Thermolysis of the benzene anion radical 18-crown-6 complex

Article abstract of DOI:10.1021/jo980700a

The C-O and G-H bonds of 18-crown-6 are activated when 18-crown-6 is complexed with the potassium salt of the benzene anion radical. Evacuated glass bulbs containing the solid anion radical salt of potassium 18-crown-6 benzene anion radical were plunged into a bath at 320 C, resulting in mini-explosions and generating a series of compounds including dioxane, 2-methyl1,3-dioxolane, divinyl ether, hydrogen, methane, and 15-crown-5. Deuterium labeling studies proved that all of these compounds originated from the 18-crown-6. Further, these labeling studies were an aid in discerning the mechanism of the decomposition. Benzene, 1,4-cyclohexadiene, and cyclohexene were also generated. The last two originated from the reaction of the anion radical of benzene with hydrogen.

Full text of DOI:10.1021/jo980700a

7694
J . Org. Chem. 1998, 63, 7694-7697
Th er m olysis of th e Ben zen e An ion Ra d ica l 18-Cr ow n -6 Com p lex
Cheryl D. Stevenson* and Grant Morgan
Department of Chemistry, Illinois State University, Normal, Illinois 61790-4160
Received April 15, 1998
The C-O and C-H bonds of 18-crown-6 are activated when 18-crown-6 is complexed with the
potassium salt of the benzene anion radical. Evacuated glass bulbs containing the solid anion
radical salt of potassium 18-crown-6 benzene anion radical were plunged into a bath at 320 °C,
resulting in mini-explosions and generating a series of compounds including dioxane, 2-methyl-
1,3-dioxolane, divinyl ether, hydrogen, methane, and 15-crown-5. Deuterium labeling studies proved
that all of these compounds originated from the 18-crown-6. Further, these labeling studies were
an aid in discerning the mechanism of the decomposition. Benzene, 1,4-cyclohexadiene, and
cyclohexene were also generated. The last two originated from the reaction of the anion radical of
benzene with hydrogen.
In tr od u ction
in the automerization mechanism.6 The extreme interest
in the high-temperature behavior ofaromaticmolecules,6-15
is in contrast with the analogous lower temperature
behaviors of their anion radicals3,5 motivated us to
investigate the thermolysis of the anion radical of the
most common of all aromatic systems: ben zen e.
Unfortunately, the solid (solvent-free) anion radical
salt of benzene is not stable toward return of the electron
back to the alkali metal.16a This problem can be circum-
vented by addition of 18-crown-6 to the system.16 The
encapsulation of the metal cation in the crown ether
thermodynamically disfavors electron return and renders
The addition of a single electron to the π cloud of
polyaromatic hydrocarbons (PAHs) activates the C-C
and C-H bonds, rendering their alkali-metal solid anion
radical salts susceptible to thermal decomposition at
temperatures just above 100 °C.1,2 This low-temperature
thermal decomposition liberates H2 and methane along
with the formation of dimers, trimers, and higher poly-
mers and is analogous to the very high temperature py-
rolysis of PAHs where H2 is also liberated3 and fullerenes
are formed.4 Hence, PAHs can be polymerized only at
temperatures above the softening point of Pyrex glass,
while thermolysis following single-electron reduction is
the only method for the low-temperature polymerization
of PAHs.
(5) (a) Stevenson, C. D.; Garland, P. M.; Batz, M. L. J . Org. Chem.
1996, 61, 5948. (b) Stevenson, C. D.; Batz, M. L.; Garland, P. M.; Reiter,
R. C.; Sanborn, M. D. J . Org. Chem. 1997, 62, 2045.
(6) (a) Scott, L. T.; Roelofs, N. H.; Tsang, T. H. J . Am. Chem. Soc.
1987, 109, 5456. (b) Scott, L. T.; Hashemi, M. M.; Schultz, T. H.;
Wallace, M. B. J . Am. Chem. Soc. 1991, 113, 9692.
(7) Zimmermann, G.; Nuchter, M.; Hopf, H.; Ibrom, K.; Ernst, L.
Liebigs Ann. Chem. 1996, 1407.
(8) Automerizations are isomerizations that are degenerate in the
absence of an substituent or isotopic label: Balaban, A. T.; Farcasiu,
D. J . J . Am. Chem. Soc. 1967, 89, 1958.
(9) (a) Poupko, R.; Muller, B. K.; Krieger, I. C.; Zimmermann, H.;
Luz, Z. J . Am. Chem. Soc. 1996, 118, 8015. (b) Poupko, R.; Zimmer-
mann, H.; Muller, B. K.; Luz, Z. J . Am. Chem. Soc. 1996, 118, 7995.
