A spontaneous fragmentation: From the Criegee zwitterion to coarctate Mobius aromaticity

Article abstract of DOI:10.1002/(SICI)1521-3773(19980803)37:13/14<1850::AID-ANIE1850>3.0.CO;2-B

The extremely fast fragmentation of the spiroozonides prepared from formaldehyde O-oxide and three-membered ring ketones proceeds via the coarctate transition state 1. The ozonides decompose at temperatures as low as -90°C to form carbon dioxide, alkene/alkyne, and formaldehyde (the topology of the structure of 1 is depicted on the right). The mechanism is in accordance with the rules for the stereochemical course of coarctate reactions.

Full text of DOI:10.1002/(SICI)1521-3773(19980803)37:13/14<1850::AID-ANIE1850>3.0.CO;2-B

COMMUNICATIONS
Keywords: combinatorial chemistry ´ peptides ´ peptidomi-
metics ´ receptors ´ sulfonamides
A Spontaneous Fragmentation: From the
Criegee Zwitterion to Coarctate Möbius
Aromaticity**
Christian Berger, Christian Bresler, Ulrich Dilger,
Daniel Geuenich, Rainer Herges,* Herbert Röttele,
and Gerhard Schröder*
[1] S. C. Zimmerman, W. Wu, Z. Zeng, J. Am. Chem. Soc. 1991, 113, 196 ±
201; R. Boyce, G. Li, H. P. Nestler, T. Suenaga, W. C. Still, ibid. 1994,
116, 7955 ± 7956; S. R. Labrenz, J. W. Kelly, ibid. 1995, 117, 1655 ±
1656; Y. Cheng, T. Suenaga, W. C. Still, ibid. 1996, 118, 1813 ± 1814;
H. Wennemers, S. S. Yoon, W. C. Still, J. Org. Chem. 1995, 60, 1108 ±
1109; C. Gennari, H. P. Nestler, B. Salom, W. C. Still, Angew. Chem.
1995, 107, 1894 ± 1896; Angew. Chem. Int. Ed. Engl. 1995, 34, 1765 ±
1768.
Dedicated to Professor William von E. Doering
on the occasion of his 80th birthday
Coarctate reactions[1] are defined as those in which two
bonds are simultaneously formed and broken at one or more
atoms. Like the pericyclic reactions, they form an independent
and coherent class of concerted reactions. Rules are available
for predicting the stereochemical course of coarctate reac-
tions,[1] similar to the Woodward ± Hoffmann rules for peri-
cyclic reactions. We now describe an unusual fragmentation
reaction which is in accordance with these stereochemical
rules and for which such a coarctate transition state was
confirmed by theoretical calculations.
[2] D. W. P. M. Löwik, S. J. E. Mulders, Y. Cheng, Y. Shao, R. M. J.
Liskamp, Tetrahedron Lett. 1996, 37, 8253 ± 8256.
[3] K. Kurz, M. G. Göbel, Helv. Chim. Acta 1996, 79, 1967 ± 1979.
[4] D. B. A. de Bont, G. H. Dijkstra, J. A. J. den Hartog, R. M. J. Liskamp,
Bioorg. Med. Chem. Lett. 1996, 6, 3035 ± 3040.
[5] This hinge was successfully applied in the synthetic receptors of Still
et al. see for example a) S. S. Yoon, W. C. Still, Tetrahedron 1995, 51,
567 ± 578; b) Y. Shao, W. C. Still, J. Org. Chem. 1996, 62, 6086 ± 6087;
c) Z. Pan, W. C Still, Tetrahedron Lett. 1996, 37, 8699 ± 8702; d) M.
Torneiro, W. C. Still Tetrahedron 1997, 53, 8739 ± 8750.
[6] U. Nagel, E. Kinzel, J. Andrade, G. Prescher, Chem. Ber. 1986, 119,
3326 ± 3343.
Tropone ethylene acetal (1) is in equilibrium with the
norcaradiene derivative 1a. Above 1008C 1 undergoes
[7] J. Skarzewski, A. Gupta, Tetrahedron Assym. 1997, 8, 1861 ± 1867.
[8] See reference [5a].
