (HPLC)
(UV detection at 280 nm: JASCO PU-987, UV-970) with a
PBB column (φ 20 × 250 mm, nacalai tesque) with a 100%
CS2 eluent. The separation and purification were done at 10
mL/min and 5 mL/min flow rate, respectively.
Results and Discussion
Observation of C36 Species Produced by the Laser-
Furnace Method. Figure 2 shows in situ LD TOF positive mass
spectra of soot products produced by the laser-furnace method.
Figure 2a-c was obtained under the conditions of pure graphite,
Ni/Co-doped graphite (Ar 100 Torr), and Ni/Co-doped graphite
(Ar 500 Torr), respectively. The temperature of the furnace was
set at 1000 °C. Only fullerene signals corresponding to C60,
A Large-Scale Synthesis of C36 Species by the DC Arc-
Discharge Method. The details of the arc-discharge apparatus

7910 J. Phys. Chem. B, Vol. 104, No. 33, 2000
Koshio et al.
Figure 4. A mass spectrum of the CS2 extract of the arc-discharge
products with a Ni/Y/C composite electrode.
A similar enhancement of C36H6 and C36H6O has been
observed for Ni(4.2 at. %)/Y(1.0 at. %)-doped graphite rods
but was not observed for Fe- and Ti-doped graphite rods.
Namely, the formation of C36 species is efficiently catalyzed
by exactly the same metal catalysts as in those for SWNTs.
Furthermore, other laser-furnace conditions, such as laser
fluence, Ar pressure, and furnace temperature, for the optimum
production of C36 species are the same as those of SWNTs.
This may lead us to speculate that the early stage of the
formation of C36 species and that of SWNTs are closely related
with each other.
A Large-Scale Production of C36 Species by DC Arc-
Discharge. In this study, we have also found that the C36
species can be produced by the arc-discharge method of a metal-
doped graphite electrode similar to the laser-furnace method.
Figure 4 shows a positive LD TOF mass spectrum of a CS2
extract of a soot produced by arc-discharge of Ni/Y/C composite
electrode under a He atmosphere of 400 Torr. Signals corre-
sponding to fullerenes such as C60, C70, and C84 were detected.
The presence of an enhanced peak due to C36H6 is particularly
noticeable. In contrast to the laser-furnace case, the signal of
oxide species, C36H6O, was not observed in the arc-discharge
products.
Similar to the laser-furnace case, the optimum production of
C36 species by the DC arc-discharge was achieved when the
optimum production of SWNTs[Ni(4.2 at. %)/Y(1.0 at. %)]26
was obtained. In general, a soot containing SWNTs shows a
sheetlike or a weblike deposit. We found that the C36 species
coexist with SWNTs in a weblike soot (deposit). Figure 5 shows
a SEM image of the soot that contains the C36 species as
observed in Figure 4. The SEM image shows bundles of SWNTs
with diameters of ca. 10-100 nm. The formation of the thicker
bundles is due to an enhanced bundle formation after extraction
and purification.
Figure 3. Expanded mass spectra of the C36 species produced by the
laser-furnace method. The insert shows a calculated theoretical mass
distribution of C36H4.
C70, and some higher fullerenes were observed. Of particular
importance is the observation that the weak peaks at 436 and
452 amu (Figure 2b) were enhanced dramatically by using a
Ni(0.6 at. %)/Co(0.6 at. %)/C composite rod at Ar 500 Torr
(Figure 2c). The signal intensity of C36H4 (436 amu) was
enhanced 12 times as large as that in Figure 2b.
The intact C36 ion (432 amu) was not observed, and only the
hydrides and oxyhydrides were detected. C36 must be quite
reactive so that, immediately after its formation, C36 incorporates
four hydrogen atoms on the reactive sites even though only a
trace amount of hydrogen sources is present in the entire laser-
furnace LD TOF MS combined system.
Figure 3a shows a high-resolution mass spectrum of the C36
species. The observed ion intensity distribution is consistent with
that of the corresponding calculated theoretical (12C ) 98.9%,
13C ) 1.1%) distribution as shown in the inset. Interestingly,
C36H4 and C36H4O in the as-produced soot were converted
entirely to C36H6 and C36H6O, respectively, upon CS2 solvent
extraction (Figure 3b). The peak intensity of the oxide species,
C36H6O, has decreased after the extraction, probably because
that the solubility of the oxide species in CS2 is much less than
that of C36H6 species.
Figure 6 shows an HPLC chromatogram of the CS2 extract
which corresponds to Figure 4. The peak due to C36 species
(shaded area called as fraction 1) appears at ca. 6 min earlier
than C60 and C70. The C36 fraction was separated from C60, C70,
and the other higher fullerenes. Fraction 1 shows three sub-
fractions, 1-1, 1-2, and 1-3, in the second HPLC stage on a
PBB column as given in Figure 7.