(10) (a) Doering, W. v. E.; Roth, W. R. Angew. Chem., Int. Ed. Engl.
1963, 2, 115. (b) Schro¨der, G. Chem. Ber. 1964, 97, 3140. (c) Merenyi,
R.; Oth, J . F. M.; Schro¨der, G. Chem. Ber. 1964, 97, 3150. (d) Seburg,
R. A.; Patterson, E. V.; Stanton, J . F.; McMahon, R. J . J . Am. Chem.
Soc. 1997, 119, 5847.
(11) Lawrence Scott has published a 15-paper series titled Thermal
Rearrangements of Aromatic Compounds. The papers addressing the
automerization of aromatic compounds are listed below and in ref 6.
(a) Scott, L. T.; Agopian, G. K. J . Am. Chem. Soc. 1977, 99, 4506. (b)
Scott, L. T.; Highsmith, J . R. Tetrahedron Lett. 1980, 21, 4703. (c) Scott,
L. T. Acc. Chem. Res. 1982, 15, 52. (d) Scott, L. T.; Kirms, M. A.; Berg,
A.; Hansen, P. E. Tetrahedron Lett. 1982, 23, 1859. (e) Scott, L. T.;
Tsang, T.-H.; Levy, L. A. Tetrahedron Lett. 1984, 25, 1661. (f) Scott,
L. T.; Roelofs, N. H. J . Am. Chem. Soc. 1987, 109, 5461. (g) Scott, L.
T.; Roelofs, N. H. Tetrahedron Lett. 1988, 29, 6857.
The hydrogen results from an anionic polymerization
of the anion radical, and the methane is believed to come
from the formation of an intermediate carbene followed
by its capture of H2.2,5 The proposed mechanism, involv-
ing carbenes, is consistent with Scott’s and co-worker’s
results concerning their 900 °C pyrolytic automerization
of naphthalene, regarding which they conclude “the
thermal automerization of naphthalene probably occurs
by reversible formation of benzofulvene, either via car-
benes or by direct dyotropic rearrangements.”6 However,
recent experiments by Hopf et al.7 showed that benzene
will automerize at temperatures as low as 800 °C when
hydrogen serves as the carrier gas. Their work strongly
suggests that the mechanism actually involves a radical-
driven process where I as opposed to II or III is involved
(12) (a) Scott, L. T.; Kirms, M. A. J . Am. Chem. Soc. 1981, 103, 5875.
(b) Becker, J .; Wentrup, C.; Katz, E.; Zeller, K.-P. J . Am. Chem. Soc.
1980, 102, 5110.
(13) Barry, M.; Brown, R. F. C.; Eastwood, F. W.; Gunawardana,
D. A.; Vogel, C. Aust. J . Chem. 1984, 37, 1643.
(14) Zeller, K.-P. Angew. Chem. 1982, 94, 448. (b) Zeller, K.-P.
Angew. Chem., Int. Ed. Engl. 1982, 21, 440.
(15) Cioslowski, J .; Piskorz, P.; Moncrieff, D. J . Am. Chem. Soc.
1998, 120, 1695.
(1) Stevenson, C. D.; Espe, M. P. J . Org. Chem. 1985, 50, 4289.
(2) Stevenson, C. D.; Rice, C. V.; Garland, P. M.; Clark, B. K. J .
Org. Chem. 1997, 62, 2193.
(3) (a) Tilicheev, M. D.; Shchitikov, V. K. Chem. Zentrbl. 1937, 4583.
(b) Kinney, C. R.; del Bel, E. Ind. Eng. Chem. 1954, 46, 548.
(4) Taylor, R. Langley, G. J .; Kroto, H. W.; Walton, D. R. M. Nature
1993, 366, 728.
(16) (a) Stevenson, C. D.; Sturgeon, B. E.; Vines, S. K.; Peters, S. J .
J . Phys. Chem. 1988, 92, 6850. (b) Weissmann, S. I.; Komarynsky, M.