[9] R. J. Pieters, I. Huc, J. Rebek, Jr., Chem. Eur. J. 1995, 1, 183 ±
192.
1
sigmatropic rearrangements (Ea ˆ 23.9 kcalmol ).[2] More-
over, above 1108C 1 decomposes via 1a to give carbon
[10] F. Diederich, D. R. Carnanague, Helv. Chim. Acta 1994, 77, 800 ± 818.
[11] It was difficult to functionalize the hindered secondary hydroxy
function in the diazabicyclononane part of the ligands with a suitable
anchor molecule to which the dye had to be attached. Ultimately, the
use of methyl p-bromomethylbenzoate in the presence of potassium
hydride and 18-crown-6 furnished the hinge part of 6.
[12] AA1±3: Gly, d- and l-Ala, d- and l-Val, d- and l-Leu, d- and l-Phe, d-
and l-Pro, d- and l-Ser, d- and l-Thr, d- and l-Asn, d- and l-Gln, d-
and l-Asp, d- and l-Glu, d- and l-His, d- and l-Lys, d- and
l-Arg.
1
dioxide, benzene, and ethene (Ea ˆ 31.6 kcalmol ).[2] The
fragmentation of 1a was interpreted as a chelotropic cyclo-
reversion to give benzene and 2-carbena-1,3-dioxolane. The
authors speculate that the decomposition of the 2-carbena-
1,3-dioxolane and the chelotropic cycloreversion might be
part of a single concerted step.[2] The parent structural
element in the fragmentation of 1a is the 4,7-dioxaspiro[2.4]-
heptane 2. If a C ± O bond in 2 is replaced by a O ± O bond (3),
[13] We attempted to reach a concentration of the receptor slightly lower
than that required for binding approximately 1% of the library beads.
These concentrations are listed in Table 1.
[14] A bissulfonamide derivative of the diazabicyclonane system gave rise
to a chair ± boat shape, whereas a bistrifluoroacetyl derivative gave
rise to a chair± chair shape: P. H. McCabe, N. J. Milne, G. A. Sim, J.
Chem. Soc. Chem. Commun. 1985, 625 ± 626
[15] M. H. J. Ohlmeyer, R. N. Swanson, L. W. Dillard, J. C. Reader, G.
Asouline, R. Kobayashi, M. Wigler, W. C. Still, Proc. Natl. Acad. Sci.
USA 1993, 90, 10922 ± 10926.
[*] Prof. Dr. R. Herges, Dipl.-Chem. D. Geuenich
Institut für Organische Chemie der Technischen Universität
Hagenring 30, D-38106 Braunschweig (Germany)
Fax: (‡49)531-391-5388
E-mail: r.herges@tu-bs.de
Prof. Dr. G. Schröder, Dr. C. Berger, Dipl.-Chem. C. Bresler,
Dr. U. Dilger, Dr. H. Röttele
Institut für Organische Chemie der Universität
Kaiserstrasse 12, D-76128 Karlsruhe (Germany)
Fax: (‡49)721-608-4825
[**] We grateful to Prof. H. J. Schäfer and Dipl.-Chem. M. Letzel,
Universität Münster, for help with the Kolbe electrolysis during the
synthesis of 22.
1850
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3713-1850 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 13/14

COMMUNICATIONS
In the reaction of 7 with 6 in [D14]methylcyclohexane,[8] the
decrease in the methylene proton signal of 5 at d ˆ 4.35 (s,
2H) corresponds to the increase in the olefin proton signal
at d ˆ 5.3 (s, 2H) of trans-di-tert-butylethene (9). In the
reaction of 10 with 6 in [D10]diethyl ether/CD2Cl2 (8/1), the
decrease in the signal for the o,o'-protons of 10 at d ˆ 8.05
(sym. m, 4H) is in agreement with the increase in the signal
for the o,o'-protons of the tolane 12 at d ˆ 7.53 (sym. m, 4H).
