On assembling polychlorinated aromatic hydrocarbons from carbon tetrachloride via dichlorocarbene intermediary by a solvothermal reaction: A reaction pattern from carbene-ylide interconversion

Article abstract of DOI:10.1021/jo048190+

(Chemical Equation Presented) The forced one-electron reduction of carbon tetrachloride with sodium in a sealed steel vessel is shown to have a narrow window of conditions to arrest the reaction at the polychlorinated aromatic hydrocarbons (PCAHs), as well as to prevent the reaction from proceeding all the way to the final stage of graphite and other carbon solids. The intermediates are quenched with toluene or benzene to give electrophilic substitution products and with water to give a quinomethine as the major product. The product pattern leads us to propose the carbene, perchlorobenzo[c,d]pyren-6-ylidene, or its reversible dimer as the major intermediate among others, that survives the severe conditions until coming into contact with these nucleophiles. Mainly from aromatic resonance stabilization, the carbene is proposed to have a delocalized singlet state analogous to a ylide electronic structure and, thus, undergoes observed ionic reactions instead of typical carbene reactions. This work serves as a mechanistic link on the structural evolution of carbon networks between molecular chemistry and nanomaterial chemistry.

Full text of DOI:10.1021/jo048190+

On Assembling Polychlorinated Aromatic Hydrocarbons from
Carbon Tetrachloride via Dichlorocarbene Intermediary by a
Solvothermal Reaction: A Reaction Pattern from Carbene-Ylide
Interconversion
Su-Yuan Xie,* Yin Peng, Meng Chen, Rong-Bin Huang, Yuan L. Chow,* and Lan-Sun Zheng
State Key Laboratory for Physical Chemistry of Solid Surfaces and Department of Chemistry, College of
Chemistry and Chemical Engineering, Xiamen University, Xiamen, 361005, China
syxie@jingxian.xmu.edu.cn; ychow@xmu.edu.cn
Received October 13, 2004
The forced one-electron reduction of carbon tetrachloride with sodium in a sealed steel vessel is
shown to have a narrow window of conditions to arrest the reaction at the polychlorinated aromatic
hydrocarbons (PCAHs), as well as to prevent the reaction from proceeding all the way to the final
stage of graphite and other carbon solids. The intermediates are quenched with toluene or benzene
to give electrophilic substitution products and with water to give a quinomethine as the major
product. The product pattern leads us to propose the carbene, perchlorobenzo[c,d]pyren-6-ylidene,
or its reversible dimer as the major intermediate among others, that survives the severe conditions
until coming into contact with these nucleophiles. Mainly from aromatic resonance stabilization,
the carbene is proposed to have a delocalized singlet state analogous to a ylide electronic structure
and, thus, undergoes observed ionic reactions instead of typical carbene reactions. This work serves
as a mechanistic link on the structural evolution of carbon networks between molecular chemistry
and nanomaterial chemistry.
Introduction
in handling any mixture of polyhaloalkanes and alkali
metals in laboratories3,4 until recent years, when in a
quick succession it was reported that such reactions could
be safely performed without an explosion in a sealed steel
reactor.5-8
This modification used a higher pressure/temperature
than those reactions in ambient conditions, and often in
the presence of metal salts as catalysts, to strip chlorine
The formation of a C-C bond through one electron-
reductive coupling of haloalkanes with alkali metals,
commonly known as the Wurtz reaction,1 has been in use
for a long time but rarely applied in more complex organic
synthesis owing to many limitations, among them the
severity of the reducing agent, alkali metals. When alkali
metals (or their mixtures) react with polyhalogenated
alkanes the process may be modified; for example, the
reactions with CCl4 are known to generate dichlorocar-
bene2,3 (:CCl2) as a reactive intermediate, accumulation
of which may cause sudden and violent reactions. For
this reason, chemists have been warned to be cautious
(4) Bretherich L. Handbook of Reactive Chemical Hazards, 3rd ed.;
Butterworth: London, 1985; p 1317.
(5) Li, Y. D.; Qian, Y. T.; Liao, H. W.; Ding Y.; Yang, L.; Xu, C. Y.;
Li, F. Q.; Zhou, G. E. Science 1998, 281, 246-247.
(6) (a) Jiang, Y.; Wu, Y.; Zhang, S. Y.; Xu, C. Y.; Yu, W. C.; Xie, Y.;
Qian, Y. T. J. Am. Chem. Soc. 2000, 122, 12383-12384. (b) Hu, G.;
Cheng, M. J.; Ma D.; Bao X. H. Chem. Mater. 2003, 15, 1470-1473.
(c) Wang, X. J.; Lu, J.; Xie, Y.; Du, G.; Guo, Q. X.; Zhang, S. Y. J.
Phys. Chem. B 2002, 106, 933-937.
(7) Lee, C. Y.; Chiu, H. T.; Peng, C. W.; Yen, M. Y.; Chang, Y. H.;
Liu, C. S. Adv. Mater. 2001, 13, 1105-1107.
(8) Hu, G.; Ma, D.; Cheng, M. J.; Liu, L.; Bao, X. H. Chem. Commun.
2002, 1948-1949.
(1) March, J. Advanced Organic Chemistry, 2nd ed.; McGraw-Hill:
New York, 1977; pp 407-409.
(2) Moss R. A. In Carbenes; Moss R. A., Jones M., Jr., Eds.; Wiley:
New York, 1973; Vol. I, pp 153-304.
(3) Fumiss B. S. Vogel’s Textbook of Practical Organic Chemistry,
4th ed.; Longman: London and New York, 1986; p 268.
10.1021/jo048190+ CCC: $30.25 © 2005 American Chemical Society
Published on Web 01/25/2005
1400
J. Org. Chem. 2005, 70, 1400-1407

Assembly of Polychlorinated Aromatic Hydrocarbons from CCl4
structural details for integrating carbon to complex
molecules. They5-9 also point to the fact that the reaction
is very sensitive to the small changes of conditions such
as temperature, duration, diluents, etc. As the intermedi-
ate PCAHs would provide key information to assign
rational mechanisms of carbon material formations, the
search of the reaction conditions and analysis of the
intermediates were undertaken. This paper describes the
separation and isolation of some PCAHs with molecular
weight (MW) <700 and the proposal of reactive inter-
mediates that can be used to examine the formation
mechanism in the early to middle stage for solid carbon
materials. While complexity of the product pattern may
discourage synthetic chemists, at this early stage of
development the proposal of major intermediates is the
first attempt in the field; it opens up the first entry to
clarify mechanistic links to molecular and nanomaterial
chemistry.