A. J . Am. Chem. Soc. 1975, 97, 1589. (c) Nelson, G. V.; von Zelewsky,
A. J . Am. Chem. Soc. 1975, 97, 6279.
10.1021/jo980700a CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/16/1998

The Benzene Anion Radical 18-Crown-6 Complex
J . Org. Chem., Vol. 63, No. 22, 1998 7695
K+{18C6}C6H6•-(solid) kinetically and thermodynami-
cally stable under inert conditions.16 We did not antici-
pate thermolysis products arising from the crown ether
itself, as the electron resides totally on the benzene
system and 18-crown-6 and 18-crown-6 containing alkali-
metal cations resist decomposition at the temperatures
of our thermolysis studies.17
Resu lts a n d Discu ssion
In a typical experiment involving 1.9 millimol (50.7 mg)
of solid potassium 18-crown-6 benzene anion radical salt
heated at 320 °C for 30 s, 0.73 millimol of noncondensable
gases, which passed through a liquid-nitrogen-cooled
U-tube were evolved. A small amount (4.4 mg) of light
condensable liquids were found in the U-tube, and 41.0
mg of a black sooty polymeric material was left in the
reaction flask.
F igu r e 1. GC trace of the low-molecular-weight materials
produced by the pyrolysis of potassium 18-crown-6 benzene
anion radical. These compounds were caught in the liquid-
nitrogen-cooled U-tube.
The noncondensable gases resulted in a pressure of
10.75 millibar in the vacuum apparatus. Upon exposure
of the gas to the Cu/CuO furnace, the pressure dropped
by 59.6% (to 4.33 millibar), and the condensable product
of the reaction in the furnace proved to be water (CuO +
H2 f H2O + Cu). Thus, 59.6% (0.435 millimol or 0.87
mg) of the original gas was hydrogen. The hydrogen
generated from the pyrolysis was also exposed to a stirred
mixture of cyclopentene containing palladium on acti-
vated carbon in toluene. This resulted in the formation
of cyclopentane. FT-IR analysis of the evolved gas proved
that the remaining 40.4% (0.29 millimol or 4.6 mg) was
methane.
The amount of condensable material, noncondensable
gases, and polymeric soot is sufficient to provide mass
balance. Further, this experiment, using various amounts
of solid potassium 18-crown-6 benzene anion radical salt,
has been performed more than 20 times with nearly
identical results. Hence, this reaction is very reproduc-
ible.
When the same pyrolysis was carried out with per-
deuterated benzene replacing benzene, a similar mixture
of gases was obtained. Exposure of these noncondensable
gases to the cyclopentene mixture did not yield the
expected dideuterated cyclopentane but again resulted
in cyclopentane. Further, the resulting methane proved
to be CH4 with no detectable traces of deuterated meth-
ane. Surprisingly, both the hydrogen and the methane
apparently came from the 18-crown-6.
Condensable material was found in the liquid-nitro-
gen-cooled U-tube resulting from the pyrolysis of
K+{18C6}C6H6•-(solid). GC analysis proved this mixture
to contain of six constituents (Figure 1). The most
abundant constituent (retention time 2.41 min) proved
to be benzene. This must have resulted from ejection of
its antibonding electron. Consistent with this was the
presence of both 1,4-cyclohexadiene and cyclohexene in
the mixture (retention times 2.51 and 2.62 min), which
must have resulted from the reaction of the benzene
anion radical with H2 (reaction 1).2,18 When K+{18C6}-
clohexene. In fact, these two materials and perdeuter-
ated benzene were the only products found to contain
deuterium resulting from any of the K+{18C6}C6D6•-(solid)
studies, as all of the remaining products resulted from
the decomposition of the crown ether.
Evidently the high thermal energy coupled with the
Coulombic pull of the potassium cation on the odd
electron results in the expected ejection of the electron
from the benzene. However, before it can be recaptured
by K+, the electron is “intercepted” by the crown ether.
The electron capture by the crown is probably due to the
electropositive nature of the oxygen atoms that is induced
by the encapsulated potassium cation. The crown, now
containing an extra electron, is in the form of a very short
lived anion radical. The 18-crown-6 anion radical is a σ
radical, which is known to cleave to an anion and a neu-
tral radical.19 Further, ether linkages in anion radicals
are known to cleave in a similar manner (Scheme 1).20
All of the products except benzene and the hydroge-
nated benzenes (Figure 1) result from the dismember-
ment of the 18-crown-6 (Scheme 1). The use of perdeu-
terated benzene did not alter the isotopic composition of
any of these decomposition products, as evidenced by both
GC-mass spectral (Figure 1) and 1H NMR analysis.