The reaction of 13 with 6 in [D10]diethyl ether can be
unambiguously monitored by means of the signals of the a-
methylene protons of the n-propyl groups of the starting
material 13 [d ˆ 2.62 (t, 4H)] and of the product 15 [d ˆ 2.1 (t,
4H)]. Decrease and increase of these signals are in agree-
ment.
the activation barrier for a concerted fragmentation should be
lowered considerably relative to 1 or 2. The bond energy of a
C ± O bond is about 85 kcalmol 1,[3] and that of an O ± O bond
only 40 kcalmol 1.[4] Derivatives of 3 are generally accessible
by 1,3-dipolar cycloaddition of formaldehyde O-oxide (6) to
three-membered ring ketones under ozone-free conditions.[5a]
The primary ozonide 5, generated from ozone and ketene
diethyl acetal (4) at 1108C was used as a source of 6.
Compound 5 decomposes between 90 and 808C in a [3‡2]
cycloreversion to give 6 and inert diethyl carbonate (Ea ˆ
1
1
12.5 kcalmol , DH=ˆ 11.5 kcalmol ).[5b]
The three-membered ring ketones 7, 10, 13, 16, 19, and 22
are known. They were added at 1108C to the primary
When 22 was treated with 6 in [D14]methylcyclohexane/
[D8]toluene (2/1), we could not assign any 1H NMR signals to
3,3,6,6-tetramethylcyclohex-1-yne (24) between
90 and
808C. This is in agreement with the literature.[9] The
behavior of 24 under the reaction conditions (in the presence
of peroxide) was not investigated. In the presence of
tetraphenylcyclopentadienone, no Diels ± Alder product was
formed. Above 908C the decrease in the methylene signal
of 5 at d ˆ 3.9 (s, 2H) was accompanied by an increase in a
signal at d ˆ 4.88 (s, 2H). After about 15 min a relative
intensity maximum was reached at
858C (about 15%,
relative to the four methylene protons of diethyl carbonate at
208C). At 808C the intensity of the signal at d ˆ 4.88
decreases faster than that of the signal of 5 at d ˆ 3.9. We
interpret the former as unambiguous evidence for the
intermediate spiroozonide 23, the fragmentation of which
forms the highly strained alkyne 24. Therefore, 23 is more
stable than the spiroozonides 8, 11, and 14.
The substantially lower thermal stability which was ex-
pected for 3 relative to 2 was proved to be correct. The
spiroozonides 8, 11, 14, 17, and 20 with the structural elements
of 4,5,7-trioxaspiro[2.4]heptane (3) and of a 4,5,7-trioxa-
spiro[2.4]hept-1-ene undergo spontaneous fragmentation be-
tween 90 and 808C.[10] The fragmentation of the spiro-
ozonides fulfils the formal criteria for reactions with a
coarctate transition state. In the course of the reaction two
C ± C bonds of the three-membered ring are broken at the
central spiro C atom, and two C ± O bonds are formed to
generate CO2. Eight electrons are involved in the reaction
(the mechanism can be formally written with four electron-
pushing arrows). A topological analysis shows that the
transition state of the reaction is stabilized by coarctate
Möbius aromaticity. According to the stereochemical rules of
coarctate reactions, the p systems of the two terminators must
be orthogonal in the transition state (Figure 1). This is
realized in an ideal manner in the spiro arrangement of the
three- and five-membered rings in the spiroozonides 8, 11, 14,
17, 20, and 23. In our nomenclature[11] the fragmentation
should be denoted as a [p2‡c2‡p4] coarctate reaction.