Results
Optimization of the Reaction Conditions. In a
Teflon-lined stainless steel reactor a mixture of carbon
tetrachloride (0.26 mol) and sodium (0.13 mol) gave no
significant reaction at 220 °C for 40 h either in the
absence or presence of CoCl2, NiCl2, CuCl2, and/or FeCl3
as catalysts. Subsequently, exploration was carried out
in a bare stainless steel vessel at this reactant ratio.
When the temperature was set at 300 °C a good amount
of toluene-soluble substances, identified as highly poly-
chlorinated aromatic hydrocarbons (PCAHs), and some
insoluble black amorphous carbon (mainly graphite) were
obtained. The latter black carbon materials increased
rapidly at the expenses of PCAHs at 400 °C and totally
dominated the crude product at 800 °C, in addition to a
small amount of iron carbide, as shown by X-ray diffrac-
tion (XRD) analysis. At 300 °C and a CCl4/Na mole ratio
of 0.26:0.13 (i.e., 2:1), PCAHs were obtained copiously and
largely consisted of those with molecular weight below
700. While with the mole ratio at 1:1 the crude product
was almost completely graphite, with the ratio at 6:1
PCAHs decreased drastically and consisted of primarily
small polychloroaromatic molecules (MW 200-400).
At 400 °C, the reaction was finished in less than 12 h
giving much black materials. At 300 °C, Na was not
totally consumed in <30 h. Thus, the standard conditions
for the present investigation were set at the reaction of
CCl4 (0.26 mol) and Na (0.13 mol) in a stainless steel
reactor at 300 °C for 40 h. The conditions led to complete
consumption of Na and gave off a small amount of yellow
fume that was assumed to be the chlorine gas when
reactor was opened. The reproducibility of the reaction
under optimal conditions was well established. Owing to
corrosion, the reactor was changed every few runs. As
this exothermic reaction builds up a moderately high
pressure and temperature during heating, the reactor
must be protected inside an appropriate shield; upon
completion the reactor can be opened safely at the room
temperature.
FIGURE 1. PCAHs isolated from the products of solvothermal
reaction of CCl4 with Na.
in an orderly fashion from polychlorohydrocarbons, that
is a reaction under solvothermal conditions.5 This tech-
nique has been successfully used to prepare nanoscale
carbon solid materials; for example, diamond powders
from CCl4,5 carbon nanotubes from hexachlorobenzene
(1 in Figure 1)6a,b and tetrachloroethylene,6c carbon onions
from hexachloropentadiene,7 and hollow spherical graph-
ite from 1.8 Allied reactions using perfluorinated hydro-
carbons or CCl4 to synthesize various carbon nanotruc-
tures (e.g., nanotubes) have also been reported.9
We have long-standing interests in the fabrication of
fullerenes and allied carbon materials.10-12 One technique
is high voltage electric discharge in liquid10 or vaporized11
chloroform and CCl4; in such a process we have isolated
perchlorinated aromatic hydrocarbons (PCAHs) that can
be construed as building blocks for fullerenes and also
identified a trace of C50 derivative,12 as well as C60 and
C70. Under the solvothermal conditions as described
above, it was indeed puzzling that the reaction was either
nil or gave completely black carbon materials even when
the 1:1 mole ratio of CCl4 to Na was used (see below).
Unfortunately, in the existing reports on carbon mater-
ials,5-9 researchers basically regard this reaction as a
routine technique with arbitrary intermediates and pay
scarce attention to the molecular transformation and
(9) (a) Kavan, L.; Hlavaty, J. Carbon 1999, 37, 1863-1865. (b) Vohs,
J. K.; Brege, J. J.; Raymond, J. E.; Brown, A. E.; Williams, G. L.;
Fahlman, B. D. J. Am. Chem. Soc. 2004, 126, 9936-9937.
(10) Huang, R. B.; Huang, W. J.; Wang, Y. H.; Tang, Z. C.; Zheng,
L. S. J. Am. Chem. Soc. 1997, 117, 55954-5955.
(11) Xie, S. Y.; Deng, S. L.; Yu, L. J.; Huang, R. B.; Zheng, L. S. J.
Phys. Chem. B 2001, 105, 1734-1738.
(12) Xie, S. Y.; Gao, F.; Lu, X.; Huang, R. B.; Wang, C. R.; Zhang,
X.; Liu, M. L.; Deng, S. L.; Zheng, L. S. Science 2004, 304, 699-699.
Product Analysis and Isolation. The exposed reac-
tion mixture was immediately quenched with toluene to
afford products which were separated into the toluene
(or ethylbenzene) soluble part and insoluble materials;
the former contained the intermediates for the formation
of the latter and was the target of the present investiga-
J. Org. Chem, Vol. 70, No. 4, 2005 1401

Xie et al.
tion. As the toluene solution contains some 50 odd
PCAHs, some of which are in trace amounts and/or high
molecular weight with limited solubility in various
solvents, it is a challenge to perform analysis, isolation,
and identification. For this, high-performance liquid
chromatography (HPLC) coupled with ultraviolet spec-
troscopy (UV) and/or mass spectrometry (MS) is a power-
ful and efficient tool at the present stage.13 This research
group, over many years’ experiences dealing with allied
reactions and related PCAHs, has established efficient
methods and know-hows in these separation techniques.
