F. Cataldo / Tetrahedron 60 (2004) 4265–4274
4271
the total concentration of the polyynes in n-hexane prepared
with Ti electrodes under similar conditions (same
current intensity and arcing time). Thus, the fundamental
contribution from the graphite electrodes in supplying
elemental carbon for the polyynes production is clear.
Furthermore, it is remarkable that with graphite electrodes
long chain polyynes C12H2, C14H2 and C16H2 were also
formed and detected in appreciable amounts, while with Ti
electrodes the longest detectable chain was C10H2 (see
Table 1).
To explain the various results in different solvents as
reported in Table 1, it is useful to recall a petrochemical
process known as ‘coking’ which involves the thermal
carbonization of an hydrocarbon22,23 (or hydrocarbon
mixture). The coke produced can be considered equivalent
to the pyrocarbon mentioned above. The coking tendency of
an hydrocarbon under pyrolytic conditions depends on its
C/H ratio and the presence of oxygen in the molecule but
essentially it depends from its enthalpy of formation.22,23
Thus, the energy necessary to produce 1 kg of carbon is
þ2325 kJ/kg in the case of n-hexane and becomes
þ9787 kJ/kg in the case of ethanol but reaches the very
high value of þ16770 kJ/kg in the case of methanol. In the
case of aromatic hydrocarbons, coking is an exothermic
process and occurs with the evolution of considerable
amount of heat. For instance, in the case of benzene, the heat
evolved in the process is 21150 kJ/kg, for naphthalene
21270 kJ/kg. The maximum value of heat emission in the
coking process is offered by acetylene: 29450 kJ/kg.22 In
these last three cases the coking process is spontaneous due
to the exothermicity of the process. Therefore, when arcing
is conducted with aromatic hydrocarbons like benzene (or
toluene8–10), because of their spontaneous tendency to
carbonization, the formation of PAH mixtures in relatively
large quantity can be observed together with pyrocarbon or
coke. Instead the amount of PAHs and coke decreases
significantly when n-hexane is arced and decreases further
when methanol is used in place of n-hexane for arcing. This
because of the unfavourable thermodynamics in the
carbonization process. As shown above the carbonization
thermodynamics of alcohols are extremely unfavourable
also because these molecules contain oxygen. Thus, in the
case of alcohols, PAH formation is extremely low to
negligible and similarly can be concluded for pyrocarbon
formation in these media.
Approximately similar results were observed when
n-hexane was replaced with benzene (Table 1): graphite
electrodes produced a higher concentration of polyynes
with a wider distribution of detectable chains in
comparison to titanium electrodes. A striking peculiarity
of benzene is the formation of a plethora of HPLC-DAD
detectable PAHs as reported in Table 1 (plus others
separated by HPLC but not identified on the basis of the
electronic absorption spectra), the distribution of the PAHs
was richer and more complete with graphite electrodes
rather than with Ti electrodes. In this case, the contribution
of the solvent to PAHs and pyrocarbon formation appears
quite evident.
Completely opposite results to those discussed in the case of
n-hexane and benzene have been observed in the case of
alcohols like methanol and ethanol (Table 1).
When methanol or ethanol was used as solvent for arcing
with Ti electrodes, the formation of carbon black derived
from solvent carbonization was reduced to a minimum.
Simultaneously, the HPLC-DAD analysis revealed the
complete absence of any polyynes in these oxygenated
solvents. Instead, only PAHs were detected but in trace
amount (Table 1). Among the PAHs detected in methanol,
biphenyl, naphthalene, acenaphthalene, phenanthrene,
anthracene were easily identified based both on their
retention times and on their peculiar UV spectral pattern
in comparison with the spectral pattern of authentic
reference compounds. Perylene and fluoranthene were also
reasonably identified based on the retention time and on the
reasonable match of reference UV spectra. In any case, the
PAHs formed under these conditions were present in at least
two orders of magnitude lower concentration than the
polyynes formed by arcing the graphite electrodes in
methanol. Moreover, PAH formation in methanol was
considerably lower than the trace amounts produced in
n-hexane.
3. Summary and conclusions
With a series of experiments, we have thrown more light on
the understanding of the process of polyyne formation by
the submerged electric arc between graphite electrodes.
Arcing graphite electrodes in distilled water produces
hydrogen-terminated polyyne chains H–(CuC)n–H, and
this demonstrates that the hydrogen is coming from water
plasmalysis at the arc temperature while carbon is vaporized
from the graphite electrodes.
Arcing of graphite electrodes in methanol or even better in
ethanol, produces the cleanest polyyne solution with
extremely small to negligible amounts of PAHs (Table 1).
Thus, these alcohols appear to date the best solvents for the
cleanest synthesis of polyynes with the electric arc
technique. As already reported for other solvents, also in
methanol and ethanol the dominant polyyne is C8H2,
accounting for more than 70 mol percent of the total
polyyne mixture. The yield of longer polyynes decreases
as the chain length grows (Table 1). Polyynes C10H2 and
C12H2 were formed in appreciable amounts together
with the C6H2. In ethanol also C14H2 and C16H2 were
detected.
Arcing graphite electrodes in liquid nitrogen with our
simple apparatus permits trapping of the polyynes formed
which consist essentially of cyanopolyynes NuC–
(CuC)n–CuN. Under the high arc temperature molecular
nitrogen is divided into atomic nitrogen, which terminates
the ends of the polyyne chains formed by the association
of elemental carbon vapour released by the graphite
electrodes. Together with the cyanopolyynes, normal
hydrogen-terminated polyynes have also been detected in
the mixture produced by arcing in liquid nitrogen. The
formation of the hydrogen-capped polyynes can be
explained because of the presence of humidity and traces
of water in the liquid nitrogen and in the reactor.
Cataldo, Franco