1
ozonide 5, which was identified in the H NMR spectrum at
1108C by means of the methylene signal of the 1,2,3-
trioxolane ring at d ˆ 4.35 (s, 2H, in [D14]methylcyclohexane),
3.9 (in [D14]methylcyclohexane/[D8]toluene 2/1), and 4.68 (in
[D10]diethyl ether). The [3‡2] cycloreversion of 5, which
begins at 908C, can be monitored by following the decrease
in intensity of this signal. We suggest[6] that 6 reacts with 7, 10,
13, 16, 19, and 22 to form the spiroozonides 8, 11, 14, 17, 20,
and 23, which with the exception of 23 (see below) decompose
spontaneously to carbon dioxide, formaldehyde, and the
alkene 9 or the alkynes 12, 15, 18, and 21. Carbon dioxide can
be extracted from the reaction mixture with a stream of
nitrogen and precipitated as barium carbonate by passing it
through a solution of barium hydroxide. Formaldehyde could
not be detected by H NMR spectroscopy.[7] The character-
1
The fragmentation of cyclopent-3-ene-1-one spiroozonide
(1,2,4-trioxaspiro[4.4]non-7-ene) is homologuous to that of
the cyclopropanone spiroozonide 3 (4,5,7-trioxaspiro[2.4]hep-
tane). According to the stereochemical rules of coarctate
reactions involving ten electrons, this reaction must proceed
via a flat transition state, which can not be realized in the spiro
ization of the hydrocarbon products 9, 12, 15, 18, and 21 is
described in the Experimental Section. The reactions of 7, 10,
13, and 22 with 6, which was generated from 5, were
monitored by low-temperature H NMR spectroscopy. Re-
sults for the temperature range 90 to 808C are discussed
1
below.
Angew. Chem. Int. Ed. 1998, 37, No. 13/14
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3713-1851 $ 17.50+.50/0
1851

COMMUNICATIONS
1
11.3 kcalmol was calculated for the fragmentation of the
1
spirocyclopropane, and DH=ˆ 16.4 kcalmol for the frag-
mentation of the spirocyclopropene. As expected, the reac-
tions are strongly exothermic: DH8 ˆ 85.2 and
1
84.8 kcalmol , respectively.
Alternatively, the following two-step reaction is conceiv-
able: In the first step the olefin or alkyne is eliminated, and in
the second step the carbene (1,2,4-trioxolane-3-ylidene) frag-
ments to give CO2 and CH2O. We did not find a stationary
point corresponding to the carbene at any of the above-
mentioned levels of theory. The corresponding structures
fragment without a barrier to give CO2 and CH2O. The two-
step mechanism therefore can be excluded with a high degree
of certainty.
Figure 1. Orbital basis (left) and topology (right) of the transition state of
the spiroozonide fragmentation. The orbital basis contains eight electrons
in eight orbitals. The spiro C atom (coarctate center) participates in two
basis orbitals, because two bonds are formed and broken simultaneously at
this atom. The topology of orbital overlap is equivalent to a coarctate
Möbius strip. For the resulting consequences for the stereochemical course
of the reaction, see the text.
Table 1. Energies of the starting materials, transition states, and products of the fragmentation of both
parent systems of the spiroozonide fragmentation calculated at the B3LYP/6-31 ‡ G* level.[a]
system. As a consequence, a coarctate
fragmentation was not observed in this
system.[10]
We performed quantum-chemical
calculations with semiempirical, ab
initio, and density functional theory
(DFT) methods[12] (PM3,[13] QCISD/6-
31G, B3LYP/6-31G, B3LYP/6-31 ‡
G*[14, 15]) on the parent systems of the
reactions presented above. We found
such a transition state at all levels of
theory. The transition states were
characterized as first-order stationary
points by harmonic frequency calcula-
tions (one imaginary frequency per
transition state), and they were con-
Eabs [a.u.]
ZPE [a.u.]
381.56811
0.100789
0
±
381.54647
0.09079
1
397.5
11.3
381.69250
0.089423
0
±
85.2
380.30805
0.076021
0
±
380.27665
0.070694
1
440.1
16.4
380.43233
0.065092
0
±
Nimag
nÄ [cm
1
]
1
Erel [kcalmol
]
0.0
0.0
84.8
[a] Eabs ˆ absolute energy, ZPE ˆ zero-point energy, Nimag ˆ number of imaginary frequencies
according to a normal coordinate analysis, nÄ ˆ wavenumber of the imaginary frequency, Erel ˆ ener-
gy relative to the starting materials.
firmed to be saddle points of a concerted fragmentation by
intrinsic reaction coordinate (IRC) calculations. Figure 2
The back reaction of the spontaneous fragmentation is a
termolecular [p2‡c2‡p4] reaction, which will be difficult to
realize unless preorganization of the components and high
pressure are used.