Detailed HPLC-UV/MS procedures for the separation
and identification of the present PCAH series were
described in the Ph.D. thesis of Y.P.;14 the method was
developed on the basis of the principle and practice in
dealing with other PCAH series.15 The toluene solution
was investigated with this analytical system to give an
HPLC chromatogram as shown in Figure S1 (see the
Supporting Information). It must be understood that
peak heights (or areas) are good indicators of abundance,
but are not strictly proportional to the concentrations,
i.e., corresponding yields. Those well-defined chromato-
graphic peaks were examined with HPLC-UV and/or
HPLC-MS; the latter MS spectra generally gave distinct
molecular ion peaks whose chlorine isotope distribution
pattern16 was utilized to judge the number of chlorine
and, in turn, to compute the molecular formula. Among
the structures of the products, perchlorobenzene 1,17
perchloronaphthalene 2,18 perchlorobiphenyl 3,19 perchlo-
ropyrene 4,20 and perchlorofluoranthene 521 (see Figure
1) have been previously characterized. In the present
experiment, they were readily identified by co-injection
HPLC and confirmed when isolated (vide infra) by direct
spectral comparisons with authentic samples available
in our group. For the mechanistic studies, early stage
intermediates must be identified to provide a reaction
pattern, i.e., major peaks of PCAHs with the MW range
500-700 should be identified. Larger molecules of MW
>700 were produced in low yields and, together with
their solubility problems, not amenable to isolation.
The crude product from the toluene solution was
sublimed at 120-160 °C and then up to 200 °C to remove
mostly 1, but in the latter range, in addition small
amounts of 2-5. HPLC-MS and co-injection with au-
thentic samples were used to identify these PCAHs. The
remaining solid was chromatographed repeatedly on a
neutral alumina column using cyclohexane as the eluent
to afford pure samples of 2-4 and 1,2,3,4,4b,4c,6,7,8,9,-
9b,9c-dodecachloro-4b,4c,9b,9c-tetrahydrocyclobuta[1,2-
a;3,4-a′]diindene-5,10-dione C18Cl12O2 (9),22 6-(p-tolyl)-6H-
1,2,3,4,5,7,8,9,10,11-decachlorobenzo[cd]pyrene C26Cl10H8
(10),23 and 6-(p-tolyl)-6H-1,11-dihydroxy-2,3,4,5,7,8,9,10-
octachlorobenzo[cd]pyrene C26Cl8H10O2 (11) as shown in
Figure 1. An ethylbenzene soluble extract was also
chromatographed in the similar way to afford perchlo-
rocoronene (C24Cl12, 6),24 and perchlorobenzo[f]indenone
C13Cl8O (7), and 1,2,3,5,6,7,8,9-octachlorocyclopenta[def]-
phenanthren-4-one C15Cl8O (8).25
Repeated column chromatography caused much loss
of materials in each run due to irreversible adsorption
and band overlaps and demanded a large outlay of time
and solvents for good separations. To circumvent these
drawbacks, a combined column and HPLC process was
also developed as an alternative. That is, an ethylbenzene
extractive was chromatographed on an alumina column
but eluted with toluene to give four blocks of mixtures.
The first fraction was recrystallized many times to give
perchlorofluoranthene 5. The last two fractions were
worked up by preparative HPLC; 6H-1,2,4,5,7,8,9,10,11-
nonachlorobenzo[cd]pyrene-3-one C19Cl9HO (12) was ob-
tained from the third fraction. In addition, three higher
molecular weight PCAHs, among others, were also
obtained in milligram scales just enough to record their
MS, infrared (IR), and/or UV spectra; this group included
C25Cl10O2 and two isomers of C26Cl7H5O and was not
investigated further.
The MS, UV, and IR spectra of these PCAHs were
recorded. X-ray crystallography was employed to identify
the structures of 5, 6, and 8-12; details of the X-ray
structure of 8-10 have been reported.22,23,25 However,
1
owing to availability and solubility of these PCAHs, H
and 13C NMR spectra were successfully obtained only
with 10, which was isolated in the largest amount. These
physical parameters are discussed in conjunction with
X-ray structures below. It should be noted that owing to
tedious and inefficient isolation procedures, product
yields could not be evaluated with reliability; some
isolated weights are given in the Experimental Section.
On the basis of the isolations, 10-12 were major and 5-9
minor products; the major products added up to a 6%
yield (75 mg) on the basis of individual calculations.
X-ray Structures and Properties. Owing to the
large steric volume of chlorine atoms, these PCAHs have
degrees of molecular strains that cause skeletal distor-
tions, which should greatly affect their physical proper-
ties. Thus detailed molecular shapes by X-ray crystal-
lography were obtained wherever good grade crystals
could be obtained. As a detailed description requires a
large space, in this paper only relevant structures and
parameters will be given to facilitate discussion. The
X-ray structures of both 5 and 6 are in record, the former
was analyzed again as (C16Cl10)2/CS2, the crystal obtained
in recrystallization from carbon disulfide (see Figure S2
in the Supporting Information). Both structures show
(13) Niessen, W. M. A. Liquid chromatography-mass spectrometry;
M. Dekker: New York, 1999.
(14) Peng, Y. Ph.D. Thesis, Xiamen University, Xiamen, June 2003.
(15) (a) Xie, S. Y.; Deng, S. L.; Huang, R. B.; Zheng, L. S. J.
Chromatogr. A 2001, 932, 43-53. (b) Peng, Y.; Xie, S. Y.; Chen, M.;
Feng, Y. Q.; Yu, L. J.; Huang, R. B.; Zheng, L. S. J. Chromatogr. A
2003, 1016, 61-69.
(16) Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectroscopic
Identification of Organic Compounds; John Wiley: New York, 1995.
(17) Brown, G. M.; Strydom, O. A. W. Acta Crystallogr. B 1974, 30,
801-804.
(18) (a) Gafner, G.; Herbstein, F. H. Nature 1963, 200, 130-132.
(b) Grasselliet, J. G.; Ritchey, W. M. Atlas of Spectral Data and
Physical Constants for Organic Com-pounds, 2nd ed.; CRC Press:
Cleveland, 1975; Vol. 3, pp 645-646.
(19) Pedersen, B. F. Acta Crystallogr. B 1975, 31, 2931-2933.
(20) Hazell, A. C.; Jagner, S. Acta Crystallogr. B 1976, 32, 682-
686.