The observed transformation from the four-hydrogenated
species, C36H4, to the six-hydrogenated species, C36H6, on the
extraction suggests that two reactive sites were newly formed
on C36H4, favorable to incorporate two additional hydrogen
atoms. This may be due to bond-breaking of the solid composed
of C36H4 upon CS2 extraction, which may lead to the formation
of a molecular C36H6 species in solution.
Figure 8a-c shows MALDI TOF mass spectra of the
fractions 1-1, 1-2, and 1-3, respectively, with a cobalt ultrafine
particle (Co-UFP)27 as a matrix. The C36 species was contained
in fraction 1-2. Fractions 1-1 and 1-3 mainly contain species
having 524 and 219 amu, respectively. At present, we think
the species due to 219 and 524 amu are reaction dissociation
products of C36 species.

TOF Mass Spectrometry of C36
J. Phys. Chem. B, Vol. 104, No. 33, 2000 7911
Figure 5. A SEM image of the purified soot produced by DC arc-discharge of a Ni/Y/C composite electrode. Bundles of SWNTs are also clearly
observed.
Figure 6. An HPLC chromatogram (10 mL/min flow rate) of the CS2
extract of the arc-discharge products with a Ni/Y/C composite electrode.
The shaded area shows the HPLC fraction containing C36 species.
Figure 8. Mass spectra of HPLC fractions: (a), (b) and (c) show
fraction 1-1, 1-2, and 1-3 in Figure 7, respectively.
TOF mass spectrum of the C36 species fraction (the shaded area
in Figure 6) without matrixes. Only a signal due to 219 amu
was observed, indicating that C36H6 dissociates almost com-
pletely to give the 219 amu species. The 219 amu species can
be assigned as C18H3, a half of C36H6. Obviously, C36H6 is
fragile upon desorption/ionization. In contrast, MALDI mass
Figure 7. An expanded HPLC chromatogram (5 mL/min flow rate)
of the C36 species fraction (the shaded area in Figure 6).
analysis of the same HPLC fraction by using Co-UFP and C60
as matrixes gave distinct peaks due to the parent C36H6 ions,
which is shown in Figure 9b,c, respectively. Evidently, the Co-
UFP and C60 matrixes play crucial roles in dissipating excess
A Magic Fragmentation of C36H6. During the mass spectral
analysis, we have found that C36 species have a strong tendency
to fragment. In fact, C36H6 was not observed by direct LD TOF
mass spectrometry without matrixes. Figure 9(a) shows a LD

7912 J. Phys. Chem. B, Vol. 104, No. 33, 2000
Koshio et al.
high reactivity. Recent theoretical calculations13,14 also suggest
such an oligomer formation of C36. The purified C36 species,
on the other hand, can exist as a molecular form in Co-UFP or
C60 matrix and does not bind together. Zettl and co-workers
reported29 that the pure C36 solid gave a very broad mass
distribution peaked at about C120 and that the K-doped C36 solid
showed a very sharp peak of C36 similar to the current MALDI
TOF mass spectrum with the Co-UFP. Of particular interest in
the present study is that C60 molecules prevent C36 species from
aggregation and that they disperse molecularly the C36 species
in the solid state or a concentrated solution.
Conclusions
We have found that C36 species were effectively produced
not only by the laser-furnace method but the DC arc-discharge
method with a metal-doped graphite composite rod containing
Ni/Co and Ni/Y, respectively. In particular, C36H6 was produced
by arc-discharge in a large scale. The production conditions for
the current C36 species are found to be similar to those of
SWNTs in terms of metal catalysts, temperature, and Ar (or
He) buffer gas pressure. A MALDI mass analysis reveals that
a molecular C36H6 gives C18H3 as a magic fragment upon
desorption/ionization.
Acknowledgment. H.S. thanks the Japanese Ministry of
Education, Science, Sports and Culture for Grants-in-Aid for
Scientific Research (B)(2)(No. 10554030) and JSPS for Future
Program on New Carbon Nano-Materials for the financial
support of the present study.
References and Notes
(1) Kra¨tschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffman, D. R.
Nature 1990, 347, 354.
Figure 9. Mass spectra of (a) purified C36H6, (b) purified C36H6
with Co ultrafine particles (Co-UFP) matrix,and (c) C36H6 with C60
matrix.
(2) Kroto, H. W. Nature 1987, 329, 529.
(3) Guo, T.; Diener, M. D.; Chai, Y.; Alford, M. J.; Haufler, R. E.;
McClure, S. M.; Ohno, T.; Weaver, J. H.; Scuseria, G. E.; Smalley, R. E.
Science 1992, 257, 1661.
(4) Feyereisen, M.; Gutowski, M.; Simons, J. J. Chem. Phys. 1992,
96, 2926.
(5) Guo, T.; Smalley, R. E.; Scuseria, G. E. J. Chem. Phys. 1993, 99,
352.
(6) Bylaska, E. J.; Taylor, P. R.; Kawai, R.; Weare, J. H. J. Phys. Chem.
1996, 100, 6966.