Experimental Section
Procedure for the low-temperature 1H NMR experiments: An excess of
ozone was passed into a solution of ketene diethyl actetal 4 in a NMR tube
at 1108C through a capillary. The precooled gas stream was adjusted such
that single gas bubbles passed through the solution. After ozonization, a
precooled stream of nitrogen was passed through the NMR tube until a
moistened KI/starch paper remained colorless. A storage bottle, in which a
solution of the three-membered ring ketone was precooled to 1108C, was
connected to the glass joint on the top of the NMR tube. The contents were
discharged by nitrogen pressure into the NMR tube, which was immedi-
1
ately introduced into the precooled H NMR measuring cell.
7 ‡ 6: 4 (14.1 mg, 0.121 mmol) in [D14]methylcyclohexane (0.5 mL), 7
(7.4 mg, 0.044 mmol) in [D14]methylcyclohexane (0.4 mL); 10 ‡ 6:
4
Figure 2. Transition state structures of both parent systems of the
spiroozonide fragmentation calculated at the B3LYP/6-31 ‡ G* level with
bond lengths [Š] and angles [8].
(20.8 mg, 0.171 mmol) in [D10]diethyl ether (0.45 mL), 10 (7.4 mg,
0.036 mmol) in [D10]diethyl ether (0.35 mL) and CD2Cl2 (0.1 mL); 13 ‡ 6:
4
(23.5 mg, 0.20 mmol) in [D10]diethyl ether (0.45 mL), 13 (7.0 mg,
0.05 mmol) in [D10]diethyl ether (0.45 mL); 22 ‡ 6: 4 (18.2 mg, 0.157 mmol)
in [D14]methylcyclohexane (0.45 mL), 22 (8.2 mg, 0.05 mmol) in [D14]meth-
ylcyclohexane (0.12 mL) and [D8]toluene (0.33 mL).
shows the geometry of the two transition states, and Table 1
lists the energies of the starting materials, transition states,
and products for the highest and most reliable theoretical
level of the calculations. The transition states are not sym-
metrical. The calculated course of the reaction is not
The [3‡2] cycloaddition of 6 with the three-membered ring ketones on a
preparative scale is similar to the low-temperature 1H NMR procedure.
Ozonization was carried out in cold traps with an additional side flask.[5]
A
precooled stream of ozone was passed through the solution of 4 at 1108C
until the mixture turned blue. Excess ozone was expelled with a precooled
synchronous but concerted.[16]
A
barrier of DH=ˆ
1852
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3713-1852 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 13/14

COMMUNICATIONS
stream of nitrogen until a moistened KI/starch paper remained colorless.
The solution of the ketone was transferred with a pipette into the side flask.
After temperature equalization (5 min), the cold solution was discharged
into the main flask with nitrogen pressure. During warm up above 658C,
a gentle stream of nitrogen was passed through the solution. In all cases,
above 608C we observed slight and above 508C pronounced precip-
itation of barium carbonate on passing the gas through an aqueous solution
of barium hydroxide.
[1] R. Herges, Angew. Chem. 1994, 106, 261 ± 283; Angew. Chem. Int. Ed.
Engl. 1994, 33, 255 ± 276.
[2] T. Fukunaga, T. Mukai, Y. Akasaki, R. Suzuki, Tetrahedron Lett. 1970,
2975 ± 2978.
[3] T. L. Cottrell, The Strengths of Chemical Bonds, 2nd ed., Butter-
worths, London, 1958; L. Pauling, The Nature of the Chemical Bond,
3rd Ed., Cornell University Press, Ithaca, NY. 1960.
[4] R. D. Bach, P. Y. Ayala, H. B. Schlegel, J. Am. Chem. Soc. 1996, 118,
12758 ± 12765.
[5] a) R. Reiser, C. Süling, G. Schröder, Chem. Ber. 1992, 125, 2493 ± 2501,
b) C. Süling, Dissertation, Technische Universität, Karlsruhe, 1992.
[6] Under the same reaction conditions, 6 reacts with tropone and 2-, or
2,7-substituted tropones to form the corresponding spiroozonides,
which can be characterized by low-temperature 1H NMR spectro-
scopy (ca. 808C). Therefore, it is plausible to assume dipolarophilic
behavior of the three-membered ring ketones towards 6. The
spiroozonides of the tropones fragment at 60 to ‡508C. The half-
lives are informative: C. Berger, S. Dietrich, U. Dilger, D. Geuenich,
H. Helios, R. Herges, P. Kirchmer, H. Röttele, G. Schröder, Angew.
Chem. 1998, 110, 1954 ± 1957; Angew. Chem. Int. Ed. 1998, 37, 1854 ±
1856.