(21) (a) Ballester, M.; Castaner, J.; Riera, J.; Pares, J. An. Quim. C
1980, 76, 157-170. (b) Xie, S. Y.; Huang, R. B.; Chen, L. H.; Huang,
W. J.; Zheng, L. S. Chem. Commun. 1998, 2045-2046.
(22) Peng, Y.; Xie, S. Y.; Long, L. S.; Huang, R. B.; Zheng, L. S. Acta
Crystallogr. E 2004, 60, o763-764.
(23) Peng, Y.; Xie, S. Y.; Huang, R. B.; Zheng, L. S. Acta Crystallogr.
E 2001, 57, o617-618.
(24) Baird, T.; Gall, J. H.; MacNicol, D. D.; Mallinson, P. R.; Michie,
C. R. J. Chem. Soc., Chem. Commun. 1988, 1471-1473.
(25) Peng, Y.; Xie, S. Y.; Deng, S. L.; Huang, R. B.; Zheng, L. S.
Acta Crystallogr. E 2004, 60, 899-900.
1402 J. Org. Chem., Vol. 70, No. 4, 2005

Assembly of Polychlorinated Aromatic Hydrocarbons from CCl4
21.2° (2.8°), 14.5° (4.2°), and 11.3° (2.0°), respectively. The
magnitude of deviations shows the size of surface humps
in these three PCAHs and 5-6 and becomes clearer when
compared with the planar structure of 8 in Figure S4
(Supporting Information). It follows that two hydroxyl
groups in 11 relieve such crowding significantly probably
owing to intramolecular hydrogen bonding, but the
quinoid (IR band at 1650 cm-1) moiety in 12 did not
alleviate the steric crowding very much. All of them show
pertinent IR, UV, and MS spectra. Compound 10 shows
the pertinent 1H NMR signals (see the Experimental
Section) and 16 13C NMR signals that require a plane of
symmetry cutting across the C-6.29 As the X-ray structure
(Figure 2) shows the lack of such a plane, the molecule
must be undergoing fast (i.e., faster than the NMR
time scale) wagging motion of the PCAH surface in the
solution phase at the cavity temperature. Such fast
conformational exchanges create an average PCAH plane
with the plane of symmetry that can afford the observed
16 13C NMR signals.16
Miscellaneous Reaction Conditions. To trace the
source of oxygen atoms present in some PCAHs, the
standard reaction conditions were maintained but the
quenching process was modified to see the effects on the
product pattern. First of all, the reaction mixture upon
opening was immediately quenched with either water or
benzene instead of toluene. In both cases the resulted
HPLC spectra were completely different and devoid of
10 and 11 as expected. In the former spectrum a very
large peak of 12 was recorded and was isolated as the
major product. In the latter case, a peak corresponding
to C25Cl10H6 (10′, i.e., 10 where tolyl ) phenyl) was
detected through HPLC-MS, although the intensity was
not as prominent as that of 10 from toluene quenching.
The implication was unmistakable that at the end of the
solvothermal reaction an intermediate accumulated and
reacted with toluene to give 10-11 and benzene to 10′
and also with water to give 12.
FIGURE 2. Crystal structure of 10.
severe twist from coplanarity arising from chlorine
overcrowding; the former arising from opposing chlorines
at the two 1,4-angular positions, and the latter from
parallel 1,3-C-Cl bonds at the six peri-positions. As the
result, the latter molecular surface has three humps
around the periphery. 9 has a twisted cyclobutane ring
with all-cis chlorine orientations; two fused indanone
rings avoid each other in space (Figure S3 in the
Supporting Information). Compound 8 has a planar
surface despite two 1,3 peri-interactions as shown in
Figure S4 (Supporting Information) that are obviously
much relieved due to tightening of the pentadienone ring.
Both 7 and 8 show typical pentadienone IR band at 1721
and 1731 cm-1, respectively. The structure of 7 is
tentatively assigned on the basis of MS, and its IR and
UV spectra that are distinctly different from those of
perchlorofluorenone26 and perchlorophenalenone,27 and
calculations using the semiempirical PM3 method.28a In
addition, density-functional theory (B3LYP/6-31G)28b cal-
culations prove the relative energy of 7 (-19.5553 kJ/
mol) is much lower than its isomers benzo[e]indenone (0
kJ/mol) and benzo[g]indenone (-10.4900 kJ/mol). Were
the X-ray structure of 7 available, it should show
significant molecular distortion due to a double peri-
interaction (i.e., three nearly parallel 1,3-C-Cl bonds).
While PCAHs 10-12 share the same PCAH skeleton,
their X-ray structures are quite different from the small
difference in substitution patterns. In both 10 and 11
(Figure 2 and Figure S5, Supporting Information) the
toluene ring is forced into a nearly perpendicular orienta-
tion with respect to the PCAH ring and does not exert
significant effects on PCAH steric-wise. Major product
10 has the most distorted PCAH surface followed by 12
(Figure S6, Supporting Information) and then 11. The
measured deviation from the average PCAH plane is
0.1844, 0.1352, and 0.0719 Å, respectively, and the
maximum (minimum) angles of distorted six-membered
ring planes against each other in the PCAH surface are
Separately, a reaction, including the weighing of Na,
was prepared inside a drybox under nitrogen and sealed.
The reaction was performed and worked up in the usual
way to be quenched by toluene; the toluene solution was
HPLC analyzed to show the complete absence of peaks
corresponding to 7, 8, and 9 and a much-reduced inten-
sity of the peak for 11. But significantly, the peak
corresponding to 12, though relatively small, was still
present. Thus oxygen atoms in 7-9 and 11 must be
derived from oxygen in the air entered into the reaction
at the beginning, and that in 12 must not be from the
air but from adventitious moisture upon the opening of
the reaction vessel.
Discussion
The published5-9 as well as present results show that
the solvothermal reaction of Na (or K) with CCl4 (or other
polychlorocarbon) in a closed vessel readily strips chlorine
atoms sequentially and integrates single carbon units to
reach a wide range of carbon materials. Our exploration
indicates that the reaction of Na with CCl4 has a narrow
window of conditions by which the reaction can be
(26) Ballester, M.; Riera, J.; Castan˜er, J.; Bad´ıa, C.; Monso´, J. M.