(7) Rohlfing, E. A.; Cox, D. M.; Kaldor, A. J. J. Chem. Phys. 1984,
81, 3322.
(8) Kroto, H. W.; Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley,
R. E. Nature 1985, 318, 162.
(9) Hunter, J.; Fye, J.; Jarrold, M. F. Science 1993, 260, 784.
(10) von Helden, G.; Hsu, M.-T.; Gotts, N.; Bowers, M. T. J. Phys.
Chem. 1993, 97, 8182.
(11) Piskoti, C.; Yarger, J.; Zettl, A. Nature 1998, 393, 771.
(12) Collins, P. G.; Grossman, J. C.; Coˆte´, M.; Ishigami, M.; Piskoti,
C.; Louie, S. G.; Cohen, M. L.; Zettl, A. Phys. ReV. Lett. 1999, 82, 165.
(13) Fowler, P. W.; Heine, T.; Rogers, K. M.; Sandall, J. P. B.; Seifert,
G.; Zerbetto, F. Chem. Phys. Lett. 1999, 300, 369.
(14) Heine, T.; Fowler, P. W.; Seifert, G. Solid State Commun. 1999,
111, 19.
Figure 10. A mass spectrum of a dried C36H6 solid prepared by
evaporating a purified CS2 extract.
(15) Jagadeesh, M. N.; Chandrasekhar, J. Chem. Phys. Lett. 1999, 305,
298.
(16) Ito, A.; Monobe, T.; Yoshii, T.; Tanaka, K. Chem. Phys. Lett. 1999,
315, 348.
(17) Coˆte´, M.; Grossman, J. C.; Cohen, M. L.; Louie, S. G. Phys. ReV.
Lett. 1998, 81, 697.
(18) Grossman, J. C.; Louie, S. G.; Cohen, M. L. Phys. ReV. B 1999,
60, 6941.
(19) Koshio, A.; Inakuma, M.; Sugai, T.; Shinohara, H. J. Am. Chem.
energies upon desorption/ionization. C60 has been also known
as an effective matrix for LD TOF MS.28 The C60 molecules,
which have a strong absorption at 328 nm close to the N2 laser
wavelength at 337 nm, effectively absorb the laser photons and
effectively transfer the corresponding photon energy to C36H6
A MALDI TOF mass spectrum of the purified solid-state
C36H6 with a higher laser fluence shows a series of peaks due
to the oligomers of C36 (i.e., C144-C216) species and/or their
fragments in addition to C36H6 (Figure 10). It is highly likely
that the purified C36H6 forms oligomers or clusters immediately
after the removal of solvent CS2 molecules owing to its very
Soc. 2000, 122, 398.
(20) Thess, A.; Lee, R.; Nikolaev, P.; Dai, H.; Petit, P.; Robert, J.; Xu,
C.; Lee, Y. H.; Kim, S. G.; Rinzler, A. G.; Colbert, D. T.; Scuseria, G. E.;
Toma´nek, D.; Fischer, J. E.; Smalley, R. E. Science 1996, 273, 483.
(21) Haufler, R. E.; Chai, Y.; Chibante, L. P. F.; Conceicao, J.; Jin, C.;
Wang, L. S.; Maruyama, S.; Smalley, R. E. Mater. Res. Soc. Symp. Proc.
1990, 206, 627.

TOF Mass Spectrometry of C36
J. Phys. Chem. B, Vol. 104, No. 33, 2000 7913
(22) Shinohara, H. Rep. Prog. Phys. 2000, 63, 843.
(23) Shinohara, H. AdV. Met. Semicond. Clusters 1998, 4, 205.
(24) Shinohara, H.; Takata, M.; Sakata, M.; Hashizume, T.; Sakurai, T.
Mater. Sci. Forum 1996, 232, 207.
(25) Takada, M.; Umeda, B.; Nishibori, E.; Sakata, M.; Saito, Y.; Ohno,
M.; Shinohara, H. Nature 1995, 377, 46.
(26) Journet, C.; Maser, W. K.; Bernier, P.; Loiseau, A.; Lamy de la
Chapelle, M.; Lefrant, S.; Deniard, P.; Lee, R.; Fischer, J. E. Nature 1997,
388, 756.
(27) Tanaka, K.; Waki, H.; Ido, Y.; Akita, S.; Yoshida, Y.; Yoshida, T.
Rapid Commun. Mass Spectrom. 1988, 2, 151.
(28) Hopwood, F. G.; Michalak, L.; Alderdice, D. S.; Fisher, K. J.;
Willett, G. D. Rapid Commun. Mass Spectrom. 1994, 8, 881.
(29) Zettl, A.; Piskoti, C.; Grossman, J. C.; Cohen, M. L.; Louie, S. G.
Electronic Properties of NoVel MaterialssScience and Technology of
Molecular Nanostructures; Kuzmany, H., Fink, J., Mehring M., Roth, S.,
Eds.; 1999; p 183.