[7] The spiroozonide derived from 2,7-di-tert-butyltropone and 6 is
crystalline. The half-life t1/2 in CD3OD at 258C is 35 min. The
spontaneous fragmentation must proceed via the valence-isomeric
norcaradiene derivative. During the reaction, a signal at d ˆ 4.6
appears in the 1H NMR spectrum whose rate of increase is in
agreement with the signals of o-di-tert-butylbenzene. We assign it to
the methylene protons of the hemiacetal CD3OCH2OD(H).
[8] A [3‡2] cycloaddition of 6 and 7 can only be performed in solvents of
low polarity.
7 ‡ 6: 4 (132.4 mg, 1.14 mmol) in pentane (7 mL), 7 (63.0 mg, 0.37 mmol) in
pentane (7 mL). The solvent was removed at 208C and 200 mbar on a
rotary evaporator. The residue was purified by chromatography (SiO2;
pentane/diethyl ether, 20/1) to give 9 (13.6 mg, 0.1 mmol, 26%). The 1H
NMR data were in agreement with the literature data.[17] A yield of 51%
was determined by 1H NMR spectroscopy of the crude product (benzene as
standard).
10 ‡ 6: 4 (132.9 mg, 1.14 mmol) in diethyl ether (6 mL), 10 (47.2 mg,
0.23 mmol) in diethyl ether (5 mL) and CH2Cl2 (3 mL); for the workup, see
above; the pressure was reduced to about 1 mbar. The residue was purified
by chromatography (SiO2, CH2Cl2) to give colorless crystals (23.6 mg
0.13 mmol, 58%) of the tolane 12, m.p. 638C (methanol). The yield of the
crude product was 76%, as determined by 1H NMR spectroscopy with
toluene as standard.
13 ‡ 6: 4 (149.0 mg, 1.28 mmol) in diethyl ether (7 mL), 13 (42.4 mg,
0.31 mmol) in diethyl ether (5 mL). The yield of the crude product was
1
67%, as determined by H NMR spectroscopy with benzene as standard.
The 1H NMR data of 15 are in agreement with those of an independently
synthesized sample.
16 ‡ 6: 4 (165.3 mg, 1.42 mmol) in diethyl ether (6 mL), 16 (63.0 mg,
0.46 mmol) in diethyl ether (6 mL). The yield of the crude product was
[9] W. Sander, O. L. Chapman, Angew. Chem. 1988, 100, 402 ± 403;
Angew. Chem. Int. Ed. Engl. 1988, 27, 398 ± 399.
1
83%, as determined by H NMR spectroscopy with benzene as standard.
The 1H NMR data of 18 are in agreement with the literature data.[18]
[10] We synthesized the spiroozonides 1,2,4-trioxaspiro[4.4]non-7-ene and
1,2,4-trioxaspiro[4.6]undeca-7,9-diene from cyclopent-3-en-1-one and
cyclohepta-3,5-dien-1-one, respectively, and 6. The half lives in
[D14]methylcyclohexane at 988C are 26 and 23 min, respectively. In
the first case cyclopent-3-en-1-one is the predominant product, in the
second case a complex product mixture is formed. Carbon dioxide and
butadiene or hexa-1,3,5-triene could not be detected.
16 ‡ 6 ‡ tetraphenylcyclopentadienone: 4 (192.0 mg, 1.65 mmol) in pen-
tane (13 mL), 16 (72.9 mg, 0.54 mmol) in pentane (6 mL) and toluene
(2.5 mL). At 208C tetraphenylcyclopentadienone (205.9 mg, 0.54 mmol) in
CH2Cl2 (8 mL) was added, and the mixture stirred for 48 h. After the usual
workup, the product was purified by chromatography (Al2O3, activity
level I, petroleum ether (b.p. 100 ± 1408C)/benzene (7/3)) as colorless
crystals (166 mg, 0.36 mmol, 67%), m.p. 222 ± 2238C (pentane/diethyl
[11] R. Herges, J. Chem. Inf. Comput. Sci. 1994, 43, 91 ± 106; according to
this nomenclature the epoxidation of alkenes with peracids is also a
[p2‡c2 ‡ p4] coarctate reaction.