J. Am. Chem. Soc. 1971, 93, 2215-2225 and related papers.
(27) Koutentis, P. A.; Chen, Y.; Cao, Y.; Best, T. P.; Itkis, M. E.;
Beer, L.; Oakley, R. T.; Cordes, A. W.; Brock, C. P.; Haddon, R. C. J.
Am. Chem. Soc. 2001, 123, 3864-3871.
(28) (a) Stewart, J. J. P. J. Comput. Chem. 1989, 10, 209-220. (b)
Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(29) Memory, J. D.; Wilson, N. K. NMR of Aromatic Compounds;
John Wiley & Sons: New York, 1982; pp 93-99.
J. Org. Chem, Vol. 70, No. 4, 2005 1403

Xie et al.
SCHEME 1
dichloro derivative 13 as a precursor; for all probability
the latter is rapidly reduced by Na to 14 aided by the
steric crowding26 at the CCl2 moiety (at C-6) and has only
a fleeting lifetime. A carbene with the sp2 hybridization
for the reactive center in the condensed aromatic system
is deduced rationally on the basis of products 10-12
isolated upon quenching and also prevalent reductive
dechlorination by Na. The singlet electronic state 14a is
rationalized to be the lowest (or very close to the lowest)
ground state as the intermediate mainly in analogy to
well-established aromatic carbene intermediates pub-
lished in the literature.2,35-37 First, the electron pair is
placed in the sp2 orbital which has a lower energy than
the empty p-orbital that creates aromatic resonance
stabilization from 18 (4n + 2) π-electrons in a circular
carbenium ion configuration.38 This is in good analogy
to extensively studied case of cycloheptatrienylidene and
diphenylcyclopropenylidene;35-37,39 in particular, the former
was shown conclusively by MNDO computations,40-43 IR
spectroscopy,38a and trapping experiments44 to have a
singlet-state electronic configurations in the ground state,
though the structure varies with a complex pattern.40,41
Second, suitably located chlorine substituents43 in the
periphery can donate electron pairs through conjugation
to the vacant carbene p-orbital to stabilize 14a, as it is a
well-established fact that heteroatom electron pairs36,45
can lower the singlet state energy to the extent of gaining
ground singlet state, e.g., in xanthilidene46 and dimethoxy-
fluorenylidene.47 Finally, the electron pair (negative
charge) in the sp2 orbital is shielded from steric crowding
of the both peri-positions that prevent its access to a
attacking species, except probably to a small proton.
Compound 14b with the electron pair localized in the
p-orbital, owing to contradiction against the reasons state
above, should have a higher energy and can scarcely
contribute to the structure of the intermediate. It must
be added that in the case of 2,5-cyclohexadienon-4-ylidene
and allied cases, this electronic configuration definitely
assume importance from a reason of negative charge
arrested at the PCAH level without going down to carbon
materials. At 300 °C in a closed reactor both reactants
must be a superheated fluid under a certain high pres-
sure. Under such extreme conditions, its reaction mech-
anism can only be speculated on the basis of the product
pattern upon quenching. The stepwise one electron
reduction of CCl4 by Na under such conditions appears
to proceed slowly to give the observed product pattern,
which is different from that of an open electric arc
discharge in CCl4 and chloroform.10,11 The latter is
initiated by high-voltage (10 kV) electron capture that
produces PCAHs containing five-membered rings (e.g.,
perchlorinated acenaphthene,21b,30 fluoranthene,21 and
corannulene10,31); those carbon skeletons with built-in
five-membered rings are generally accepted precursors
as building blocks for fullerenes.32,33 The present reaction
produces PCAH precursors for mostly graphite except a
small amount of perchlorofluoranthene; its early stage
reactions could be formally written (among other pos-
sibilities) as in Scheme 1 according to the accepted one-
electron reduction mechanism.2 Thus, we interpret that
dichlorocarbene and tetrachloroethene are the basic
building blocks for the observed PCAHs. This interpreta-
tion shares the same mechanistic pathway, at least at
the initial stage, with the former arc discharge reac-
tion,10,11 and supported by the formation of some small
PCAHs, such as 1 and 2, in common. At the pressure and
temperature under the reaction conditions, tetrachloro-
ethene is more likely to undergo thermal [2 + 2]- and [2
+ 2 + 2]-type cycloadditions34 more rapidly than to
undergo reductive chlorine elimination. Dichloroacetylene
is less likely an intermediate, but that cannot be proven
at this stage. Another possibility is that in the present
reaction dichlorocarbene may react with anion NaCCl3
faster than dimerize by itself; this remains to be studied.
While under such severe reaction conditions an exten-
sive scatter of products as observed is unavoidable, good
yields of 10 and its congeners 11 and 12 are indeed
surprising and remarkable. These three products share
the same PCAH frame that suggests the major pathway
for this reaction; namely, under the conditions the
reaction coalesces to an intermediate that must be
persistent owing to structural reasons until contacting
with benzenes or water to undergo irreversible reactions.
As shown in Scheme 2, we propose perchlorobenzo[c,d]-
pyren-6-ylidene in its singlet state2,35 14a as the reactive
species for the intermediate and the corresponding gem-
(35) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry B,
3rd ed.; Plenum Press: New York, 1990; pp 511-532.
(36) Platz, M. S.; Maloney V. M. In Kinetics and Spectroscopy of
Carbenes and Diradicals; Platz, M. S., Ed.; Plenum: New York, 1990;
pp 239-352.
(37) Kirmse W. In Advance in Carbene Chemistry; Brinker, U. H.,
Ed.; JAI Press: Greenwich, CT, 1994; pp 1-51.
(38) (a) McMahon, R. J.; Abelt, C. J.; Chapman, O. L.; Johnson, J.
W.; Kreil, C. L.; Le Roux, J. P.; Mooring, A. M.; West, P. R. J. Am.