Attack of the methylene radical on the second oxygen
followed by cleavage of the carbon-second oxygen bond
results in dioxane, which is found at retention time 2.98
min in Figure 1. If water is simultaneously lost, the
result is the formation of divinyl ether, which was also
found (retention time of 1.46 min in Figure 1). The peak
at retention time 2.29 min (Figure 1) proved to result
from 2-methyl-1,3-dioxolane, which simply results from
1,2-hydrogen migration followed by attack of the methine
carbon radical upon the oxygen atom. Note that the
fragment left after the departure of dioxane, 2-methyl-
1,3-dioxolane, or divinyl ether can undergo an analogous
loss of another molecule.
No water was found in the reaction products. How-
ever, it would react with the potassium hydride (formed
via reaction 1) and any potassium metal to yield hydrogen
(17) 18-Crown-6 samples that were dissolved in tetrahydrofuran
saturated in potassium iodide, followed by solvent removal, resisted
any thermal decomposition at 320 °C.
(18) Ichikawa, M.; Soma, M.; Onishi, T.; Tamaru, K. Bull. Chem.
Soc. J pn. 1970, 43, 3672.
(19) Kimura, N.; Takamuku, S. J . Am. Chem. Soc. 1995, 117, 8023.
(20) (a) Guthrie, R. D.; Przemyslaw, M. J . Am. Chem. Soc. 1986,
108, 2628. (b) Bartak, D. E.; Koppang, M. D.; Woolsey, N. F. J . Am.
Chem. Soc. 1985, 107, 4692.
C6D6•-(solid) was substituted for K+{18C6}C6H6•-(solid),
the analogous products proved to be 1,2,3,4,5,6-hexadeu-
terio-1,4-cyclohexadiene and 1,2,3,4,5,6-hexadeuteriocy-

7696 J . Org. Chem., Vol. 63, No. 22, 1998
Stevenson and Morgan
Sch em e 1
the same products were obtained from the 18-crown-6
and the CH4 and H2 were found to come from the crown
as opposed to the naphthalene.
The addition of electrons to polyaromatic hydrocarbons
results in thermal C-C and C-H bond activation;1,2
however, if 18-crown-6 is present, there is more bond
activation in the crown than in the PAH. The remark-
able affinity that the macrocyclic ethers have for metal
cations is of fundamental importance in coordination
chemistry, and the presence of a crown ether assists the
electron transfer from a metal to an acceptor (M + crown
+ A f M+-crown-A•-). The thermally induced return
of the electron is, on the other hand, a much more
involved process.
It has been previously observed by Bickelhaupt and
co-workers21 that (2-methoxy-1,3-xylylene)-15-crown 4
undergoes a rather efficient ether cleavage in the pres-
ence of organomagnesium reagents at 100 °C (reaction
2). This reaction does not involve cleavage of the outer
perimeter of the crown ether structure, but it does show
that the ether linkages can be activated in the presence
of organic anions. Richey et al.22 did, however, observ
cleavage of one of the short bridges of 2,1,1-cryptand to
vinyl and (metalated) hydroxyethyl groups, when solu-
tions of the cryptand were exposed to dialkylmagnesium
compounds. It seems likely that we are observing an
analogous activation of the ether linkages in the 18-
crown-6 in the presence of the very basic benzene anion
radical.
gas, which is observed. Once the crown ether is broken,
some electron return to the exposed K+ would be ex-
pected, thus providing a mechanism for the generation
of potassium metal.
The liberation of methane and hydrogen were expected,
as these are typically the major gas-phase products
produced during anion radical pyrolysis.2,5,23 The hy-
drogen is lost upon polymerization of the anion radical.
The new carbon-carbon bonds replace the carbon-
hydrogen bonds in the monomer.2,5,23 The methane is
believed to come from the hydrogenation of a radical
The granular black sooty material, which remained in
the reaction flask after the pyrolysis, was exposed to I2.
The fast atom bombardment mass spectra and laser
desorption mass spectra of the I2-quenched soot shows
the presence of only high-molecular-weight material.
After the addition of I2 (in ether) and treatment with
sodium thiosulfate to remove the excess iodine, the soot
was extracted with toluene for 12 h in a Soxhlet ap-
paratus. Submission of this toluene extract to laser
desorption mass spectral analysis revealed the presence
of a vast number of high-molecular-weight polymeric
compounds with prominent m/e values of 1104, 856, and
582. These heavy materials are expected from the
formation of polyethers. Closure of the radical produced
after the initial dioxane formation (Scheme 1) would
produce some 15-crown-5. Indeed, prominent peaks at
m/e 220 (the molecular weight of 15-crown-5) and 204
(loss of O) suggests the presence of 15-crown-5 in the
toluene extracts.