1
ether, 2/1; 224 ± 2258C[19]). The H NMR data of the cyclooctanotetraphe-
nylbenzene are in agreement with the literature data.[19]
19 ‡ 6: 4 (241.0 mg, 2.07 mmol) in pentane (14 mL) was ozonized at
1108C. Then 19 (62.7 mg, 0.51 mmol) in pentane (2 mL) and toluene
(6 mL) was added, and the reaction flask was immersed in a cold bath at
788C. After 15 min tetraphenylcyclopentadienone (200 mg, 0.52 mmol)
in CH2Cl2 (8 mL) was introduced into the reaction flask from a separate
flask with N2 pressure. The mixture was kept for 48 h at 788C, worked up
as usual, and purified by chromatography as described above. Colorless
crystals (120 mg, 0.27 mmol, 52%), m.p. 221 ± 2228C (petroleum ether);
222 ± 2238C.[20a] The 1H NMR data of the cycloheptanotetraphenylbenzene
are in agreement with the literature data.[20b]
[12] W. Kohn, A. D. Becke, R. G. Parr, J. Phys. Chem. 1996, 100, 12974 ±
12980.
[13] J. J. Stewart, J. Comput. Chem. 1989 10, 209 ± 221.
[14] A. D. Becke, J. Chem. Phys. 1993, 98, 5648 ± 5652.
[15] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[16] A synchronous mechanism is not expected because the starting
materials are not symmetrical. The trioxolane ring of the starting
materials is not flat, and the bonds of the three-membered ring are not
equal in length.
[17] T. W. Wallace, I. Wardell, K.-D. Li, P. Leeming, A. D. Redhouse, S. R.
Challand, J. Chem. Soc. Perkin Trans. 1 1995, 2293 ± 2308.
[18] P. Caubere, G. Coudert, Bull. Soc. Chim. Fr. 1973, 3067 ± 3070.
[19] E. V. Dehmlow, M. Lissel, Liebigs Ann. Chem. 1980, 1 ± 13.
[20] a) G. Wittig, J. Meske-Schüller, Justus Liebigs Ann. Chem. 1968, 711,
65 ± 75; b) H. Dürr, G. Scheppers, Justus Liebigs Ann. Chem. 1970,
734, 141 ± 154.
22 ‡ 6: 4 (214 mg, 1.83 mmol) in pentane (8 mL), 22 (75.3 mg, 0.46 mmol)
in pentane (2 mL) and toluene (6 mL). There were no 1H NMR signals that
could be assigned to 3,3,6,6-tetramethyl-1-(3,3,6-trimethyl-hept-6-en-1-
ynyl)cyclohexene (isomer of the dimer of 24)[21] in the spectrum of the
crude product.
22 ‡ 6 ‡ tetraphenylcyclopentadienone: 4 (239.5 mg, 2.06 mmol) in pen-
tane (8 mL), 22 (84.4 mg, 0.51 mmol) and tetraphenylcyclopentadienone
(202.4 mg, 0.53 mmol) both in pentane (2 mL) and toluene (6 mL). The
mixture was kept for 15 min at 788C and worked up as usual. There are
no signals in the 1H NMR spectrum that could be assigned to the Diels ±
Alder product. According to a control experiment, tetraphenylcyclopenta-
dienone is virtually inert towards 6.
[21] C. N. Bush, D. E. Applequist, J. Org. Chem. 1977, 42, 1076 ± 1079.
Received: May 30, 1997
Supplemented version: March 12, 1998 [Z10494IE]
German version: Angew. Chem. 1998, 110, 1951 ± 1954
Keywords: aromaticity ´ fragmentations ´ ozonolysis ´ spiro
compounds ´ transition states
Angew. Chem. Int. Ed. 1998, 37, No. 13/14
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3713-1853 $ 17.50+.50/0
1853