Chem. Soc. 1987, 109, 2456-2469. (b) Kuzaj, M.; Luerssen, H.;
Wentrup, C. Angew. Chem., Int. Ed. Engl. 1986, 25, 480-482. (c)
Janssen, C. L.; Schaefer, H. F. J. Am. Chem. Soc. 1987, 109, 5030-
5031.
(39) Tomioka H. In Carbene Chemistry; Bertrand, G., Ed.; Marcel
Dekker: New York, 2002; pp 103-147.
(40) Schreiner, P. R.; Karney, W. L.; Schleyer, P.v. R.; Borden, W.
T.; Hamilton, T. P.; Schaefer, H. F. J. Org. Chem. 1996, 61, 7030-
7039.
(41) Wong, M. W.; Wentrup, C. J. Org. Chem. 1996, 61, 7022-7029.
(42) Harrison, J. F.; Liedtke, R. C.; Liebman, J. F. J. Am. Chem.
Soc. 1979, 101, 7162-7168.
(30) Mack W. Tetrahedron Lett. 1966, 7, 2875-2877.
(31) (a) Cheng, P.-C.; Scott L. T. Abstracts of Papers, 208th National
Meeting of the American Chemical Society; American Chemical
Society: Washington, DC, 1994; 87-ORGN Part 2. (b) Scott, L. T. Pure
Appl. Chem. 1996, 68, 291-300. (c) Seiders, T. J.; Elliott, E. L.; Grube,
G. H.; Siegel, J. S. J. Am. Chem. Soc. 1999, 121, 7804-7813. (d)
Samdal, S.; Hedberg, L.; Hedberg, K.; Richardson, A. D.; Bancu, N.;
Scott, L. T. J. Phys. Chem. A 2003, 107, 411-417.
(43) (a) Baird, N. C.; Taylor, K. F. J. Am. Chem. Soc. 1978, 100,
1333-1338. (b) Murray, K. K.; Leopold, D. G. J.; Miller, T. M.;
Lindbergr, W. C. J. Chem. Phys. 1988, 89, 5442-5453.
(44) Harris, J. W.; Jones, W. M. J. Am. Chem. Soc. 1982, 104, 7329-
7330.
(45) Bertrand G. In Carbene Chemistry; Bertrand, G., Ed.; Marcel
Dekker: New York, 2002; pp 177-199.
(32) Rabideau, P. W.; Sygula, A. Acc. Chem. Res. 1996, 2, 9, 235-
242.
(46) Lapin, S. C.; Schuster, G. B. J. Am. Chem. Soc. 1985, 107,
4243-4248.
(33) Mehta, G.; Rao, H. S. P. Tetrahedron 1998, 54, 13325-13370.
(34) Isaacs, N. S. Physical Organic Chemistry; Longman Scientific
& Technical: Essex, 1987; pp 701-753.
(47) Chuang, C.; Lapin, S. C.; Schrock, A. K.; Schuster, G. B. J. Am.
Chem. Soc. 1985, 107, 4238-4243.
1404 J. Org. Chem., Vol. 70, No. 4, 2005

Assembly of Polychlorinated Aromatic Hydrocarbons from CCl4
SCHEME 2
delocalization.48 Whereas in general, alkyl and phenyl
carbenes have a triplet diradical in the ground state,2,35
in the present case triplet diradical 14c is believed to
have a comparable or even higher energy than 14a on
the reasons as discussed above. While attempted calcula-
tions of the energy difference between 14a and 14c have
not been successful, we speculate that the singlet-triplet
energy split is, in analogy to those above aromatic
carbenes,35-37,39,48bc fairly small in the order of few kilo-
calories either ways. Thus, under the reaction conditions,
there is no agent to intercept carbene 14a except a small
amount of intruding oxygen from the beginning that
allows its survival. The related oxygenation will be
discussed later. The concept of resonance stabilized
carbenes is similar to that described for heteroatom
electron pair stabilized singlet carbenes49 appeared while
this is in preparation; both have ylide electronic struc-
tures. Alternatively, carbene 14a may survive under the
conditions as its dimeric form with a severely twisted Cd
C bond in analogy to perchloro-9,9-bifluorenylidene;50 the
weak CdC bond allow the dimer to reversibly generate
14a under the thermal conditions. Even under severe
steric crowding from peri-bis-chloro interactions, we may
envisage certain kind of dimeric association through a
long bonding. The discussion of other carbenes (see below)
can share similar alternative possibilities.
Having proposed carbene 14 as the intermediate, the
observed reaction pattern must be discussed. The benzyl
derivatives corresponding to 10 and 11 have not been
identified in isolable quantities among complex products,
although small amounts may have escaped the detection.
This indicates that carbene 14 has feeble (or lacks) singlet
insertion and triplet radical reactivity. Those singlet
carbenes cited above also tend to be more facile in
addition than insertion, and do not insert into the R-CH
bonds of tetrahydrofuran.2 While the access for spectro-
scopic examinations of the intermediate is beyond reality
at this stage, singlet carbene 14a can be perceived as a
(48) (a) Bucher, G.; Sander, W. J. Org. Chem. 1992, 57, 1346-1351.
(b) Lin, Y. S.; Jiang, S. Y.; Huang, T. C.; Lin, S. J.; Chow, Y. L. J. Org.
Chem. 1998, 63, 3364-3370. (c) Sole, A.; Olivella, S.; Bofill, J. M.;
Anglada, J. M. J. Phys. Chem. 1995, 99, 5934-5944.
(50) Molins, E.; Miravitlles, C.; Espinosa, E.; Ballester, M. J. Org.
Chem. 2002, 67, 7175-7178.
(49) Kirmse, W. Angew. Chem., Int. Ed. 2004, 43, 1767-1769.
J. Org. Chem, Vol. 70, No. 4, 2005 1405

Xie et al.
SCHEME 3
carbene center is not severely hindered, they are rapidly
scavenged by floating oxygen. In Scheme 3, the trans-
formation of 16 via 18 to 8 is used as the model to
illustrate the oxidation process to ketonic PCAHs 7-9.