It should be noted that the pyrolysis described above
at 320 °C can be carried out under milder conditions with
similar results but with fewer condensable products.
Benzene makes up a larger portion of the condensable
products at lower pyrolysis temperatures. When the
pyrolysis is carried out for 20 s at 200 °C, benzene and a
small amount of cyclohexene (in a 17:1 ratio) are the only
products caught in the U-tube.
•
(•CH2CH2O- - - + H2 ) CH4 + CH2O- - -) or via the
capture of hydrogen by a carbene intermediate.2,5,23
Exp er im en ta l Section
The anion radical salts were generated under high-vacuum
conditions via reduction of neutral benzene on a freshly
distilled potassium mirror. The benzene, which was dried over
sodium-potassium alloy, served as the solvent.15 After com-
plete reduction, the dark-colored anion radical solutions were
poured into an evacuated round-bottom Pyrex bulb, and excess
benzene was removed under reduced pressure while the
contents of the bulb were stirred magnetically. The remaining
solid anion radical salt was then exposed to an open vacuum
for several more hours to complete the solvent removal.
The pyrolysis studies were carried out by plunging the bulb
containing the solid anion radical salts into a hot (320 °C) sand
bath. This procedure results in an apparent explosion of the
salt and the generation of a puff of smoke. The bulb containing
(21) (a) Gruter, G. M.; van Klink, G. P. M.; Akkerman, O. S.;
Bickelhaupt, F. Organometallics 1993, 12, 1180. (b) Gruter, G. M.; van
Klink, G. P. M.; Akkerman, O. S.; Bickelhaupt, F. Organometallics
1991, 10, 2535.
(22) Squiller, E. P.; Whittle, R. R.; Richey, H. G., J r. Organometallics
1985, 4, 1154.
The experiments described above were repeated using
naphthalene and perdeuterated naphthalene in place of
the benzene, and identical results were obtained. Again
(23) Stevenson, C. D.; Garland, P. M.; Batz, M. L. J . Org. Chem.
1996, 61, 5948.

The Benzene Anion Radical 18-Crown-6 Complex
J . Org. Chem., Vol. 63, No. 22, 1998 7697
the pyrolyzed material was then connected to the high-vacuum
system. The temperature of the sand bath was monitored with
a nickel-chromium vs nickel-aluminum thermocouple by
Omega (-200-1250 °C).
The noncondensable (at liquid-nitrogen temperatures) gases
were exposed to a 300 °C Cu-CuO furnace. This resulted in
a pressure drop due to the oxidation of the H2 to form water,
which was collected in a U-tube. The water produced in the
furnace was caught in the adjoining U tube, which was
immersed in liquid N2. H2 and D2 discrimination was carried
explosively toward air prior to I2 reoxidation. This explosive
nature is probably due to the presence of highly reactive
anionic living polymers.23 After reoxidation, the ether and I2
were removed under reduced pressure, and the solids were
extracted with toluene. Both the toluene-soluble and toluene-
insoluble materials were submitted to matrix-assisted laser
desorption ionization mass spectroscopy (MALDI-MS) analysis.
The liquid-nitrogen-cooled U-tube was sealed from the line
on both sides. The tube was then broken open under chilled
conditions, and either CDCl3 (for NMR analysis) or dodecane
(for GC-mass spectral analysis) was added. The NMR spectra
were recorded on a Varian Mercury 400 MHz instrument, and
the GC mass spectral data were collected on a Hewlett-
Packard HP 5890 GC system connected to a Hewlett-Packard
5973 mass selective detector.
out by replacing the NMR tube with
a bulb containing
cyclopentene with the hydrogenation catalyst. In separate
experiments mixtures of methane and hydrogen were injected
into the system to verify that the hot CuO would selectively
oxidize the H2, leaving the CH4. This vacuum system was
fitted with a 10 cm gas IR cell, allowing us to take an IR
(Nicolet 5SXC FTIR spectrometer) sample at any time.
After completion of the pyrolysis, I2 and ether were distilled
into the bulb containing the reaction mixture. This was done
to oxidize any reactive anionic material. It should be noted
that the black sooty material generated via the pyrolysis reacts
Ack n ow led gm en t. We thank the National Science
Foundation (Grant CHE-9617066) for support of this
work.
J O980700A