It is also expected that 14a may react with oxygen in
the same way after surface crossing, albeit more slowly
from steric crowding at the carbene center, to afford the
corresponding ketone. We did find a medium peak having
C19Cl10O in the crude toluene extract (see Figure S1,
Supporting Information), but could not isolate this com-
pound, presumably it was difficult to come out of column.
Finally, we assume that a carbene similar to 14 but with
two OH substituents at the 1,11-position must be the
corresponding precursor to 11. However, we have no
simple mechanism to incorporate an oxygen molecule into
the particular 1,3-position.
Going back to Figure 1, while PCAHs 1-3 are common
to solvothermal reactions of CCl4, and their structures
are too small and early in the process to tell the story in
the skeletal integration process, those of 4, 6, 10-12
suggest that they (or their precursors) are the intermedi-
ates for graphite.8 PCAHs 5 and 8 are most likely the
precursors to perchlorocorannulene,10,11,31 which incor-
porates a five-membered ring to cause curving of molec-
ular surfaces and has been assumed to be a precursor of
carbon onions and fullerenes.7,12,32,33 Their low yields
suggest that such final carbon materials must be minor
products. We have difficulty to fit ketones 7 and 9 into
these reaction schemes and suspect that they are dead
alleys in the reaction. On the basis of the known
pericyclic reactions,57 there should be more than one
acceptable mechanistic pathway that can be proposed for
the sequential integration of building blocks, :CCl2 and
Cl2CdCCl2, to give PCAHs in Figure 1. At this stage, with
limited information it is not profitable to go into deeper
mechanistic discussions.
zwitterion (or ylide) of the localized electron pair in the
sp2 σ-orbital and an extensively delocalized positive
charge in the p-orbitals over the aromatic frame.36,37 The
reaction of such a ylide must reflect its electronic
structure; with water it should be most facile by initial
protonation and with a benzene ring by the π-complex
formation. The former protonation is in agreement with
nucleophilic reactivity of cycloheptatrienyliden51 and
allied carbenes;36,37,45 this leads to 12 after elimination
of HCl. The latter π-complex should collapse by an
electrophilic bond formation in the aromatic substitution
pattern leading to 10 and 11. These reaction pathways
must be energetically the lowest and should be lower
than the activation energy required for insertion or
hydrogen abstraction for this type of delocalized and
hindered carbenes.39,45 Thus, we propose here carbene 14
that does not undertake typical carbene reactions but
preferentially does ordinary ylide ionic reactions in the
ground state as the intermediate. Together with those
delocalized singlet carbenes,39,45,51a they may be classified
as anomalous carbenes.
Triplet carbenes are known to be trapped by oxygen
with a diffusion-controlled rate to form a carbonyl oxide
(or ylide),36,52 known as the Criegee’s ozonolysis inter-
mediate from the original proposal,53 which reacts with
water to give a carbonyl compound and H2O2; for ex-
ample, 18 in Scheme 3. Noting that the carbonyl oxygen
atom in minor PCAHs 7-9 is derived from oxygen in the
intruding air, we suggest that these ketones are derived
from the corresponding carbenes 15-17 (i.e., structural
analogues to 14); 15-17, in turn, should come from the
corresponding vicinal dichloro-derivatives as logical pre-
cursors through reductive dechlorination by Na (see
Scheme 1) and probably exist as their dimeric forms in
the reaction solution (vide supra).50 These perchloro
aromatic carbenes 15-17, regardless of their electronic
configuration in the ground state, should have a small
singlet-triplet energy split54 from the reasons discussed
above; even if 15-17 possess a ground singlet state, they
can cross over the energy surface at the high reaction
temperature to react as a triplet with oxygen.55,56 As their
Finally, it is necessary to comment on a small amount
of what appears to be a yellow chlorine gas observed
when the vessel is opened. It is envisaged that chlorine
could be generated by thermolysis of suitable perchloro
intermediates at the point when all Na was consumed.
It is known that perchlorodiphenylmethane undergoes
rapid thermal dissociation to form a chlorine atom and
stabilized radical;26 many stabilized (or persistent) radi-
cals from steric reasons and extensive delocalization58
from thermal dissociation are well documented. In view
of observed severe distortion in the vicinity at C-6 in 10-
11 (see Figure 2), gem-dichloro precursor 13 must be just
as (or even more) seriously crowded sterically and a good
candidate to undergo homolysis at 300 °C to a stabilized
radical 13′ that should afford chlorine. There is no
information to suggest the thermolytic fate of 13′ in the
product pattern; at 300 °C many reactions may occur to
relieve strains due to the double 1,3-peri interaction of
three parallel C-Cl bonds in 13′. We may venture to
speculate the gaining of greater stabilization in addition
(51) (a) Duell, B. L.; Jones, W. M. J. Org. Chem. 1978, 43, 4901-
4903. (b) Christiansen, L. W.; Waali, E. E.; Jones, W. M. J. Am. Chem.
Soc. 1972, 94, 2118-2119.
(55) Casal, H. L.; Tanner, M.; Werstiuk, N. H.; Scaiano, J. C. J. Am.
Chem. Soc. 1985, 107, 4616-4620.
(52) Ganzar, G. A.; Sheridan, R. S.; Liu, M. T. H. J. Am. Chem. Soc.
1986, 108, 1517-1520.
(56) Griller, D.; Hadel, M.; Nazran, A. S.; Platz, M. S.; Wong, P. C.;
Savino, T. G.; Scaiano, J. C. J. Am. Chem. Soc. 1984, 106, 2227-2235.
(57) Isaacs, N. S. Physical Organic Chemistry; Longman: Essex,
1987; Chapter 14.
(53) House, H. O. Modern Synthetic Reactions, 2nd ed.; W. A.
Benjamin: Reading, MA, 1972; pp 359-387.
(54) Grasse, P. B.; Zupancic, J. J.; Lapin, S. C.; Hendrich, M. P.;
Schuster, G. B. J. Org. Chem. 1985, 50, 2352-2356.
(58) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry A,
2nd ed.; Plenum Press: New York, 1984; p 627.
1406 J. Org. Chem., Vol. 70, No. 4, 2005

Assembly of Polychlorinated Aromatic Hydrocarbons from CCl4
matographed on an alumina column and cyclohexane in the
similar method as above to afford, among others, 6 (1 mg), 8
(1 mg), and 7 (10 mg) in normalized weights.
to relieve strains may help further homolysis of the
remaining C6-Cl to give 14.
In conclusion, the forced one-electron reduction of CCl4
with Na under the solvothermal reaction produces PCAH
intermediates that are on its way to be integrated into
graphite and allied carbon solids. Under the controlled
conditions as defined, perchlorobenzo[c,d]pyrene-6-yl,
carbene (14) is deduced to be the major intermediate that
can be quenched with benzenes and water to afford
corresponding products 10-12. Mainly from aromatic
resonance stabilization carbene 14 is proposed to have a
delocalized singlet state analogous to an ylide electronic
structure and should undergo observed ionic reactions
instead of prototype carbene reactions. Under the reac-
tion conditions, 14 exists as a stable carbene-ylide
resonance hybrid or alternatively stabilized as a revers-
ible carbene dimer. Whereas the product pattern is
complex, the integration process of one carbon unit to 14
provides a critical insight into mechanisms of carbon solid
formations. The reaction should and could be improved
to be a stepping stone to nanomaterial chemistry.
The other half of the ethylbenzene extract was taken up in
an alumina column and eluted with toluene to afford four
impure fractions A-D in the order of elution. Fraction A was
recrystallized from cyclohexane three times to afford 5 (6 mg).
The fractions C and D were separated with a preparative
HPLC on an ODS column (ECONO-PREP C18 column, 250
mm in length and 21.2 mm in diameter) employing methanol/
toluene/water (75:25:5 volume ratio) as the mobile phase
through repeat operations. From fraction C, 12 (20 mg) was
isolated. Also, from fraction C and D few milligrams each of
C25Cl10O2 and two C26Cl7OH5 isomers were obtained; their IR,
UV, and MS spectral data as well as some details of isolating
each member of these products were recorded in the thesis of
Y.P.14
X-ray Crystallographic Analysis. X-ray quality single
crystals of these PCAHs (except 7) were obtained by crystal-
lization from the cyclohexane and/or toluene solution, except
5 which afforded the CS2 complex in good crystalline form. The
X-ray structural analyses data were collected at 298 K on a
SMART 1000 CCD diffractometer (for compounds 6 and 9-12)
or an Enraf-Nonius CAD4 diffractometer (for compounds 3,
5, and 8) equipped with a Mo-target X-ray tube (λ ) 0.71073
Å). The structure was resolved with direct methods and refined
with full-matrix least-squares using the SHELX-9759 software
package.
The deviation from the average PCAH plane was measured
as the distance of the farthest carbon atom perpendicular to
the average PCAH plane, which was calculated by a least-
squares method59 from pertinent PCAH carbons coordinates.
In Figure 2 for 10, for example, the average PCAH plane was
computed from the 19 PCAH carbon atoms (i.e., C1-C19);
whereas the deviation from the PCAH plane was the vertical
distance between C6 and the PCAH plane; C6 was farthest
form the PCAH plane than others. The angle of distorted six-
membered ring planes was determined by the dihedral angle
between two neighbor six-membered ring planes sharing a
common bond.
Experimental Section
Forced Reaction of Carbon Tetrachloride with So-
dium. A clean piece of freshly cut sodium (ca. 3 g) and CCl4
(25 mL) were placed in a stainless steel autoclave (40 mL
capacity). Sodium stored in paraffin was first cleaned on a
filter paper and skinned to remove the covering sodium oxide.
The sealed autoclave was heated in an oven to 300 °C for 40
h. This was allowed to cool to the room temperature for
uncapping which was immediately followed by the addition of
toluene (30 mL). A small puff of yellow fume that was assumed
to be the chlorine gas gave off when the reactor was opened.
The mixture quenched by toluene was allowed to stand in the
air for evaporation; the remaining crude product was washed
several times with water to leach out NaCl. The filtered solid
was dried at 100 °C and sublimed at 160-200 °C to give a
white crystalline mixture (150 mg) of perchlorobenzene and
some 2 and 3. The remaining solid was divided into two equal
parts of organic mixtures (10 g total). The first part was
triturated with toluene (20 mL) and filtered through a 0.45
µm pore filter to remove carbon black and other inorganic
materials. The second part was treated similarly with ethyl-
benzene (20 mL). The toluene and ethylbenzene extracts were
analyzed with HPLC-UV/MS; a HPLC spectrum of the
toluene extract is shown in Figure S1 (Supporting Informa-
tion).
The products were isolated by preparative chromatography,
and the weight yields (here and all others) were calculated on
the basis of the total crude. An aliquot (2 mL) of the toluene
extract was taken up on an alumina column and eluted with
cyclohexane to afford impure fractions. These impure fractions
each containing a major component were rechromatographed
in the similar mode repeatedly to obtain small amounts of pure
products. After a series of chromatography, the isolated weight,
normalized to the total crude, in a typical experiment were 2
(20 mg), 3 (15 mg), 4 (2 mg), 10 (40 mg), 9 (1 mg), and 11 (15
mg). One-half of the ethylbenzene extract (10 mL) was chro-
Acknowledgment. This work was supported by the
National Natural Science Foundation of China (Grant
No. 20371041 and 20273052), the Ministry of Science
and Technology of China (Grant No. 2002CCA01600),
and the Ministry of Education of China (03096). The
authors thank Mr. Hua-Ping Li for his generous help
in structural computations.
Supporting Information Available: Description of start-
ing materials and analytical methods; analytical and spectro-
scopic data for compounds 5-12; typical HPLC-UV chromato-
gram of the toluene-soluble products and ORTEP drawing for
the structures of 5-CS2, 8, 9, 11, and 12. X-ray crystallographic
structural data (CIF) for compounds 11 and 12. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO048190+
(59) Sheldrick, G. M. SHELX-97, Program for Crystal Structure
Analysis (Release 97-2); University of Go¨ttingen: Germany, 1997.
J. Org. Chem, Vol. 70, No. 4, 2005 1407